Network Design Fundamentals

Network Design Fundamentals 1.
Network design fundamentals
dalmazio.mandich@alcatel-lucent.it -1- All rights reserved © 2008

Network Design Fundamentals 1.0
INDEX OF CONTENTS
1. GEODESY ...................................................................................................................... - 7 -
1.1. ORTHODROMIC DISTANCE BETWEEN TWO POINTS OF TERRESTRIAL SURFACE. .................... - 7 -
1.2. Haversine formula .......................................................................................................... - 7 -
1.3. Coordinates conversion ..................................................................................................- 10 -
2. DIGITAL RADIO LINKS DIMENSIONING CRITERIA .............................................................- 16 -

2.1. General ........................................................................................................................- 16 -
2.2. The radio refractive index and its gradient ........................................................................- 19 -
2.3. Flat fading outage .........................................................................................................- 43 -
2.4. The fading occurrence factor P0 .......................................................................................- 45 -
2.5. Fading margin ...............................................................................................................- 55 -
2.6. The selective fading outage ............................................................................................- 61 -
2.7. Reflection .....................................................................................................................- 66 -
2.8. Frequency re-use system outage .....................................................................................- 75 -
2.9. Unprotected channel global outage ..................................................................................- 81 -
2.10. The protected channel....................................................................................................- 82 -
2.11. Two diversity receiver : ITU method ................................................................................- 86 -
2.12. Four receivers diversity ..................................................................................................- 88 -
2.13. Diversity techniques for frequency re-used systems ....................................................................- 90 -
2.14. MULTICHANNEL SYSTEMS ..............................................................................................- 90 -
2.15. General guidelines .........................................................................................................- 91 -
3. RAIN ATTENUATION ......................................................................................................- 95 -

3.1. Introduction ..................................................................................................................- 95 -
3.2. ITU-R 837-2 .................................................................................................................- 95 -
3.3. Rain attenuation characteristics.......................................................................................- 97 -
3.4. Rain outage ..................................................................................................................- 99 -
3.5. General guidelines .......................................................................................................... - 102 -
4. QUALITY .................................................................................................................... - 103 -

4.1. International standards: from ITU-T G.821 to ITU-T G.828 ............................................... - 103 -
4.2. Medium independent and radio specific Recs. .................................................................. - 105 -
4.3. ITU-T G.821 ............................................................................................................... - 106 -
4.4. ITU-T G.826 and G.828 ................................................................................................ - 109 -
4.5. G826 – G 828 Error Performance Objectives .................................................................. - 114 -
4.6. G.826/8 - National/International Portion of the End-to-End Path ....................................... - 115 -
4.7. ITU-T G. 821 vs. ITU-T G. 826/8 ................................................................................... - 116 -
4.8. ITU-T G. 826-8 Related Recommendations ..................................................................... - 117 -
4.9. Annex 1: Relationship between PDH path performance monitoring and the block-based
parameters (G.826 / Annex B - extract) .................................................................................. - 132 -

5. UNAVAILABILITY ......................................................................................................... - 133 -

5.1. Introduction ................................................................................................................ - 133 -
5.2. Unavailability definition ................................................................................................ - 134 -
5.3. ITU-T G. 821 Unavailability evaluation ........................................................................... - 135 -
5.4. ITU-T G.827 (03/2000) superseded .............................................................................. - 136 -
5.5. ITU-T G.827 (09-2003) ................................................................................................ - 145 -
5.6. Some considerations about path topologies .................................................................... - 151 -
5.7. Reliability ................................................................................................................... - 155 -
6. REFERENCES ............................................................................................................. - 156 -
7. US PATH AVAILABILITY CALCULATION METHOD .............................................................. - 157 -

INDEX OF FIGURES
Fig. 1 – Parameters affecting quality and availability evaluations ................................... - 17 -
Fig. 2 – Flat and selective fading .............................................................................. - 17 -
Fig. 3 – Received spectrum affected by selective fading ............................................... - 18 -
Fig. 4 – Measured dry(blue), wet(red) end global refractivity (green line) values ............. - 20 -
Fig. 5 – Surface water vapor density ( winter ) ........................................................... - 21 -
Fig. 6 - Surface water vapor density ( spring )............................................................ - 21 -
Fig. 7 - Surface water vapor density ( summer ) ......................................................... - 22 -
Fig. 8 - Surface water vapor density ( autumn ) ......................................................... - 22 -
Fig. 9 – Standard atmosphere .................................................................................. - 23 -
Fig. 10 – IGRA stations layout .................................................................................. - 24 -
Fig. 11 - IGRA stations with available data on at least one day of the year (dashed line) and
at least 80% of the days in the year (solid line). ................................................... - 25 -
Fig. 12 – G (dN/dh) classes ..................................................................................... - 27 -
Fig. 13 – Flat earth model ....................................................................................... - 27 -
Fig. 14 – Typical propagation condition vs K ............................................................... - 28 -
Fig. 15 – M classes ................................................................................................. - 29 -
Fig. 16 – Measured N (on the left) and modified refractivity M (on the right) vs elevation . - 29 -
Fig. 17 – G relevant to first km ( February ) ............................................................... - 30 -
Fig. 18 - G relevant to first km ( May ) ...................................................................... - 30 -
Fig. 19 - G relevant to first km ( August ) .................................................................. - 31 -
Fig. 20 - G relevant to first km ( November ) ............................................................. - 31 -
Fig. 21 - G statistics for the lowest 100 m ( February ) ................................................ - 32 -
Fig. 22 - G statistics for the lowest 100 m (May)......................................................... - 32 -
Fig. 23 - G statistics for the lowest 100 m (August) ..................................................... - 33 -
Fig. 24 - G statistics for the lowest 100 m (November) ................................................ - 33 -
Fig. 25 – ITU ( red dots ) and “ITU enhanced” ( black line ) refractivity data template...... - 35 -
Fig. 26 – Focusing and defocusing ............................................................................ - 36 -
Fig. 27 Blackout measurements. N (abscissa) vs elevation (ordinates) ........................... - 37 -
Fig. 28 – Received field levels during blackout ............................................................ - 38 -
Fig. 29 – Duct measurements. N (abscissa) vs elevation (ordinates) .............................. - 39 -
Fig. 30 – Definition of parameters describing a) Surface, b) Elevated Surface and c)
Elevated ducts ................................................................................................. - 40 -
Fig. 31 – Refractivity and ducts data report template .................................................. - 41 -
Fig. 32 – Typical behavior of RX level depending on propagation environment................. - 42 -
Fig. 33 – P0........................................................................................................... - 43 -

Fig. 34.................................................................................................................. - 44 -
Fig. 35 – Fading statistics ( P0 =0.1) ......................................................................... - 44 -
Fig. 36 - Fading statistics ( P0 =10) .......................................................................... - 45 -
Fig. 37 – P0 value (4 GHz , S = 42 m ) ..................................................................... - 46 -
Fig. 38 – Milano Pero – Caronno regression line .......................................................... - 47 -
Fig. 39 – Abbate Argento – Bari regression line .......................................................... - 47 -
Fig. 40 – Probability G < -100 NU (February) ............................................................. - 50 -
Fig. 41 - Probability G < -100 NU (May) ................................................................... - 51 -
Fig. 42 - Probability G < -100 NU (August) ................................................................ - 51 -
Fig. 43 - Probability G < -100 NU (November) ........................................................... - 52 -
Fig. 44 – Measured P0 : hilly terrain near the sea, 4 GHz 60 km .................................... - 55 -
Fig. 45 – Measured P0 : Mediterranean Sea , 2 GHz 140 km ......................................... - 55 -
Fig. 46 – Gas attenuation for different relative humidity values ..................................... - 59 -
Fig. 47 - ATPC behavior versus fading ...................................................................... - 61 -
Fig. 48 – Selective fading probability vs. P0 ................................................................ - 62 -
Fig. 49 – Measured signature with ALCATEL Ka and Kb parameters ( Ts=symbol time , τ =
6.3 ns) ............................................................................................................ - 63 -
Fig. 50 - ...............................................................................................................- 75 -
Fig. 51 – X Polar Interference Reduction Factor .......................................................... - 80 -
Fig. 52 – High XPD antenna (right) and antenna with high XPD (left) comparison ............ - 81 -
Fig. 53 – Frequency diversity configuration ................................................................ - 83 -
Fig. 54 – Space diversity configuration ...................................................................... - 83 -
Fig. 55 – Hybrid diversity configuration ..................................................................... - 83 -
Fig. 56 – Space and frequency diversity with 2 receivers.............................................. - 83 -
Fig. 57 – Space and frequency diversity with 4 receivers.............................................. - 84 -
Fig. 58 – Space and frequency diversity with 4 receivers and 4 antennas ....................... - 84 -
Fig. 59 - 3+1 Space and frequency diversity with 4 receivers and 4 antennas ................. - 84 -
Fig. 60 – Diversity improvement............................................................................... - 85 -
Fig. 61 – Four receiver diversity configuration ............................................................ - 89 -
Fig. 62 - Rain outage % vs. AR depending on latitude ................................................ - 100 -
Fig. 63 – ITU reference for tropical rain attenuation .................................................. - 101 -
Fig. 64 – ............................................................................................................. - 101 -
Fig. 65 – Hypothetical Reference Digital Path (HRDP) ITU-T G. 821 apportionment ........ - 107 -
Fig. 66 – SDH frame structure ............................................................................... - 110 -
Fig. 67 – BER vs. errored blocks relationship ............................................................ - 113 -
Fig. 68 – Quality parameters flow chart ................................................................... - 114 -

Fig. 69 - ITU-T G. 826/828 apportionment criteria .................................................... - 116 -

Fig. 70 - ITU-T G. 826/828 apportionment criteria .................................................... - 116 -
Fig. 71 – HRDP national portion.............................................................................. - 122 -
Fig. 72 – Mono-directional unavailability .................................................................. - 134 -
Fig. 73 – Bi-directional unavailability ....................................................................... - 134 -
Fig. 74 - ITU-T G.827 : type a Path......................................................................... - 136 -
Fig. 75- ITU-T G.827 : type b Path ......................................................................... - 137 -
Fig. 76 – ITU-T G.827 : availability ratio depending on PCE ........................................ - 139 -
Fig. 77 – ITU-T G.827 : Values for mean time between digital path outage calculations .. - 140 -
Fig. 78 – A comparison between ITU-T G.827 and ITU-R F. 1492 availability ratio
parameters .................................................................................................... - 141 -
Fig. 79 – A comparison between ITU-T G.827 and ITU-R F.1492 for mean time between
digital path outage .......................................................................................... - 142 -
Fig. 80 – International PCE .................................................................................... - 145 -
Fig. 81 – ITU-T G. 827 : Unavailability Ratio ( 1 – AR ) Objectives ( S=standard, H=high)- 148 -
Fig. 82 - F. 1703 : AR objectives for international portion .......................................... - 149 -
Fig. 83 - F. 1703 : AR objectives for national portion : Red (long haul), green (short haul)
and blue (Access) ........................................................................................... - 149 -
Fig. 84 – F. 1703 : OI objectives for international portion .......................................... - 150 -
Fig. 85 – F. 1703 OI objectives for national portion : Red (long haul), green (short haul) and
blue (Access) ................................................................................................. - 151 -
Fig. 86 – Linear topology network ........................................................................... - 152 -
Fig. 87 – Redundant topology network .................................................................... - 152 -

1. Geodesy
1.1. ORTHODROMIC DISTANCE BETWEEN TWO POINTS

OF TERRESTRIAL SURFACE.
Let the points be: A(lat1, lon1) and B(lat2, lon2)
Δlat = lat1 - lat2 Δlon = lon1 - lon2
⎛ Δlat ⎞ ⎛ Δlon ⎞
a = sin² ⎜ ⎟ + cos(lat1 )*cos(lat2 )*sin² ⎜ 2 ⎟ (1)
⎝ 2 ⎠ ⎝ ⎠
⎛ a ⎞
d(A,B) = 2*R*atn ⎜
⎜ 1-a ⎟⎟
⎝ ⎠
1.2. Haversine formula

R = earth’s radius (mean radius = 6,371km)
Δlat = lat2 - lat1 Δlon = lon2 - lon1
2 2
⎛ Δlat ⎞ ⎛ Δlon ⎞
a = sin ⎜ ⎟ + cos(lat1 )*cos(lat2 )*sin ⎜ ⎟ (2)
⎝ 2 ⎠ ⎝ 2 ⎠
c = 2*atan2 ( a, 1-a ) d = R*c
The Haversine formula remains particularly well-conditioned for numerical computa-

tion even at small distances. (R W Sinnott in Sky and Telescope, 1984; the ‘half-
versed-sine’ is (1-cosθ)/2, or sin²(θ/2) ). Since the earth is actually roughly ellip-
soidal with a radius varying between about 6,378km (equatorial) and 6,357km (po-
lar) there are small errors in using spherical geometry. This means that errors from
assuming spherical geometry might be up to about 0.5%, worst case, depending on
latitude and direction of travel.
lat1+lat2 b2
mean latitude l = a = 6378 b = 6357 e = 1-
2 a2
local radius of curvature travelling N/S is
R' = a*
(1 - e²)
3
(3)
(1 - e².sin²l) 2
travelling perpendicular to that is

a
R'' =
(1 - e².sin²l)
You could then take a weighted average of the two local radii of curvature depend-

ing on direction of travel. Errors would then be negligible over small distances ( be-
low 0.02% at distances of 1,000km ).
Δlon=lon1 -lon2
(4)
Azimuth = atan2 ⎡⎣-sin ( Δlon) *cos(lat2 ), cos(lat1 )*sin(lat2 )-sin(lat1 )*cos(lat2 )* cos(Δlon)⎤⎦
The problem of determining the great circle distance on a sphere has been around
for hundreds of years, as have both the Law of Cosines solution (not recommend-
ed) and the Haversine formula (2). An UNRELIABLE way to calculate distance on a
spherical Earth is the Law of Cosines for Spherical Trigonometry
a =sin (lat1 ) * sin (lat2 ) b = cos (lat1 ) * cos (lat2 ) * cos (lon2 - lon1 )
c = acos ( a + b ) d=r*c (5)
d = r * acos ⎡⎣sin (lat1 ) * sin (lat2 ) +cos (lat1 ) * cos (lat2 ) * cos (lon2 - lon1 ) ⎤⎦
Latitude r Latitude r
0 110.57 50 111.23
5 110.58 55 111.32
10 110.61 60 111.41
15 110.65 65 111.49
20 110.70 70 111.56
25 110.77 75 111.62
30 110.85 80 111.66
35 110.94 85 111.69
40 111.04 90 111.69
45 111.13
Under this assumption the antenna bearings are

⎡ sin (lon2 -lon1 ) ⎤
AzimuthA = atn ⎢ ⎥
⎢⎣ cos (lat1 ) tan (lat2 ) − cos (lat1 ) cos (lon2 -lon1 ) ⎥⎦
(6)
⎡ sin (lon1 -lon2 ) ⎤
AzimuthB = atn ⎢ ⎥
⎢⎣ cos (lat2 ) tan (lat1 ) − sin (lat2 ) cos (lon1 -lon2 ) ⎥⎦
If the coordinates site A, bearing θ and great circle distance d are known , B coordi-
nates are
⎡ ⎛d ⎞ ⎛ d ⎞⎤
lat2=asin ⎢cos (θ ) * sin ⎜ ⎟ * cos (lat1 ) + sin (lat1 ) * cos ⎜ ⎟ ⎥
⎣ ⎝r⎠ ⎝ r ⎠⎦
⎡ ⎛d⎞ ⎤
⎢ cos ⎜ r ⎟ − sin (lat1 ) * sin (lat2 ) ⎥
lon2 = lon1 ± acos ⎢ ⎝ ⎠ ⎥ (7)
⎢ cos (lat1 ) * cos (lat2 ) ⎥
⎢ ⎥
⎣ ⎦
+acos if θ<180° -acos if θ>180°

Although this formula is mathematically exact, it is unreliable for small distances

because the inverse cosine is ill-conditioned.
cos (5 degrees) = 0.996194698 ; cos (1 degree) = 0.999847695
cos (1 minute) = 0.9999999577 ; cos (1 second) = 0.9999999999882
cos (0.05 sec) = 0.999999999999971
1.2.1.1. The radius of the Earth, R

The historical definition of a "nautical mile" is "one minute of arc of a great circle of
the earth". Since the earth is not a perfect sphere, that definition is ambiguous. The
standard (SI) value for the length of a nautical mile is 1.852 km = 1.852/1.609344
international miles (that is, approximately 1.15078 miles - either "international" or
"U.S. statute"). Thus, the implied official circumference is 360 degrees times 60
minutes/degree times 1.852 km/minute = 40003.2 km. The implied radius is the
circumference / 2π: R = 6367 km = 3956 mi
The shape of the Earth is well approximated by an oblate spheroid. The radius of
curvature varies with direction and latitude and the radius of curvature of an ellip-
soidal Earth in the plane of the meridian is given by R' = a * (1 - e2) / (1 - e2 *
(sin(lat))2)3/2 where a is the equatorial radius, b is the polar radius, and e is the
eccentricity of the ellipsoid = √ (1 - b2/a2) and the radius of curvature in a plane
perpendicular to the meridian and perpendicular to a plane tangent to the surface is
given by N = a/√ (1-e2*(sin(lat)2)) .Use of the geometric mean of these two radii
of curvature for all azimuths, as it produces errors of order of magnitude 0.1% for
distances within 500 km (300 mi) at 60 degrees latitude is suggested. The formula
for that average is no more complicated than either of its components: r = √ (R' *
N) becomes r = a*√ (1-e2)/(1-e2*(sin(lat))2) . Using these formulas with a =
6378 km, b = 6357 km, e=0.081082 gives the following table of values for the Ra-
dii of Curvature:
latitude r R' N
00 degrees 6357 km 6336 km 6378 km
15 degrees 6360 km 6340 km 6379 km
30 degrees 6367 km 6352 km 6383 km
45 degrees 6378 km 6367 km 6389 km
60 degrees 6388 km 6383 km 6394 km
75 degrees 6396 km 6395 km 6398 km
90 degrees 6399 km 6399 km 6399 km
Note that the radius of curvature for an ellipsoid is not the same as the distance

from the surface of the ellipsoid to the center. In fact, the radius of curvature in-
creases as the radius decreases. Also, be aware that a variety of ellipsoids with
slightly different parameters have been fit to the Earth; the preferred ellipsoid may
depend on the region in which you are most interested. ( 1.3.2 )
1.3. Coordinates conversion

This module must support the following coordinate conversions:
o Conversions between geodetic, geocentric, and local Cartesian coordinate
systems,
o Conversions between geodetic coordinates (latitude and longitude ) and vari-
ous types of map projection coordinates (easting and northing in linear units),
o Conversions between geodetic coordinates and Military Grid Reference Sys-
tem (MGRS) or World Geographic Reference System (GEOREF) grid coordinates,
o Transformations of geodetic or geocentric coordinates between different
global or local horizontal datums,
o Transformations between ellipsoid heights and geoid (or local MSL) heights
o Conversions that combine two or more of the above operations.
1.3.1. Coordinate systems

o Lambert conic conformal
o Lambert conic near-conformal
o Mercator
o Cassini-Solder
o Transverse Mercator
o Oblique Mercator
o Stereographic
o New Zealand map grid
o Tunisia Mining Grid
o American polyconic
1.3.2. Reference Ellipsoids

Ellipsoid Semi-major 1/flattening
dalmazio.mandich@alcatel-lucent.it - 10 - All rights reserved © 2008

axis
Airy 1830, 6377563.396 299.3249646
Modified Airy 6377340.189 299.3249646
Australian National 6378160 298.25
Bessel 1841 (Namibia) 6377483.865 299.1528128
Bessel 1841 6377397.155 299.1528128
Clarke 1866, 6378206.4 294.9786982
Clarke 1880, 6378249.145 293.465
Everest (India 1830) 6377276.345 300.8017
Everest (Sabah Sarawak) 6377298.556 300.8017
Everest (India 1956) 6377301.243 300.8017
Everest (Malaysia 1969) 6377295.664 300.8017
Everest (Malay. & Sing) 6377304.063 300.8017
Everest (Pakistan) 6377309.613 300.8017
Modified Fischer 1960 6378155 298.3
Helmert 1906 6378200 298.3
Hough 1960 6378270 297
Indonesian 1974 6378160 298.247
International 1924 6378388 297
Krassovsky 1940 6378245 298.3
GRS 80 6378137 298.257222101
South American 1969 6378160 298.25
WGS 72 6378135 298.26
WGS 84 6378137 298.257223563
Datum Ellipsoid Region of use
Adindan Clarke 1880 Burkina Faso; Cameroon; Ethiopia; Mali;

MEAN FOR Sudan; Senegal
Afgooye Krassovsky 1940 Somalia

Ain el Abd 1970 International 1924 Bahrain; Saudi Arabia
American Samoa 1962 Clarke 1866 American Samoa Islands
Anna 1 Astro 1965 Australian National Cocos Islands
Antigua Island Astro 1943 Antigua (Leeward Islands)
Arc 1950 Botswana; Burundi; Lesotho; Malawi; MEAN
Clarke 1880
FOR Swaziland; Zaire; Zambia; Zimbabwe
Arc 1960 Kenya; Tanzania
Ascension Island 1958 Ascension Island
International 1924
Astro Beacon E 1945 Iwo Jima

Astro DOS 71/4 St Helena Island

Astro Tern Island (FRIG) 1961 Tern Island
Astronomical Station 1952 Marcus Island
Australian Geodetic 1966 Australia; Tasmania
Australian National
Australian Geodetic 1984 Australia; Tasmania
Ayabelle Lighthouse Clarke 1880 Djibouti
Bellevue (IGN) International 1924 Efate & Erromango Islands
Bermuda 1957 Clarke 1866 Bermuda
Bissau Guinea-Bissau
International 1924
Bogota Observatory Colombia
Bukit Rimpah Bessel 1841 Indonesia (Bangka & Belitung Ids)
Campo Inchauspe Argentina
International 1924
Canton Astro 1966 Phoenix Islands
Cape Clarke 1880 South Africa
Cape Canaveral Clarke 1866 Bahamas; Florida
Carthage Clarke 1880 Tunisia
Chatham Island Astro 1971 New Zealand (Chatham Island)
Chua Astro International 1924 Paraguay
Corrego Alegre Brazil
Dabola Clarke 1880 Guinea
Deception Island Clarke 1880 Deception Island; Antarctia
Djakarta (Batavia) Bessel 1841 Indonesia (Sumatra)
DOS 1968 New Georgia Islands
International 1924
Easter Island 1967 Easter Island
Estonia;Coord. System 1937 Bessel 1841 Estonia
Cyprus; Egypt ; England; Channel Islands;
Scotland; Shetland Islands ; Finland; Nor-
way; Greece; Iran; Italy (Sardinia); Italy
(Sicily); Malta; MEAN FOR Austria; Belgium;
Denmark; Finland; France; W Germany; Gi-
European 1950 International 1924 braltar; Greece; Italy; Luxembourg; Neth-
erlands; Norway; Portugal; Spain; Sweden;
Switzerland MEAN FOR Austria; Denmark;
France; W Germany; Netherlands; Switzer-
land ,MEAN FOR Iraq; Israel; Jordan; Leba-
non; Kuwait; Saudi Arabia; Syria ;Portugal;
Spain,Tunisia

MEAN FOR Austria; Finland; Netherlands;

European 1979 International 1924
Norway; Spain; Sweden; Switzerland
Fort Thomas 1955 Clarke 1880 Nevis; St. Kitts

Gan 1970 Republic of Maldives
Geodetic Datum 1949 New Zealand
International 1924
Azores (Faial; Graciosa; Pico; Sao Jorge;
Graciosa Base SW 1948
Terceira)
Guam 1963 Clarke 1866 Guam
Gunung Segara Bessel 1841 Indonesia (Kalimantan)
GUX 1 Astro Guadalcanal Island
International 1924
Herat North Afghanistan
Hermannskogel Datum Bessel 1841 (Namibia) Croatia-Serbia,Bosnia-Herzegovina
Hjorsey 1955 Iceland
Hong Kong 1963 International 1924 Hong Kong
Hu-Tzu-Shan Taiwan
Indian Everest (India 1830) Bangladesh
Indian Everest (India 1956) India; Nepal
Indian Everest (Pakistan) Pakistan
Indian 1954 Thailand
Indian 1960 Vietnam (Con Son Island)
Everest (India 1830)
Indian 1960 Vietnam (Near 16øN)
Indian 1975 Thailand
Indonesian 1974 Indonesian 1974 Indonesia
Ireland 1965 Modified Airy Ireland
ISTS 061 Astro 1968 South Georgia Islands
ISTS 073 Astro 1969 International 1924 Diego Garcia
Johnston Island 1961 Johnston Island
Kandawala Everest (India 1830) Sri Lanka
Kerguelen Island 1949 International 1924 Kerguelen Island
Kertau 1948 Everest (Malay. & Sing) West Malaysia & Singapore
Kusaie Astro 1951 International 1924 Caroline Islands
Korean Geodetic System GRS 80 South Korea
L. C. 5 Astro 1961 Clarke 1866 Cayman Brac Island
Leigon Ghana
Clarke 1880
Liberia 1964 Liberia
Luzon Clarke 1866 Philippines
M'Poraloko Gabon
Clarke 1880
Mahe 1971 Mahe Island

Massawa Bessel 1841 Ethiopia (Eritrea)

Merchich Clarke 1880 Morocco
Midway Astro 1961 International 1924 Midway Islands
Minna Cameroon; Nigeria
Montserrat Island Astro 1958 Montserrat (Leeward Islands)
Clarke 1880
Oman (Masirah Island); Saudi Arabia ;
Nahrwan
United Arab Emirates
Naparima BWI International 1924 Trinidad & Tobago
Alaska ; Bahamas ; Canada (Alberta; British
Columbia;Manitoba; Ontario; New Bruns-
wick; Newfoundland; Nova Scotia; Quebec;
Northwest Territories; Saskatchewan , Yu-
kon ) , Canal Zone , Cuba ; Greenland
North American 1927 Clarke 1866
(Hayes Peninsula), MEAN FOR Antigua;
Barbados; Barbuda; Caicos Islands; Cuba;
Dominican Republic; Grand Cayman; Jamai-
ca; Turks Islands ; MEAN FOR Belize; Costa
Rica; El Salvador ; Guatemala ; Honduras ;
Nicaragua, Mexico
North American 1983 GRS 80 Alaska , Aleutian Ids, Canada, CONUS ,
Hawaii , Mexico; Central America
North Sahara 1959 Clarke 1880 Algeria
Obs. Meteorologico 1939 International 1924 Azores (Corvo & Flores Islands)
Old Egyptian 1907 Helmert 1906 Egypt
Old Hawaiian Clarke 1866 Hawaii, Kauai, Maui; MEAN FOR Hawaii;
Kauai; Maui; Oahu
Oman Clarke 1880 Oman
Ordnance Survey Great Britain England; Isle of Man; Wales; Scotland;
Airy
1936 Shetland Islands
Pico de las Nieves Canary Islands
International 1924
Pitcairn Astro 1967 Pitcairn Island
Point 58 Burkina Faso & Niger
Clarke 1880
Pointe Noire 1948 Congo
Porto Santo 1936 Porto Santo; Madeira Islands
International 1924 Bolivia; Colombia; Ecuador; Guyana; MEAN
Provisional South Chilean 1963
FOR Chile; Colombia; Peru; Venezuela
Puerto Rico Clarke 1866 Puerto Rico; Virgin Islands
Pulkovo 1942 Krassovsky 1940 Russia
Qatar National International 1924 Qatar

Qornoq Greenland (South)

Reunion Mascarene Islands
Rome 1940 Italy (Sardinia)
S-42 (Pulkovo 1942) Krassovsky 1940 Hungary; Poland; Czechoslavakia; Latvia;
Kazakhstan; Albania; Romania
S-JTSK Bessel 1841 Czechoslavakia (Prior 1 JAN 1993)
Santo (DOS) 1965 Espirito Santo Island
Sao Braz International 1924 Azores (Sao Miguel; Santa Maria Ids)
Sapper Hill 1943 East Falkland Island
Schwarzeck Bessel 1841 (Namibia) Namibia
Selvagem Grande 1938 International 1924 Salvage Islands
Sierra Leone 1960 Clarke 1880 Sierra Leone
Argentina; Bolivia; Brazil; Chile; Colombia;
Ecuador; Ecuador (Baltra; Galapagos); Gu-
South American 1969 South American 1969
yana; Paraguay; Peru; Trinidad & Tobago;
Venezuela
South Asia Modified Fischer 1960 Singapore
Tananarive Observatory 1925 International 1924 Madagascar
Timbalai 1948 Everest (Sabah Sarawak) Brunei; E. Malaysia (Sabah Sarawak)
Tokyo Bessel 1841 Japan; Okinawa; South Korea
Tristan Astro 1968 International 1924 Tristan da Cunha
Viti Levu 1916 Fiji (Viti Levu Island)
Clarke 1880
Voirol 1960 Algeria
Wake Island Astro 1952 International 1924 Wake Atoll
Wake-Eniwetok 1960 Hough 1960 Marshall Islands
WGS 1972 WGS 72 Global Definition
WGS 1984 WGS 84 Global Definition
Yacare Uruguay
International 1924
Zanderij Suriname

2. Digital radio links dimensioning criteria
2.1. General
The performance evaluation of radio links is based on the calculations of
Ö physical parameters concerning the hop geometry
• Free space losses
• Additional losses (branching, attenuators …)
• Anomalous propagation condition ( reflection , obstruction … )
• Received field nominal level
Ö parameters characterizing the standard propagation along the hop
• Fading occurrence factor P0
• Percentage of deep or selective fading periods
Ö quality performance relevant to the unprotected channel
Ö improvement obtainable adopting diversity techniques ( 1+1 )
Ö worsening factor for a multichannel ( n+1 ) system
Performance evaluation of digital microwave radio links must be imple-

mented in terms of unavailability and quality that are representative pa-
rameters relevant to DRRS.
As far as the propagation path characteristics are concerned, rain atten-

uation predominates at frequencies >17 GHz, while multipath attenua-
tions and/or distortion predominates at frequencies < 10 GHz.
DRRS should be mainly designed in terms of unavailability at frequen-

cies above 17 GHz and error performance at frequencies below about
10 GHz, while in the range 10-17 GHz both objectives should be consi-
dered.
All kind of quality and unavailability objectives MUST be referenced by

the tool and must be included in each result summary relevant to single
hop and/or end-to-end performance

Fig. 1 – Parameters affecting quality and availability evaluations
In the most general case we can assume that the overall outage time, relevant to
propagation , for a digital microwave radio link will result from the composition of
the following effects:
Ö Flat fading ( including rain and gas attenuation ) generating outages related
to Signal/Noise degradation
Ö Selective fading generating outages related to Inter-symbol interference
Ö Cross polar interference ( only for Frequency re-use systems )
Other causes of impairments (reflection, obstruction ..) will be analyzed later.
ATMOSPHERIC
MULTIPATH
FLAT SELECTIVE
FADING FADING
ANALOG DIGITAL ANALOG DIGITAL
THERMAL THERMAL INTERMODULATION INTERSYMBOL

NOISE NOISE INTERFERENCE
FADING EXCEEDS
MARGIN OVER DISTORSION PRODUCES
THRESHOLD EYE CLOSURE AND
LOSS OF SYNC.
Fig. 2 – Flat and selective fading

Fig. 3 – Received spectrum affected by selective fading
Although the same transmission issues affect analogue and digital sig-
nals, their reaction to frequency selective fading are quite different. Fre-
quency-selective fading can be viewed in the frequency domain and in
the time domain it is called multipath delay spread and it is defined as
the overall path delays between the first to the last pulses reaching the
receiver (the multipath echo).The formal exact definition of the selective
fading in a given environment is the limit of the hop total outage when
the power margin tends to infinite:
Selective fading outage = limMargin→∞ (total fading outage)

(8)
so the ‘flat fading’ is consequently defined as Flat fading outage=total
fading outage-selective fading outage.
The main difference between flat (slow) and selective (fast) fading is
that the second one doesn’t depend on the fading margin (or effective
link system gain)

2.2. The radio refractive index and its gradient
2.2.1. Introduction
Ö The index of refraction n is the ratio between the speeds of light in a va-
cuum and through some medium.
c
v= (9)
n
Where c = speed of light (vacuum) and v = speed of light (medium)
Ö Snell’s Law states that a ray passing from a medium of higher refractive in-
dex into a medium of less refractive index is bent away from the normal.
Ö In the real atmosphere (where the density is higher because of gas and wa-
ter molecules ) the propagation speed is slower and in a “standard” envi-
ronment atmospheric density (depending on pressure, temperature and
humidity) decreases with altitude increasing.
Ö Radio waves passing from dense air to thinner air undergo a change of di-
rection proportional to the densities difference due to increasing of propaga-
tion speed with decreasing of transmission medium density.
2.2.2. Radio refractivity1

Ö Refractive index values are so near to 1 that radio refractivity defined as N
= (n - 1) x 106 is used . N values are related to pressure, temperature and
humidity by
P e e
N = Nd + Nw =77.6 -5.6 +3.75 x 105 2
T T T (10)
P e 77.6 ⎛ e⎞
N ≈ 77.6 + 3.75 x 105 2 = P + 4810 ⎟
T T T ⎜⎝ T⎠
where Nd is the dry term, Nw the wet term and

• P atmospheric pressure [ hP or mb ]
• T temperature [ °K ]
• e vapor pressure [ hP ]
Temperature and humidity are the predominant factors for changes in

refractivity as depicted in the following Fig. 4 reporting some month of
measured N components in southern India.
1
Rec. ITU-R P.453-9 : The radio refractive index: its formula and refractivity data

Fig. 4 – Measured dry(blue), wet(red) end global refractivity (green line) values
The relationship between water vapor pressure e and relative humidity is:
⎛ a*T ⎞
⎜ ⎟
es = 6.1078 e⎝ T + b ⎠ Teten's formula (11)
where
• e = H es / 100
• H relative humidity (%)
• T Celsius temperature (°C)
• es saturation vapor pressure (hPa) at the temperature T (°C).
• a = 7.5 , b = 237.3 °C for T >0 °C
• a = 9.5 , b = 265.5 °C for T <0 °C
Moreover vapor pressure e and water vapor density ρ [gm-3] are related as e =
ρT/216.7
Representative values of ρ 2 are reported in Fig. 5 - Fig. 8
2
Rec ITU-R P.836-3 : Water vapour: surface density and total columnar content

Fig. 5 – Surface water vapor density ( winter )
Fig. 6 - Surface water vapor density ( spring )

Network De
esign Funda
amentals 1.0
0
Fig. 7 - Surfa
ace water vapo
or density ( sum
mmer )
Fig. 8 - Surfa
ace water vapo
or density ( autumn )
T
The ITU ta
ables relev
vant to ann
nual values
s of surface
e water va
apor densitty
e
exceeded f
for 1, 2, 3,
3 5, 10, 20
0, 30 and 50% of th
he year mu
ust be auto
o-
m
matically accessed
a by the tool
dalmazio.ma
andich@alcate
el-lucent.it - 22 - All rights
r reserve
ed © 2008
2.2.3. Standard atmosphere3

Assuming a sea level temperature value T0=288.15 °K , the existing relationship
between T and altitude h is
T(h)=T0+Γ*h with Γ ( T lapse rate ) = -6.5 °K/Km (12)
In the same way starting with a sea level pressure value p0=1013.25 hPa, the ex-
isting relationship between p and T versus altitude h is:
g
⎛ Γ ⋅ h ⎞ Γ ⋅R
p(z) = p0 ⋅ ⎜1 − ⎟ (13)
⎝ T0 ⎠
where
• g acceleration due to gravity = 9.81 [ms-2]
• R dry air gas constant = 287.04 [Jkg-1°K-1]
Standard Atmosphere
1100 310
950 290
temperature [K]
pressure [mbar]
800 270
650 250
500 230
350 210
200 190
0 1 2 3 4 5 6 7 8 9 10 11
height [km]
Fig. 9 – Standard atmosphere
Finally the distribution of water vapor in the atmosphere may be approximated as
⎛ -h ⎞
ρ (h)=ρ 0 exp ⎜ ⎟ (14)
⎜ ⎟
⎝ h0 ⎠
where h0 = 2 km and ρ0 = 7.5 [g m-3]
Water-vapor density decreases exponentially with increasing altitude, up to an alti-
tude where the mixing ratio e (h)/P(h) = 2*10–6. Above this altitude, the mixing ra-
tio is assumed to be constant.
3
Rec ITU-R P.835-3 Reference standard atmospheres

Under these assumptions the long-term mean dependence of n upon h is expressed
⎛ -h ⎞
by N(h) = 1 + N0 * 10-6 * exp ⎜ ⎟ (15)
⎝ h0 ⎠
N0 :average value of atmospheric refractivity extrapolated to sea level
h0 :scale height (km).
N0 and h0 can be determined statistically for different climates.
For reference purposes a global mean of the height profile of refractivity

may be defined for terrestrial path , by N0 = 315 and h0 = 7.35 km
2.2.4. Refractivity measured data

4
The tools should allow the user to access IGRA database containing a significant
number of measurements based on the following parameters
• Pressure values rounded to 1 hPa
• Temperature rounded to 0.1 °C
• Dew point depression rounded to 0.1 °C
• Wind direction and speed
Fig. 10 – IGRA stations layout
4
http://www.ncdc.noaa.gov/oa/climate/igra/index.php

Fig. 11 - IGRA stations with available data on at least one day of the year (dashed line) and at
least 80% of the days in the year (solid line).
The tool must consider that elevation is given as geo-potential height

rather then true height that is necessary for this application
The words "height" and "geo-potential height" are sometimes carelessly

interchanged. However, they are subtly different. Height (or altitude)
refers to the absolute distance above sea-level. Geopotential height is
closely related to altitude but accounts for variations of gravity within
latitude and height as reported in Annex1

SYDNEY 28 May 1999

3500
3000
2500
Heigth [m]
2000
1500
1000
500
0
0 5 10 15 20
temperature [°C]
mandatory mandatory+significant
Adopting these data we can base our calculation on figures much more
reliable of those actually available.
Basing on these data devoted analysis for reflection, obstruction and

anomalous propagation can be performed by means of suitable ray trac-
ing algorithm.
2.2.4.1. Measurement caveat

Data derived from radiosonde measurements must be carefully validated and used
considering some severe limitations related to the following items :
• Profile inaccuracy due to limited sampling due to rapid transit in the lower 100 m
• Universally synchronized sampling ( 00:00 – 12:00 UTZ)
• Large spacing between stations
• Uncertainties regarding initial conditions
• Possible errors in humidity sensing under rapidly changing conditions
• Interpolations between mandatory/significant levels instead of fixed altitudes
• Differences in accuracy and sensitivity due to different instrumentations
• Radiosondes trajectories
2.2.5. The refractivity gradient G and K factor

The refractivity gradient (G) is simply the change in N with height and denoted as
“dN/dh” or “ΔN/ Δ h”.

Network De
esign Funda
amentals 1.0
0
Fig
g. 12 – G (dN/d
dh) classes
S
Sharp chan
nges in G value
v vs. height caus
se the abno
ormal bend
ding of radio
w
waves: in a normal atmospher
a re G decrea
ases linearrly but in an
a abnormal
a
atmospher
re, G chang
ges both negatively and
a positiv
vely in a no
on linear fa
a-
s
shion.
The
e K factor, defined
d as K = 157 / (1
157 + G), allows
a the ra
adio path to
o be drawn as
a
a sttraight line on a terrain
n path profile.
Fig. 13 – Flat ea
arth model
Typ
pical K value
es are :
normal K = 4/3 ( standard atmo
osphere )
sub-standarrd K = 2/3
super-stand
dard K = infinity
Depending on K values
s different propagation
p conditions can
c be expe
erienced, as
s summarize
ed
in next Fig.. 14 , and statistical
s de
efinition of the relevan
nt probabilitties are man
ndatory for a
correct netw
work plannin
ng. Multipatth propagation, the pre
evalent caus ded transmis-
se of degrad
ms of occurrrence, resullts from anomalous attmospheric structures. A
sion reliabillity in term
statistical description of s occurring in nature is
o refractivitty gradients s illustrated
d qualitative
ely
in Fig. 14, where the curve (bla
ack line) is the interpo
olation of m
measured IT
TU data (re
ed
points) corresponding to differen
nt atmospheric structu
ures. The shape
s of th
he curve ca
an
change drastically with
h geophysic
cal location.. The central portion of
o the curve
e around th
he
gradient of 40 N-units//km, the sta
andard norm
mal refractiv
ve index gradient, reprresents linea
ar
dalmazio.ma
andich@alcate
r reserve
ed © 2008
gradients that affect terrain clearance and change the relative phases of ground-reflected
rays. The multipath fading segment is centered on a -157 N-units/km value. Multipath fad-
ing occurs when an elevated atmospheric layer with a refractive gradient near this value is
sandwiched within a more "normal" atmosphere. Generally, the probability of the daytime
propagation regime is the largest . Of all the anomalous propagation regimes, however, the
multipath fading regime has the largest probability.
Fig. 14 – Typical propagation condition vs K
2.2.6. Modified refractivity

M =Modified Refractive index = N + (h/a)*106
N = Radio Refractivity
h = height in m a.s.l.
a = Earth radius (6375 Km)
M-profiles are used to graphically depict changes in radio refractivity in the lower
atmosphere. In fact for a spherical earth with an atmosphere having a uniform re-
fractive index, rays are curved relative to the earth and this is approximately
equivalent to a flat earth with an atmosphere in which the refractive index varies
with height as h/a as described by the Modified Refractive Index.

Fig. 15 – M classes
Fig. 16 – Measured N (on the left) and modified refractivity M (on the right) vs elevation
2.2.7. Worldwide G data

Refractivity gradient (G) figures are the most important parameter to be adopted in
order to evaluate implementation criticism of every microwave link.
Unfortunately, depending on the adopted propagation modeling, different represen-
tation of this values are given , generating a great confusion in some readers.
The most important representations of this parameter will be given in the following
together with the relevant theoretical approach they’re referring to.

2.2.7.1. G relevant to the first km of the atmosphere5
Fig. 17 – G relevant to first km ( February )
Fig. 18 - G relevant to first km ( May )
5
Rec ITU-R P.453-9 The radio refractive index: its formula and refractivity data

Fig. 19 - G relevant to first km ( August )
Fig. 20 - G relevant to first km ( November )
Fig. 17-Fig. 20 were derived from 1955-1959 data set from ONLY 99 ra-
diosonde sites.
The ΔN values were not reduced to a reference surface
These data are not useful for to microwave link dimensioning considering that
Ö The ΔN values were not reduced to a reference surface
Ö Inside the first km of the atmosphere the existence of different layers cha-
racterized by different values of “local” gradients is very probable
2.2.7.2. G statistics for the lowest 100 m

Gradient probability <-100 NU/km for the lowest 100 m of the atmosphere are
used to estimate the probability of occurrence of ducting and multipath conditions

Fig. 21 - Fig. 24 were derived from 1955-1959 data set from ONLY 99
radiosondes sites
Fig. 21 - G statistics for the lowest 100 m ( February )
Fig. 22 - G statistics for the lowest 100 m (May)

Fig. 23 - G statistics for the lowest 100 m (August)
Fig. 24 - G statistics for the lowest 100 m (November)
Actually these data are not useful for to microwave link dimensioning
considering that according to the latest ITU Rec. ( ITU-R P.530-10 ) re-
fractivity gradient data are referred to the lower 65 m and they’ve been
derived from a totally different database.
This set of data can be used ONLY to evaluate P0 according to previous

ITU recommendations

2.2.7.3. G statistics for the lowest 65 m

The refractivity gradient was derived from ONLY two years (1992-1993) of initia-
lization data (4 times a day) of the numerical weather forecast of the ECMWF and
was stored in database table with a resolution of 1.5° in both latitude and longi-
tude.
The data range from 0° to 360° in longitude and from +90° to –90° in latitude.
For a location different from the grid points , the refractivity gradient at
the desired location can be derived by performing a bi-linear interpola-
tion on the values at the four closest grid points.
These data files, obtained from the Radiocommunication Bureau, must

be used by the tool to define the G statistics.
2.2.7.4. Enhanced refractivity gradient distributions6

The tool must give the possibility to enhance the G statistics evaluation obtaining
complete statistical distribution near the ground level over the lowest 65 m and
100m of the atmosphere evaluated basing on the median value GMed of the refrac-
tivity gradient and the surface refractivity value, Ns.
GMed may be computed from the probability P that the G < GDn as
GDn + 30
GMed = − 30 with E0 =Lgt ( | GDn | ) (16)
(1/ P − 1)1/E0
Equation (16) is valid for the interval –300 NU/km ≤ GDn ≤ – 40 NU/km.
For GDn ≤ GMed, the cumulative probability P1 of GDn may be obtained from:
1
P1 = E1
⎡⎛ D − Med 1.6 * B ⎞ 120 ⎤
1 + ⎢⎜ n + ⎟ ⎥
⎣⎢⎝ B 120 ⎠ B ⎦⎥
⎛ ⎞ (17)
⎜ ⎟
0.3 GMed − Ns + 210 ⎜ 2 × GDn − GMed
B= E1 = Lgt + 1⎟
2 ⎜ ⎛ B ⎞ 6.5 ⎟
⎜ ⎜ ⎟ +1 ⎟
⎝ ⎝ 67 ⎠ ⎠
Eq. (17) is valid if GMed>120 NU/km and for the interval 300 NU/km<GDn<50 NU/km.
For GDn > GMed, the cumulative probability P2 of GDn is computed as:
6

1
P2 = 1 − E1
⎡⎛ G − GMed 1.6 * B ⎞ ⎡ 100 ⎤ 2.4 ⎤
1 + ⎢⎜ Dn + ⎟ ⎥
⎣⎢⎝ B 120 ⎠ ⎢⎣ B ⎥⎦ ⎦⎥
⎛ ⎞ (18)
⎜ ⎟
0.3 GMed − Ns + 210 2 × GDn − GMed
B= E1 = Lgt ⎜ + 1⎟
2 ⎜ ⎛ B ⎞ 6.5 ⎟
⎜ ⎜ ⎟ +1 ⎟
⎝ ⎝ 67 ⎠ ⎠
Eq.(18) is valid if GMed >–120 NU/km and for the interval 300 NU/km < GDn < 50 NU.
The final result , compared with the raw ITU data , is included in
Fig. 25 – ITU ( red dots ) and “ITU enhanced” ( black line ) refractivity data template
2.2.8. NON standard atmosphere

This condition arises when some of the previous defined parameters exhibits a
“strange” behavior ( for instance T increase with h ).
2.2.8.1. Focusing and defocusing

If refractive index gradient is not constant up-fading ( focusing ) or deep fading
phenomena can happen

Fig. 26 – Focusing and defocusing
2.2.8.2. Sub-refraction
A cold front can produce a surface based sub-refractive layer due to cold advection.
Cold advection is the process in which the wind blows from a region of cold air to a
region of warmer air : it results in cooling of the warmer region. As the cold advec-
tion persists, temperatures in the warmer region will begin to decrease as the cold-
er air moves into the region of warmer air.
Sub-refraction is the abnormal upward bending of radio waves under weather con-
ditions such that the refractive index increase with the altitude.
A substandard atmosphere is produced by an abnormally high surface temperature
or increasing water vapor content with height.
The greater the refractive index, the slower the radio wave travel, hence for an in-
crease in refractive index with altitude, the upper part of the wave front is slowed,
causing rays (perpendicular to wave fronts) to curve upwards increasing earth bulg-
ing effect with possible diffraction problems
2.2.9. Super-refraction
Warm advection is the process in which the wind blows from a region of warm air to
a region of cooler air. As the warm advection persists, temperatures in the colder
region will begin to increase as the warmer air moves into the region of colder air.
A super standard atmosphere is produced by rise in temperature with increasing
height ( temperature inversion ) , or a decrease in water vapor content , or both.
Radio refractivity decreases faster than normal with height : an example of this
kind of anomalous propagation is the passage of cool air over a warm body of wa-
ter. Water evaporation increases the atmospheric humidity and the low tempera-
ture near the surface is a sign of temperature inversion.
Low temperature and high humidity increase the atmospheric density near the sur-

face causing an abnormally high downward bending of wavefront : in these condi-

tions, K = ∞
2.2.10. Blackout
During this phenomenon the received fields are characterized by very long ( up to
some hours ) absolute propagation failures that can’t be justified in terms of poor
fading margin or space diversity efficiency or antenna/equipment failures consider-
ing that the effects are the same for all the channels in both the stations.
The occurrence of blackout fading is very rare : it can be roughly esti-

mated observing the probability of a refractivity gradient < -157 NU/km
, but when it happens the effects are total and catastrophic because it’s
generally a largely frequency independent go/no-go phenomenon.
It typically occurs only late in the evening and early in the morning.
An estimation of expected probability of these events should be fore-

seen by the tool
Typical refractive index behaviors indicating blackout are included in Fig. 27
Fig. 27 Blackout measurements. N (abscissa) vs elevation (ordinates)
Typical received field levels are reported in next figure

Fig. 28 – Received field levels during blackout
2.2.11. Ducts
The meteorological conditions which lead to duct formation can be summarized as
Ö Evaporation over the sea
Shallow surface-based ducts are frequently observed over the oceans and results
from the humidity-temperature contrasts between air and sea in a shallow transi-
tion layer adjacent to the surface ( influenced by the wind ).They are characterized
by geographical , seasonal and diurnal variations
Ducts intensity is greater at lower latitudes ( typical duct depth is 5 m over the
north sea and 10-15 m in the Mediterranean ) , during summer and daytime and
they can also form ( also if for shorter periods ) over land areas due to evaporation
over wet surfaces after rain or over large bodies of water where steam fog is ob-
served.
Ö Anticyclonic subsidence
Elevated ducts ( their height is about 3 km above the surface ) rise in mid-latitudes
during subsidence associated with anticyclones where a temperature inversion is
formed by subsiding air. Stratocumulus sheets are a symptom of this phenomenon
: semi-permanent areas of stratocumulus occur around the British islands , in west-
ern Europe and in the sub-tropics where they could result in semi-permanent re-
gion of ducts.
Ö Nocturnal radiative cooling over land
Night-time radiative cooling of a land surface under clear skies leads to a formation
of a temperature inversion , whilst dew deposition leads to an increase of humidity

with height : the duct arises if the wind is light enough to reduce dew deposition .
typically also radiation fog is present and a duct may form during the early stages
of fog .
However , as the fog thickens , the temperature inversion migrates to the fog top ,
where a negative hydrolapse may develop , thus weakening the duct.
Ö Advection
Such ducts are significant for low-level coastal region and they may form when
warm , dry continental air passes over a cooler sea cooling and moistening the low-
est layers. This effect can reinforce a pre-existing evaporation duct increasing its
depth . this kind of duct are usually confined to within 100 km downwind of a lee
shore. Advection ducts should also be observed when warm moist air is advected
over a cooler sea , resulting in the formation of sea fog with a duct near the top of
the fog.
The following Fig. 29 a proposed output of this analysis is included
Fig. 29 – Duct measurements. N (abscissa) vs elevation (ordinates)
Atmospheric ducts may cause deep slow fading, strong signal enhancement, and
multipath fading on terrestrial line-of-sight links and may also be the cause of sig-
nificant interference on transhorizon paths. It is therefore of interest to describe the
occurrence of ducts and their structure.
Ducts are described in terms of modified refractivity defined as:
M(h)=N(h)+157*h ( M-units ) (19)
where h [km] is the height.

Fig. 30 illustrates the modified refractivity as a function of height above ground and
the definitions of duct types. Ducts can be of three types: surface based, elevated-
surface, and elevated ducts. Due to rather few cases of elevated-surface ducts in
comparison with surface ducts, the statistics have been derived by combining these
two types into one group called surface ducts. Surface ducts are characterized by (
referring to Fig. 30 )
Ö strength, Ss (M-units) or Es (M-units)
Ö thickness, St (m) or Et (m).
Two additional parameters are used to characterize elevated ducts:

Ö base height of the duct Eb (m), and Em (m)
Ö height within the duct of maximum M.
h a) b) c)
Et Em
St
Eb
St
M
Ss Ss Es
Fig. 30 – Definition of parameters describing a) Surface, b) Elevated Surface and c) Elevated

ducts7
7

Fig. 31 – Refractivity and ducts data report template

Network De
esign Funda
amentals 1.0
0
Fig. 32 – Typiical behavior of RX level depe

ending on prop
pagation enviro
onment.
dalmazio.ma
andich@alcate
r reserve
ed © 2008
2.3. Flat fading outage

Flat fading outage probability Of can be expressed as
−M
O f = P0 * 10 10
(20)
where
P0 is the fading occurrence factor ( 2.4 )
M is the fading margin ( 2.6 )
Of represent a probability and the corresponding number of monthly

seconds can be obtained multiplying this number to 2.592*106 ( = 3600
* 24 * 30 = number of seconds in one month )
2.3.1. Some consideration about P0 calculations

We’ll see that different methods are available to foresee the fading occurrence fac-
tor depending on link characteristics but the physical meaning ( i.e. method inde-
pendent ) of this parameter is reported in the next Fig. 33
Fading probability
1E+0
1E-1
P=P0*10^(-Fading/10)
1E-2
Probability
1E-3
1E-4
1E-5
-10 0 10 20 30 40 50
Fading [dB]
Gauss Cdf Rayleigh Cdf Total Cdf
Fig. 33 – P0
Deep fading distribution function for the unprotected channel is charac-

terized by an asymptotic Rayleigh slope of 10 dB / decade ( 10 dB of
fading vs 1 decade of probability ) : P0 value can be evaluated as the
crossing point between this slope and the level representing the nominal

received level power ( 0 dB of fading ).
The nominal level is the received power level when the link is not affected by fading
and differs from the most probable received power level depending on the median
depression.
Median
depression
Log-normal
Rayleigh
Fig. 34
Fading probability
1E+0
1E-1
Probability
1E-2
1E-3
1E-4
1E-5
-10 0 10 20 30 40 50
Fading [dB]
Gauss pdf Rayleigh pdf Total pdf

Fig. 35 – Fading statistics ( P0 =0.1)

Fading probability
1E+0
1E-1
Probability
1E-2
1E-3
1E-4
1E-5
-10 0 10 20 30 40 50
Fading [dB]
Gauss pdf Rayleigh pdf Total pdf

Fig. 36 - Fading statistics ( P0 =10)
The relationship between P0 and the Median Depression is given by
⎛ 2 ⎞
(1.8 * P00.6 ) ⎛ P ⎞
K = 10*Lgt ⎜ 1 + 10 +⎜ 0
⎟
⎟
⎜ ⎝ log(2) ⎠ ⎟
⎝ ⎠
2.4. The fading occurrence factor P0

The evaluation of flat fading Outage Time Of concerning an unprotected bearer is
based on the parameter P0 that it’s a fading indicator related to deep fading , and
indirectly , selective fading probabilities and can be evaluated according to the fol-
lowing methods
Ö CCIR Rep. 338-6.
Ö Barnett - Vigants
Ö ITU -R P 530-5/6
Ö ITU -R P 530-7/8
Ö ITU -R P 530-9/10
2.4.1. CCIR Rep. 338-6 Method

This approach is based on the relationship
3
f⎛D⎞
P0 = 0.2 * a * b * = 4 * 10−7 * a * b * f * D3
4 ⎜⎝ 50 ⎟⎠ (21)

Ö a climatic coefficient
2.4 for maritime hops
1 for flat hops
0.7 for hilly hops
0.3 for mountain
-1.3
Ö b roughness factor = (S /15) with 6 m < S < 42 m
b=0.26 (S=42) to b=3.3 (S=6) so the maximum 0.2*a*b (equivalent to

the Po of a 50km, 4GHz link) is approx 0.2*a*b=1.6 (maritime and flat
profile) and the minimum approx 0.2*a*b=0.016 (mountain with high
variation of the profile). With the same frequency and length the Po var-
iation is two decades from min to max.
Ö f frequency [GHz]
Ö d [km] is the length.
1
0.8
0.6
P0
0.4
0.2
0
10 30 50 70 90
Hop length [ Km ]
SEA FLAT HILL MOUNTAIN
Fig. 37 – P0 value (4 GHz , S = 42 m )
2.4.2. Roughness factor

This parameter takes into account the atmospheric layering ( more the atmosphere
is well mixed less is the extent of fading activity ) and its value , representing the
effects of the terrain , is evaluated according to the following procedure :
Ö From the profile of the terrain along the path, obtain the terrain heights h at
intervals of 1 km, beginning 1 km from one terminal and ending 1-2 km
from the other.
Ö Using these heights, carry out a linear regression with the “method of least
squares” to obtain the linear equation of the “average” profile
h(x)=a0 x+a1
(22)
where x is the distance along the path and

∑ x∑h
∑ xh- n
n
n
∑ h-a ∑ x
0
a0 = n
2
a1= n n
⎛ ⎞ n
⎜∑x⎟
∑n x - n ⎠
2 ⎝ n
(23)
with n equal to the number of profile height samples.
Ö From (23) calculate h(0) and h(d), the heights of the average profile at the
ends of the path, and the heights of the antennas above the average path
profile:
h1=he -h(0) h2=hr -h(d) (24)
compute the standard deviation of each height point (1 Km spaced and excluding
first and last km) respect the height of the linear regression straight line Y = a0+a1x
n n
n ∑ X *∑ Y
i i
∑XY − i i
i =1
n
i=1
1⎛ n n
⎞ 1 ⎛ n 2⎞
C2 = ; C1 = ⎜ ∑ Yi − C2 ∑ Xi ⎟ S= ⎜ ∑ ( Yi − Yi′) ⎟
i =1
; (25)
⎛ n ⎞
2
n ⎝ i=1 ⎠ n − 1 ⎝ i=1 ⎠
⎜ ∑ Xi ⎟
i =1
n
∑ Xi2 − ⎝ i=1 ⎠
i =1 n
Fig. 38 – Milano Pero – Caronno regression line
Fig. 39 – Abbate Argento – Bari regression line

2.4.3. ITU-R P.530 (5/6)

Method for detailed link design at small percentages of time
Ö Obtain the geo-climatic factor K as:
overland links for which the lower of the transmitting and
(5.4CLatCLon ) 1.5
K = 10 pL receiving antennas is less than 700 m above mean sea
level
overland links for which the lower of the transmitting and
(6.0CLat CLon ) 1.5
K = 10 pL receiving antennas is higher than 700 m above mean sea
level
links over medium-sized bodies of water , coastal areas
K = 10(4.8CLatCLon ) pL1.5
beside such bodies of water , or regions of many lakes
links over large bodies of water , or coastal areas beside
K = 10(4.4CLatCLon ) pL1.5
such bodies of water
where the coefficient CLat of latitude ξ is given by:

CLat Latitude ξ
0 ξ ≤ |53°|
–5.3 + ξ /10 |53°| <ξ < |60°|
0.7 ξ ≥ |60°|
Tab. 1 - Latitude coefficients
and the longitude coefficient CLon, by:

CLon
0.3 Europe and Africa
–0.3 North and South America
0 Asia, Australia
Tab. 2 – Longitude coefficients
If the antenna heights he and hr (m above sea level or some other reference height)
are known, calculate the magnitude of the path inclination |εp| in milliradians from:
hr – he
εp =
d
(26)
where the path length d is in km.
Ö From the profile of the terrain along the path, obtain the terrain heights h at
intervals of 1 km, beginning 1 km from one terminal and ending 1-2 km
from the other.
Ö Using these heights, carry out a linear regression with the “method of least
squares” to obtain the linear equation of the “average” profile according to
eq. (22)

Ö The regression coefficients can be calculated from the relations according to

(23) and calculate h(0) and h(d), the heights of the average profile at the
ends of the path, and the heights of the antennas above the average path
profile:
h1=he -h(0) h2=hr -h(d) (27)
Ö Calculate the “average” grazing angle ϕ corresponding to a 4/3 Earth radius
model (i.e. ae = 8500 km)
h1+h2 ⎡
1-m (1+b ) ⎤
2
ϕ= (28)
d ⎣ ⎦
where:
d2 h1 − h2 m+1 ⎡π 1 ⎛ 3c 3m ⎞⎤
m= ; c= ; b=2 cos ⎢ + arccos ⎜ ⎟⎟ ⎥ (29)
4ae (h1 + h2 ) (h1 + h2 ) 3m 3 3 ⎜ 2 (m + 1)3
⎣⎢ ⎝ ⎠ ⎦⎥
K d3.3 f 0.93
p0 = %
(1 + ε p )1.1ϕ 1.2
(30)
Ö Eq. (30) has been derived from
• fading data paths with lengths in the range 7-95 km
• frequencies in the range 2-37 GHz
• path inclinations for the range 0-24 mrad
• grazing angles in the range 1-12 mrad.
It is applicable only to narrow-band systems and for fade depths greater than about 15 dB
or the value exceeded for 0.1% of the worst month, whichever is greater.
2.4.4. ITU-R P.530 (7/8)

Obtain the geo-climatic factor K from the following empirical relation in the climatic
variable pL ( % of time that the refractivity gradient in the lowest 100 m of the at-
mosphere is more negative than –100 NU/km in the estimated average worst
month):
- ⎡7+0.1( C0 − CLat − CLon )⎤⎦ 1.5
K = 5*10 ⎣ pL (31)
where
Ö C0 coefficient is evaluated according to Tab. 3
Ö CLat coefficients according to Tab. 1
Ö CLon coefficients according to Tab. 2
Altitude of lower antenna and type of link terrain C0
Low altitude antenna (0-400 m) – Plains:

(Partially) overland links : lower-antenna altitude < 400 m a.s.l. 0

Low altitude antenna (0-400 m) – Hills:
(Partially) overland links : lower-antenna altitude < 400 m a.s.l. 3.5
Medium altitude antenna (400-700 m) – Plains:

(Partially) overland links : lower-antenna altitude in the range 2.5
400-700 m a.s.l.
Medium altitude antenna (400-700 m) – Hills:
(Partially) overland links : 400 m < lower-antenna altitude < 700 m 6
a.s.l.
High altitude antenna ( >700 m) – Plains:

(Partially) overland links : lower-antenna altitude > 700 m a.s.l. 5.5
High altitude antenna ( >700 m) – Hills:
(Partially) overland links : lower-antenna altitude > 700 m a.s.l. 8
High altitude antenna ( >700 m) – Mountains:
(Partially) overland links : lower-antenna altitude > 700 m a.s.l. 10.5
Tab. 3
The value of the climatic variable pL in (9) is estimated by taking the probability
(%) to exceed the highest value of the –100 NU/km gradient from the maps given
in Fig. 40 - Fig. 43.
Fig. 40 – Probability G < -100 NU (February)

Fig. 41 - Probability G < -100 NU (May)
Fig. 42 - Probability G < -100 NU (August)

Fig. 43 - Probability G < -100 NU (November)
Ö Coastal links over/near large bodies of water.

K is estimated as :
⎧⎪K (r ) = 10(1 – rc ) log Ki + rc log K cl

for K cl ≥ K i
K=⎨ l c (32)
⎪⎩Ki for K cl < K i
where rc is the fraction of the path profile below 100 m altitude above the mean
level of the body of water in question and within 50 km of the coastline and Ki is
given by (31) and
K cl =2.3*10-4 *10(
- 0.1C0 - 0.011|ξ |)
(33)
with C0 given in Tab. 3.

Ö Coastal links over/near medium-sized bodies of water
K can be estimated from:
⎧⎪K (r ) K = 10(1 – rc ) log Ki + rc log K cm

for K cm ≥ Ki
K=⎨ m c (34)
⎪⎩Ki for K cm < Ki
with
K cm =100.5 (log Ki + log K cl ) (35)
with Kcl given by equation (33)

From the antenna heights he and hr calculate the magnitude of the path inclination
| εp | (mrad) as

hr – he
εp = (36)
d
where d, is the path length (km).
P0 = Kd3.6 f 0.89 (1 + ε p )1.4 % (37)
This equation was derived from fading data on paths with

Ö lengths in the range 7-95 km
Ö frequencies in the range 2-37 GHz
Ö the maximum considered path inclination is 24 mrad.
2.4.5. ITU-R P.530 (9-10-11)

The geo-climatic factor K for the average worst month is evaluated as:
K = 10−3.9−0.003dN1 Sa −0.42 (38)
where dN1 is the point refractivity gradient in the lowest 65 m of the atmosphere
not exceeded for 1% of an average year (provided on a 1.5° grid in latitude and
longitude in Recommendation ITU-R P.4531 ) and Sa is the regional roughness de-
fined as the standard deviation of terrain heights (m) within a 110 km * 110 km
area based on GTOPO30 or SRTM30 data.
Kd 3.2 10 0.032f − 0.00085hL

P0 = % (39)
(1 + | εp | )0.97
where
Ö f is the frequency (GHz)
Ö hL is the altitude of the lower antenna (i.e. the smaller of he and hr)
Ö path inclination εp is evaluated according to (36)
Ö the geo-climatic factor K is obtained from equation (39)
Considering that factor K has been derived from multiple regressions on fading data
for 251 links in various geo-climatic regions of the world with path lengths d in the
range of 7.5 to 185 km, frequencies in the range of 450 MHz to 37 GHz , the follow-
ing limitations are applied :
• |εp| < 37 mrad
• lower antenna altitudes hL in the range of 17 to 2300 m
• refractivity gradients dN1 in the range of –860 to –150 NU/km
• Sa in the range of 6 to 850 m

2.4.6. Final remarks about P0 calculations

The following points must be underlined to avoid troubles considering that P0:
Ö it’s only a fading indicator
Ö it’s NOT a probability
Ö it’s not a dimensional quantity
Ö adopting ITU method it must be expressed in %
Ö it’s value should be > 1 ( > 100% )
Ö it must be used to evaluate flat and selective fading outage probabilities
Pay attention to properly use roughness and/or link slope values.

They’re similar and their physical meaning is , roughly speaking ,a
measure of atmospheric layering : if You look at Fig. 38 - Fig. 39 it’s
quite easy to verify that in both the cases we’re very near the terrain
and the link slope should be assumed to be totally negligible!
Sometimes the results obtained using different calculation methods are

significantly ( more than 1 decade ) different between them : in this
case You must verify if the problem is related to a single link or to the
whole region and to find the critical parameter in order to see if it’s real-
ly associated with a particular behavior
It’s strongly suggested to use the same approach for both P0 evaluation
and diversity improvement calculations : ALCATEL method is recom-
mended
ITUR-P530-10 (p.2.3.2) gives a different reference for the complete fad-

ing probability distribution, that is referred to the median fading. Giving
that there is no rules for estimate the median depression it is very diffi-
cult to estimate the probability for each fading level. Furthermore the
ITUR-P530-10 estimation for any level of fading is not in agreement
with the Po definition (the asymptotic distribution is to be referred to
the computed nominal level not to the median one otherwise it will be
useless to compute the outage probability).

Monthly fading occurence factor ( P0 )
8% 7.20%
7% 6.05%
5.3%
6%
5% 3.9%
4% 2.8%
3%
1.7% 1.4%
2%
1% 0.3%
0%
ITU
Jul 99
Jun 99
Sep 99
ALCATEL
May 99
Aug 99
Apr 99
Fig. 44 – Measured P0 : hilly terrain near the sea, 4 GHz 60 km
NAXXAR - M.T E LAURO

Fading oc c urrenc e fac t or ( P0 )
5
4.71
4
m easured P0
2067 L
1951 L
3 3.20
2.35 2067 H
2.37
1951 H
AVERAGE
2
1.62 1.30
1.39
0.91
1 0.66 0.62
0.44 0.41
0.32
0
m ar 96
m ay 96
ju n 9 6
ju l 96
nov 9 5
aug 96
jan 96
apr 96
sep 95
sep 96
feb 96
dec 95
oct 9 5
Fig. 45 – Measured P0 : Mediterranean Sea , 2 GHz 140 km
2.5. Fading margin

This quantity is the difference between the maximum signal level that can be re-
ceived and the equipment threshold level ( the received level corresponding to a
given quality in terms of errored blocks or bits ).
M = PRx - PThr (40)
Ö PThr : threshold value related to the specific radio equipment depending on

Capacity , modulation, configuration … = 10 * LGT ( KTBF )+ S/N
Ö PRX : the received power ( 2.5.1 ) corresponding to the maximum value of
Tx power
Ö M is the Fade Margin referred to BER relevant to the specific quality pa-

rameter
Typical threshold values relevant to 10-6 BER depending on modulation and capacity
are reported in the next table
Mod. Levels
Nyquist
System ( coherent detection S/N
bandwidth
)
2 10.5 B
4 13.5 B/2
PSK
8 18.8 B/3
16 24.4 B/4
16 20.5 B/4
32 23.5 B/5
QAM
64 26.5 B/6
128 29.5 B/7
-6
Tab. 4 – Typical uncoded 10 BER threshold levels
2.5.1. Received nominal level

The received nominal level is defined as
PRx = PTx - A fsl + AG + GTx + GRx - A f,Rx - A f,Tx - ABR - A - Tol-ATPC (41)
Ö PRx received nominal level

Ö PTx transmitted power
Ö Afsl propagation free-space loss ( 2.5.1.1 )
Ö AG gas attenuation ( 2.5.1.2 )
Ö GTx transmit antenna gain ( 2.5.1.3 )
Ö GRx receive antenna gain ( 2.5.1.3 )
Ö Af,Tx loss in the transmit feeder
Ö Af,Rx loss in the receive feeder
Ö ABR loss in the RF branching (filters) system
Ö A other attenuation
Ö Tol tolerance
Ö ATPC Automatic Transmit Power Control range ( 2.5.1.4 )
A worldwide suggested tolerance for all Alcatel computation should be

fixed. Giving that it should take into account error on pointing the an-
tenna and losses/gain tolerances (waveguide, antennas, branching, ag-
ing) the tolerance to be used should be 1dB per link + an additional fac-
tor for high gain antennas t.b.d.

2.5.1.1. Propagation Free-space loss8,9

Afsl is the propagation free-space loss and depends upon the operating frequency f
[GHz] and the hop length d [km] according to
A fsl = 92.44 + 20 * LGT ( d * f ) [dB] (42)
2.5.1.2. Gas attenuation10

The specific gas attenuation ( dB/km ) is due to dry air and water vapor compo-
nents and can be estimated using the following simplified algorithms considering
that absolute difference between the attainable results and the line-by-line calcula-
tion is generally less than 0.1 dB/km and reaches a maximum of 0.7 dB/km near
60 GHz.
For dry air, the attenuation γo (dB/km) for frequencies < 54 GHz is given by:
⎡ 7.34 rp2 rt3 0.7296912 b rp1.4954 rt–1.6032 exp [ –2.5280 (1– rt )] ⎤ 2

γo = ⎢ + ⎥ f × 10
–3
(43)
⎣⎢ f + 0.36 rp rt (54 – f)a + b
2 2 2
⎦⎥
with
η2 4a
a = 0.7982 * ln( ) b=
η1 η1
η1 = 6.7665 r –0.5050
p r
t
0.5106
exp [1.5663 (1– rt )] – 1 (44)
η2 = 27.8843 rp–0.4908 rt0.8491 exp [0.5496 (1– rt )] – 1
where
Ö f frequency [GHz]
Ö rp p / 1013
Ö rt 288/(273 + t)
Ö p pressure [hPa]
Ö t temperature [°C] (
8
Rec ITU-R P.341-5 The concept of transmission loss for radio links
9
Rec ITU-R PN.525-2 Calculation of free space attenuation
10
Rec ITU-R P.676-6 Attenuation by atmospheric gases

2
Ö Rec. ITU-R P.1510
Ö data )
For water vapor, the attenuation γw (dB/km) is given by:
⎧
⎪ ⎡ 3.84 ξ w1 g22 exp (2.23 (1–rt ))
γ w = ⎨0.0313 rp rt2 +0.00176 ρ rt8.5 +rt2.5 ⎢
⎪ ⎣ (f–22.235)2 +9.42 ξ 2w1
⎩
10.48 ξ w2 exp (0.7 (1–rt )) 0.078 ξ w3 exp (6.4385 (1–rt ))
+ +
(f–183.31)2 +9.48 ξ 2w2 (f–321.226)2 +6.29 ξ 2w3
3.76 ξ w4 exp (1.6 (1–rt )) 26.36 ξ w5 exp (1.09 (1–rt ))
+ 2 2
+ (45)
(f–325.153) +9.22 ξ w4 (f–380)2
17.87 ξ w5 exp (1.46 (1–rt )) 883.7 ξ w5 g557 exp (0.17 (1–rt ))
+ 2
+
(f–448) (f–557)2
302.6ξ w5g752 exp ( 0.41(1 − rt ) ) ⎤ ⎪⎫ 2
+ ⎥ ⎬ f ρ × 10 –4
( f-752 )
2
⎥⎦ ⎭⎪
with
ξ w1 = 0.9544 rp rt0.69 + 0.0061 ρ ξ w 2 = 0.95 rp rt0.64 + 0.0067 ρ

ξ w3 = 0.9561rp rt0.67 + 0.0059 ρ ξ w 4 = 0.9543 rp rt0.68 + 0.0061 ρ (46)
ξ w 4 = 0.9543 rp rt0.68 + 0.0061 ρ
2 2 2
⎛ f – 22.235 ⎞ ⎛ f – 557 ⎞ ⎛ f – 752 ⎞
g22 = 1 + ⎜ ⎟ g557 = 1+ ⎜ ⎟ g752 = 1+ ⎜ ⎟ (47)
⎝ f + 22.235 ⎠ ⎝ f + 557 ⎠ ⎝ f + 752 ⎠
where ρ is the water-vapor density [g m-3].
For a microwave path of L km the attenuation AG may be written as
AG = γ L = (γ o + γ w ) L (48)

Fig. 46 – Gas attenuation for different relative humidity values
2.5.1.3. Antenna system

Antenna gain depends on area A, efficiency η and wavelength λ as
4 πη A
G= (49)
λ2
Typical antenna gain (η = 55% ) should be evaluated as
GdB = 17.8 + 20 * LGT ( dm * fGHz ) (50)
If you double the antenna diameter or the frequency band its gain in-
creases of 6 dB
Theoretical antenna Radiation Pattern Envelope can be evaluated depending on the

value given by D/λ . Assuming
Ö G(ϕ) gain relative to an isotropic antenna
Ö ϕ off-axis angle (degrees)
Ö D antenna diameter
Ö λ wavelength
If D/λ > 100
2
⎛ D⎞
G(ϕ )=G max -0.0025 ⎜ ⎟ 0°<ϕ <ϕ m
⎝λ⎠
G(ϕ )=G1 ϕ m <ϕ < ϕ r (51)
G(ϕ )=32-log(ϕ ) ϕ r <ϕ < 48°
G(ϕ )= − 10 48°<ϕ < 180°
If D/λ < 100
2
⎛D⎞
G(ϕ )=G max -0.0025 ⎜ ⎟ 0°<ϕ <ϕ m
⎝λ⎠
λ
G(ϕ )=G1 ϕ m <ϕ < 100
D (52)
D λ
G(ϕ )=52-10*log( ) − 25* log(ϕ ) 100 <ϕ < 48°
λ D
D
G(ϕ )= − 10 -10*log( ) 48°<ϕ < 180°
λ
Ö G1 gain of the first side-lobe = 2 + 15 log (D/λ)
-0.6
20λ ⎛D⎞
ϕm = Gmax - G1 ϕr = 15.85 ⎜ ⎟ (53)
D ⎝λ⎠
Moreover D/λ (expressed in the same unit) may be estimated from the following

expression:
D 69.3
≈ (54)
λ θ
where θ is the beamwidth (–3 db) [degrees].
If θ is given, Gmax may be approximately estimated by:
D
Gmax (dBi ) ≈ 44.5-20*log(θ) 20 log ≈ Gmax − 7.7 (55)
λ
The previous formulae must be used by the tool to evaluate antenna RPE
if its model is not found in the antenna database.
All the previous relationships are applicable ONLY in far field ( Fraunhofer ) region :
it means that the distance ( D ) between the two antennas of diameter d must be
d2
D >> (56)
λ
In the following table the minimum distances [ m ] depending on working frequen-
cy and antenna diameter are reported
0.3 0.6 0.9 1.2 1.8 2.4 3 3.7 4.6 6
4 6 24 54 96 216 384 600 913 1411 2400
7 11 42 95 168 378 672 1050 1597 2469 4200
8 12 48 108 192 432 768 1200 1825 2821 4800
11 17 66 149 264 594 1056 1650 2510 3879 6600
13 20 78 176 312 702 1248 1950 2966 4585 7800
15 23 90 203 360 810 1440 2250 3423 5290 9000
18 27 108 243 432 972 1728 2700 4107 6348 10800
25 38 150 338 600 1350 2400 3750 5704 8817 15000
38 57 228 513 912 2052 3648 5700 8670 13401 22800
Tab. 5 – Far field distance [m] depending on frequency[GHz] and antenna diameter [m]
Check about far field condition must be performed
2.5.1.4. ATPC
In standard hops the maximum TX power (≈ +30 dBm) is really necessary only dur-
ing a very low fraction of time: less than 1‰ : Automatic Transmitted Power Con-
trol allows to reduce the unnecessary RF power by 10-20 dB considering that TX
power is increased (up to 100 dB/sec) to its maximum value only during deep fad-
ing periods.
ATPC doesn’t modify the effective margin vs. flat fading
If ATPC is included the received nominal value calculated by the tool

should be the value relevant to the minimum transmitted power.
The system margin can depend on ATPC considering that for some radio

family the maximum transmitted power ( with ATPC enabled ) could be

different by the transmitted power if ATPC is disabled.
Main ATPC advantages are significant reduction of :

Ö RF power device stress
Ö Interference towards other systems
Ö RF environmental pollution
Ö Power consumption
Fig. 47 - ATPC behavior versus fading
2.6. The selective fading outage

Two-ray model is adopted : it’s assumed that during multipath propagation ( that
happens with a probability η ) two different paths are available for transmitted sig-
nal to reach the receiver.
The second (reflected) ray is characterized by amplitude , delay ( τ ) and phase
shift (ϕ )
Basing on these statements the channel transfer function is
(
H(jϖ ) = a 1 − b e[
± (jϖ τ ±ϕ )]
) = a (1 − b e ± j(ϖ −ϖ 0 )τ
) (57)
where
Ö a is the amplitude of direct ray
Ö b is the normalized ( to the direct one ) amplitude of second ray
Therefore the transfer function can be written as
P(ϖ ) = a (1 + b2 ) − 2bcos ⎡⎣(ϖ − ϖ 0 )τ ⎤⎦ (58)

Three main methods are used to develop outage prediction methods for the selec-
tive effects of multipath propagation:
Ö Fade margin method where the dispersive fade margin DFM is determined
for a measured or predicted dispersion ratio DR.
Ö Signature curve methods based on measurements for a particular system
by means of a two-ray simulator.
Ö Linear amplitude dispersion method (LAD) where amplitude and delay dis-
tortion caused by the propagation effects are estimated in terms of probabil-
ity.
ITU-R P. 530 approach is based on the signature (according to ALCATEL

method)
2.6.1. Selective fading probability

Selective fading probability ( η ) is the percentage of time characterized by deep /
selective fading.
This parameter is used to evaluate selective fading outage and is related to the fad-
ing occurrence factor P0 as
η =1-exp(-0.2*p0.75
0 ) (59)
1
0.1
0.01
0.001
0.001 0.01 0.1 1 10 100
P0
η 1−η
Fig. 48 – Selective fading probability vs. P0
2.6.2. The signature

Signatures can be measured by generating signal distortions by means of a two-ray
simulator where a direct ray, and a secondary ray of relative amplitude b and de-

layed by t ( 6.3 ns. as reference ) is assumed.

Signature value Kn depends on
bit rate
modulation type ( levels and code )
equalization methods
BER.
Delay
-10
-15
Δf0=Ka/Ts
-20
dB
-25
-30
-35
B*=-20Lgt(KB*τ/Ts)
-40
-15 -10 -5 0 5 10 15
MHz
Fig. 49 – Measured signature with ALCATEL Ka and Kb parameters ( Ts=symbol time , τ = 6.3
ns)
The “notch” of this fade is f0 respect to the channel center frequency, and has a
depth
B = -20 logλ with λ = 1 - b (60)
The signature corresponds to the locus of critical couples of B* ( notch depth ) and
f0 ( notch position ) generating the reference BER.
Non-minimum phase fades can be taken into account by previous equation by
means of negative values of the delay τ or b values >1.
Signature must be used not only for selective fading evaluation but also
to verify the effects of reflected ray from ground or obstacles
When scaling the critical B values at high delays with signature formula
[-20log(Kb* t/Ts)], one should take into account that the demodulators
have normally a not linear scaling when delay approaches the symbol
time and equalizer length. So it is suggested not to use direct scaling
beyond these values. When required, a specific measurements of the
equipment sensitivity should be obtained.

2.6.3. Selective fading outage

According to ALCATEL method ( and ITU-R 530) the probability that the reference
BER , generated by intersymbol interference due to channel selectivity is higher
than the project objective ( considering minimum and non-minimum phase contri-
butions ) is
⎛ τ2 τ2 ⎞
Ps (BER > 10 −n ) = 2.15η ⎜ WM × 10 −BM/20 m + WNM × 10 −BNM/20 m ⎟ (61)
⎜ τr τ r ⎟⎠
⎝
Ö M = minimum phase, NM = not minimum phase
Ö Wx : signature width (GHz)
Ö Bx : signature depth (dB)
Ö τr : the reference delay (ns) used to obtain the signature ( 6.3 nsec )
Ö τm : echo delay mean value = 0.7 *(D/50)1.3 with D = hop length [km]
Ö η: Selective fading probability
ALCATEL signature area is normalized respect to the symbol time ac-

cording to latest ITU standard.
2
⎛τm ⎞
− Bc
Ts
Ps (BER > 10 ) = 4.3 ηKn ⎜ with Kn = Ts Δf *
-n
⎟ 10 20 (62)
⎝ Ts ⎠ τr
Ö Δf signature bandwidth [GHz]
Ö Bc notch producing a given BER [dB] assumed symmetric for min-
imum and non-minimum phase (otherwise Kn is the average of the two
computed values)
Ö Ts symbol time depending on capacity and modulation [ns]
Ö τm echo delay mean value [ns]
2.6.4. The dispersive fade margin approach

The first step is to determine the DFM of a radio system on a path with known dis-
persion ratio of DR0
This value, (dB), is taken as a reference dispersive fade margin (DFMR).
Then the DFM that would be measured or predicted on a path with a dispersion ra-
tio of DR is given by:
⎛ DR ⎞ TIBPD
DFM = DFMR - 10 log ⎜ ⎟ with DR = (63)
⎝ DR0 ⎠ TSFF * BF2
where

Ö TIBPD amount of time that a chosen In Band Power Dispersion value is ex-
ceeded
Ö TSFF amount of time that a chosen Single Frequency Fading value is ex-
ceeded
Ö BF bandwidth correction factor (the ratio of 22 MHz to the measurement
bandwidth).
The final value of outage due to selective fading , Os , is
-DFM
Os = P0 *10 10
(64)
More detailed model implementation is included in ANNEX 1
2.6.5. Signature, Bellcore Dispersive Fade Margin and

DFM
According to Bellcore TR-TSY-000752 , Issue 1 , October 1989
⎛ S ⎞
FdB = 17.6-10*Lgt ⎜ w ⎟ (65)
⎝ 158.4 ⎠
where Sw is the signature area , assuming the signature as a rectangle of height Bn
( Notch depth in dB ) and width w (MHz) leading to
Bn
-
S w = 2 * w * e 3.8 (66)
Stating
dBns = 20 * Lgt( τ m ) (67)
the final relationship is
DFM = FdB -7.64-0.643*Bn -dBns (68)
DFM figures must be available in the equipment database
Composite fade margin calculation must be performed ONLY to compare

the results obtained from different tools and/or customer requirements
Using composite fade margin method the amount of diversity improve-

ments , especially for frequency re-used configuration , MUST be agreed

2.7. Reflection
Two different reflection analysis must be implemented depending on the hop profile
- Sea reflection ( applicable if the link profile is flat )
- Ray tracing in all the other case considering that the existence and the location of
reflection point is strongly related to refractivity statistics and that more than one
reflected ray can exist.
Alternative analysis using Fresnel zone, may be very useful to fast graphical identi-
fication of potential reflection zones. The method require to draw all the Fresnel
zones from 1 to a maximum number of half wavelength, in order to verify if there is
any area of the hop profile that is stable inside two consecutive Fresnel ellipses
(that means to have less than half wavelength difference in path delay).
Reflections from ground out of the line-of-sight plane, may be also computed if the
elevation data are known on the adjacent area. These kind of computation are not
‘standard’ because are time consuming but modern computer may perform them
very fast and it may be very useful to have them into the tools.
2.7.1. Sea reflection calculation algorithm

Step 1: Calculate the complex permittivity of the Earth’s surface in the vicinity of
the surface reflection areas from:
σ
η=εr -j18 (69)
f
where εr is the relative permittivity and σ is the conductivity (S/m). Estimate εr and
σ from the information given in Recommendation ITU-R P.527 or according to theo-
retical interpolation reported in the next section.
Step 2: Calculate the grazing angle for the range of effective k values
h1 + h2 ⎡ d2
ϕ= 1 − m(1 + b2 )⎤ m=
d ⎣ ⎦ 4 k a (h1 +h2 )
(70)
⎡⎛ m+1 ⎞ ⎤ ⎧⎪ π 1 ⎡3 h − h ⎛ 3*|m| ⎞ ⎤ ⎫⎪
b=2* ⎢⎜ ⎟ ⎥ *cos ⎨ + *acos ⎢ *
1 2
* ⎜ ⎥
3 ⎟ ⎬
⎣⎝ 3|m| ⎠ ⎦ ⎩⎪
3 3 ⎢⎣ 2 h1 + h2 ⎝ (m+1) ⎠ ⎥⎦ ⎭⎪
Step 3: Calculate the reflection coefficient of the surface and the same range of k
values from:
sin ϕ − C
ρ= (71)
sin ϕ + C
where:

⎧C = η − cos2 ϕ H polarization
⎪
⎨ η − cos2 ϕ (72)
⎪C = V polarization
⎩ η2
The reflection coefficient must be modified depending of the type of terrain surface
according to the content of the following fig.
Finally a check about roughness value Tc, depending on the grazing angle ψ, must
be applied using the following relationship
λ
Tc = (73)
16 sin(ψ)

Network De
esign Funda
amentals 1.0
0
In case
c of mea
asured data the reflectio
on coefficien
nt can be ev
valuated as
10 ΔE/10 + 1 − 2 × 10 ΔE/ 20
ρeff
e = (74)
10 Δ E/10 − 1
ere ΔE (dB) is the difference betwe
whe een maximu
um and minimum RX le
evels.
ep 4: Calculate the dive
Ste ergence factor of the Earth’s surface from:
1 − m(1 + b2 )
D= (75)
1 + m(1 − 3b2 )
Ste
ep 5: Calculate the length, L1, of the 1st Fres
snel zone elllipse on the
e Earth’s su
ur-
face along the path from:
4f h1 h2
d 1+
L1[km] = 300d (76)
f(h1 + h2)2
1+
300d
d the width, W1, in the transverse direction fro
and om:
3 × 10−4 d
W1km = (77)
f
ere h1 and h2 are in m and d in k
whe km. Assume that the 1s
st Fresnel zone
z ellipse is
cen
ntered at the c point of reflection of an
e geometric a obvious s
surface refle
ection.
Ste
ep 6: If the
ere is clearly
y only a po
ortion(s) of the 1st Fre
esnel ellipse
e that will be
b
spe
ecularly refle
ecting, estim
mate the len
ngth Δx (km
m) of this po
ortion. Then estimate th
he
spe
ecular-reflec
ction factor from:
f(h1 + h2)4 (Δ x)
x2
Rs = (78)
300h1 h2 d3
ere again h1 and h2 arre in m and d in km. Ottherwise, assume that Rs = 1.

whe
Ste
ep 7: If the surface witthin the 1stt Fresnel elllipse is som
mewhat roug
gh, estimate
e the
surrface roughn
ness factor from:
f
g2
1+
2 40 π f σ h sinϕ
Rr = 2 g= (79)
g 3
1+
2
(2.35 + 2π )
with σh (m) th
he standard
d deviation of surface height abo
out the regrression curv
ve
thro
ough that portion
p of the path profiile within the 1st Fresnel ellipse. Otherwise,
O as-
sum
me that Rr = 1.
dalmazio.ma
andich@alcate
r reserve
ed © 2008
Step 8: Calculate the effective reflection coefficient for the relevant range of effective
k values from:
ρeff = ρDR sRr (80)
The level of the reflected wave(s) relative to the direct wave can now be estimated
Step 9: Calculate the angles [degree] between the direct and surface reflected
wave(s) at sites 1 and 2 for the relevant range of effective k values obtained in Step 3
from:
0.18 ⎡ h1 h1 − h2 d2 ⎤
α1 = ⎢ − − ⎥
π ⎣ d1 d 12.74k ⎦
(81)
0.18 ⎡ h2 h2 − h1 d1 ⎤
α2 = ⎢ − − ⎥
π ⎣ d2 d 12.74k ⎦
Step 10: Estimate the loss in level of the surface reflected signal(s) relative to the di-
rect signal introduced by antenna discrimination from :
The total loss due to antenna discrimination ( including ,if any, tilting effects) can
then be estimated from its RPE or from
⎡ ⎛ α + α ⎞2 ⎛ α + α ⎞ 2 ⎤
L a = 12 ⎢⎜ 1 t1
⎟ + ⎜
2 t2
⎟ ⎥ dB (82)
⎢⎣⎝ α a1 ⎠ ⎝ α a2 ⎠ ⎥⎦
where
αa1 and αa2 are the half-power beam widths of the antennas
αt1 and αt2 are the angles with which the antennas are tilted upwards.
Step 11 The overall loss in level of the surface reflected wave(s) is then given by:
L s = L a − 20log ρeff dB (83)
Step 12: If one or both antennas are tilted upwards, the corresponding loss in level of
the direct signal in normal propagation conditions (average k kav) is given by:
⎡ ⎛ α ⎞2 ⎛ α ⎞ 2 ⎤
L d (k av ) = 12 ⎢⎜ t1 ⎟ + ⎜ t2 ⎟ ⎥ dB (84)
⎢⎣⎝ α a1 ⎠ ⎝ α a2 ⎠ ⎥⎦
In super- or sub-refractive conditions, Ld (k) can be estimated from:
⎡⎛ α − α ⎞ 2
⎛ α t2 − α d ⎞
2
⎤
L d (k) = 12 ⎢⎜ t1 d
⎟ + ⎜ ⎟ ⎥ dB (85)
⎢⎣⎝ α a1 ⎠ ⎝ α a2 ⎠ ⎥⎦
where the angle-of-arrival of the direct signal is given approximately by :
⎛1 1 ⎞
α d = − 0.0045d ⎜ − ⎟ degrees (86)
⎝ k k av ⎠

Step 13: The maximum possible fade depth in normal propagation conditions (kav)
from destructive interference between the direct and surface-reflected signals can be
calculated from:
(
A max = − 20 log 10−L d / 20 − 10−L s / 20 ) dB (87)
In super-refractive or sub-refractive conditions in which the direct signal also un-

dergoes an additional loss 0.5Ladd (e.g. due to beam spreading in super-refractive
conditions) and the surface-reflected signal a gain −0.5Ladd, the maximum possible
fade depth is given by:
⎛ − L d + 0.5L add L − 0.5L add

− s ⎞
A max = − 20 log ⎜⎜ 10 20
− 10 20
⎟⎟ dB (88)
⎝ ⎠
The tools must give the opportunity to modify the tilt angles of the an-
tennas in order to evaluate how to minimize surface multipath fading or
surface multipath amplitude distortion, or a combination of the two.
The overall analysis must be performed following a statistical approach :

the calculations must be carried out for each k value considering the re-
levant occurrence probability
In case of space diversity configuration the behavior regarding each an-

tenna and the combined signal must be included in order to verify the
effectiveness of antenna spacing according to the template included in
the next figure

2.7.2. Theoretical formula for standard surface reflection

coeff.
2.7.2.1. Salt water

fMHz <2253.5895
εr =70
fMHz >2253.5895
−1
⎡0.014114535-5.2122497*10-8 *f+5.8547829*10-11 *f 2 ⎤
εr = ⎢ ⎥
⎢⎣-7.6717423*10-16 *f 3 +2.9856318*10-21 *f 4 ⎥⎦
(89)
fMHz <1106.207
σ =5
fMHz >1106.207
3.8586749+9.1253873*10-4 *f+1.530992*10-8 *f 2
σ=
1-2.1179295*10-5 *f+6.5727504*10-10 *f 2 -1.9647664*10-15 *f 3
2.7.2.2. Fresh water

fMHz <6165.776
εr =80
fMHz >6165.776
79.027635-3.5486605*10-4 *f+8.210184*10-9 *f 2
εr =
1-2.2083308*10-5 *f+2.7067836*10-9 *f 2 -1.0007669*10-14 *f 3
(90)
fMHz <5776.157
1-2.5539582*10-3 *f-3.7853169*10-5 *f 2
σ=
201.97103+1.2197967*10-2 *f-1.728776*10-6 *f 2
fMHz >5776.157
2
⎛ -0.65750351+6.6113198*10-4 *f+1.4876952*10-9 *f 2 ⎞
σ = ⎜⎜ ⎟⎟
⎝ 1+5.5620223*10-5 *f+3.0140816*10-10 *f 2 ⎠

2.7.2.3. Wet ground

fMHz <1312.054
εr =30
fMHz >1312.054
857.94335+5.5275278*10-2 *f
εr =
1-8.9983662*10-5 *f+8.8247139*10-8 *f 2
fMHz >4228.11
915.31026-4.0348211*10-3 *f+7.4342897*10-7 *f 2
εr =
1-9.4530022*10-6 *f+4.892281*10-8 *f 2
fMHz <15454.4 (91)

-3 -5 -8 2
σ=5.5990969*10 +8.7798277*10 *f+6.2451017*10 *f
-7.1317207*10-12 *f 3 +4.2515914*10-16 *f 4 -1.240806*10-20 *f5
+1.3854354*10-25 *f 6
fMHz >15454.4
σ =0.8756665+4.7236085*10-5 *f+2.6051966*10-8 *f 2
-9.235936*10-13 *f 3 +1.4560078*10-17 *f 4 -1.1129348*10-22 *f5
+3.3253339*10-28 *f 6
2.7.2.4. Medium dry ground

fMHz <4841.945
εr =15
fMHz >4841.945
215.87521+2.6151055*10-3 *f+1.9484482*10-7 *f 2
εr =
1-7.6649237*10-5 *f+1.2565999*10-8 *f 2
(92)
fMHz <4946.751
2
⎛ 2.4625032*10-2 +1.8254018*10-4 *f-2.664754*10-8 *f 2 ⎞
σ= ⎜ ⎟
⎜ +7.6508732*10-12 *f 3 -7.4193268*10-16 *f 4 ⎟
⎝ ⎠
fMHz >4946.751
2
⎛ 0.17381269+1.2655183*10-4 *f-1.6790756*10-9 *f 2 ⎞
σ= ⎜ ⎟
⎜ +1.1037608*10-14 *f 3 -2.9223433*10-20 *f 4 ⎟
⎝ ⎠

2.7.2.5. Very dry ground

εr =3
fMHz <590.8924
σ=0.0001
fMHz <7131.933
σ=2.2953743*10-4 -8.1212741*10-7 *f+1.8045461*10-9 *f 2 (93)
-12 3 -15 4 -19 5
-1.960677*10 *f +1.256959*10 *f -4.46811*10 *f
-23 6 -26
+9.4623158*10 *f -1.1787443*10 *f +7.9254217*10-31 *f 8
7
-2.2088286*10-35 *f 9
fMHz >7131.933
2
⎛ -0.049560275+2.9876572*10-5 *f-3.0561848*10-10 *f 2 ⎞
σ= ⎜ ⎟
⎜ +1.1131828*10-15 *f 3 ⎟
⎝ ⎠
2.7.2.6. Ice ( -1°C )

εr =3
fMHz <300
3.8814567*10-5 +9.878241*10-6 *f+7.9902484*10-8 *f 2
σ= (94)
1+8.467523*10-2 *f-9.736703*10-5 *f 2 +3.269059*10-7 *f 3
fMHz >300
1.24348*10-4 +8.6808*10-7 *f+7.27017*10-11 *f 2 -2.6417*10-14 *f 3 +1.375*10-18 *f 4
σ=
1+2.824598*10-4 *f-6.755389*10-8 *f 2 +2.8728975*10-12 *f 3 -1.87958*10-18 *f 4
2.7.2.7. Ice ( -10 °C )

εr =3
fMHz <8753.398
1 (95)
σ=
(
(51852.543+389.58894*f ) * 1-8.1212741*10-7 *f+6.832108*10-5 *f 2 )
fMHz >8753.398
σ=4.13105*10-5 +2.03589*10-7 *f-3.1739*10-12 *f 2 +4.52331*10-17 *f 3
2.7.3. Reflection evaluation based on ray tracing

A proprietary algorithm ,successfully developed in cooperation with Politenico di Mi-
lano, should be implemented by the tool in such a way that the final results can in-
dicate the value of RX level depending on

• Antenna position and tilt

• Real link profile
• Statistic of refractive index
• Terrain surface
Basing on our experience and on the available literature this is the most
effective method to be adopted.
Obviously other calculation approaches can be used considering that no

standard reference exits about this matter.
The connection between this calculation and the possibility to use

measured refractivity data decreases in a significant way the risk asso-
ciated with this issue.

2.8. Frequency re-use system outage
2.8.1. Introduction
Polarization is the characteristic of electromagnetic waves related to the orientation
and rotation of the electrical (E) or magnetic (H) vector.
Fig. 50 -
Ö Polarization is a very convenient and simple method to enlarge the isolation

between two signals increasing the spectrum usage.
Ö Isolation (XPI) of 30 – 40 dB can be obtained adopting available antennas.
Ö By using orthogonal polarization, two independents information channels us-
ing the same frequency can be transmitted over a single link.
Ö However, during fading periods, the antenna’s cross-polarization discrimina-
tion (XPD) is reduced and significant interference from adjacent or re-used
channel can be observed.
Ö Cross Polar Interference Cancellers (XPIC) are used to reduce the effects of
cross-polar interference.
2.8.2. Terminology
The following are the definitions of cross polarization terms according to Rec. ITU-R
P.-310
Ö CROSS POLARIZATION
The appearance , in the course of propagation, of a polarization component which is ortho-
gonal to the expected polarization
Ö DEPOLARIZATION
A phenomenon by virtue of which all or part of the power of a radio wave transmitted with a

given polarization may no longer have a defined polarization after propagation

Ö CROSS-POLARIZATION DISCRIMINATION - XPD
For a radio wave transmitted with a given polarization , the ratio at the reception of the
power received with the expected polarization to the power received with the orthogonal po-
larization
Ö CROSS-POLARIZATION ISOLATION - XPI
For two radio waves transmitted with the same frequency with the same power and ortho-
gonal polarization, the ratio of the co-polarized power in a given receiver to the cross-
polarized power in that receiver
XPI and XPD depend both on the antenna characteristics and on the
propagation medium.
2.8.3. Minimum mathematical background

The generic expression of the electric strength vector in TEM wave propagation
along z-axis is
E(t) = Re(Ee j(ωt-kz) ) = Re ⎡⎣ (uxEx + uyEy ) e j(ωt-kz) ⎤⎦ =

(96)
= ux Ex cos ( ωt − kz + ϕ x ) + uy Ey cos ( ωt − kz + ϕ y )
Ö E is a complex vector whose components Ex = |Ex|exp(jϕx) and Ey = |

Ey|exp(jϕy) are two phasors
Ö ux and uy are unit vectors in the x- and y-directions
Ö ω is the angular frequency
Ö t is the time
Ö k is the wave number.
The general way to describe a dual polarization channel is based on the relationship
existing between the transmitted ETX and received ERX vectors that can be defined
as
⎛E ⎞ ⎛E ⎞
ETX = ⎜ VTX ⎟ ERX = ⎜ VRX ⎟ (97)
⎝ EHTX ⎠ ⎝ EHRX ⎠
First we must consider that in this kind of application dual-polarized antennas are
adopted and we must characterized their RPE according to the following statements
⎛g x RV ⎞ ⎛g x TV ⎞
ARX = ⎜ RV ⎟ A TX = ⎜ TV ⎟ (98)
⎝ x RH gRH ⎠ ⎝ x TH gTH ⎠
where g is the gain in the used polarization and x is the relative gain relevant to the

other polarization
Also the transmission medium must be characterized considering that in the real
propagation condition a fraction of energy transmitted on one polarization could
reach the other one in RX side. We must define the propagation medium matrix as
⎛S SVH ⎞ − jϑij − j(ϖ τij +ϕij )

S = ⎜ VV ⎟ with Sij = lije = lije (99)
⎝ SHV SHH ⎠
The final formula becomes

ERX = A TX • S • ARX • ETX = M • ETX (100)
defining the equivalent global transfer matrix as :
⎛M MVH ⎞
M = A TX • S • ARX with M = ⎜ VV ⎟ (101)
⎝ MHV MHH ⎠
whose generic term ij gives the received strength component along the i axis when
a unit vector alone is transmitted along the j axis.
It means that all the following components must be taken in account to define the
parameters relevant to frequency re-used equipment.
TX Antenna RX Antenna
Transmis-
In- parame- parame-
Output sion Input Output
put ter ter
medium
V g V VV V g V a)
H x V VV V g V b)
H g H VH V g V c)
H g H HH H x V d)
TX Antenna RX Antenna
Transmis-
In- parame- parame-
Output sion Input Output
put ter ter
medium
H g H HH H g H e)
V x H HH H g H f)
V g V HV H g H g)
V g V VV V x H h)
The contents of the previous tables can be expressed as

Ö MVV = gTXV * SVV * gRXV a)
Ö MVH = xTXV * SVV * gRXV + gTXH * SVH * gRXV + gTXH * SHH * xRXV b)+c)+ d)
Ö MHH = gTXH * SHH * gRXH e)

Ö MHV = xTXV * SHH * gRXH + gTXV * SHV * gRXH + gTXV * SVV * xRXH f)+g)+h)
The resulting electromagnetic received field E is:
EVo = MVV * EVi + MVH * EHi
EHo = MHV * EVi + MHH * EHi
According to the previous statements we can define Cross Polar Discrimination
(XPD) and Cross Polar Isolation (XPI ) as:
⎡E ⎤ ⎛M ⎞
XPD V = −20*LGT ⎢ Ho ⎥ = −20*LGT ⎜ HV ⎟
⎣ EVo ⎦EHi=0 ⎝ MVV ⎠
⎡E ⎤ ⎛M ⎞
XPDH = −20*LGT ⎢ Vo ⎥ = −20*LGT ⎜ VH ⎟
⎣ EHo ⎦EVi=0 ⎝ MHH ⎠
(102)
⎛M ⎞
XPIV = −20*LGT ⎜ VH ⎟
⎝ MVV ⎠EHi =EVi
⎛M ⎞
XPIH = −20*LGT ⎜ HV ⎟
⎝ MHH ⎠EHi =EVi
According to experimental results and many other references we can assume that
XPIV ≅ XPDV and XPIH ≅ XPDH
This statement is very important because XPD arise naturally out of

propagation measurements whereas XPI is of interest to system engi-
neers since it defines the channel isolation available for frequency re-
use.
2.8.4. XP outage
The combined effect of multipath propagation and the cross-polarization patterns of
the antennas governs the reductions in XPD occurring for small percentages of
time.
To compute the effect of these reductions in link performance the following step-
by-step and procedures ( proposed by ALCATEL and actually adopted as ITU stan-
dard included in ITU-R P. 530 ) should be used:
Ö Compute
⎧⎪ XPDg + 5 for XPDg ≤ 35

XPD0 = ⎨ (103)
⎪⎩ 40 for XPDg > 35
where XPDg is the manufacturer’s guaranteed minimum XPD at boresight for both
the transmitting and receiving antennas, i.e., the minimum of the transmitting and

receiving antenna boresight XPDs.

Ö Evaluate the multipath activity parameter (η ) 2.6.1
Ö Determine
⎛k η⎞
Q = −10 ∗ LGT ⎜ xp ⎟
⎝ P0 ⎠
⎧0.7 one transmit antenna (104)
⎪
k xp = ⎨ ⎡ −6 ⎛ s t ⎞
2
⎤
⎪1 − 0.3 exp ⎢ − 4 × 10 ⎜λ⎟ ⎥ two transmit antennas
⎩ ⎣⎢ ⎝ ⎠ ⎦⎥
In the case where two orthogonal polarized transmissions are from different anten-
nas, the vertical separation is st (m) and the carrier wavelength is λ (m).
Ö Derive the parameter C as C = XPD0 + Q
Ö Calculate the probability of outage Pxp due to clear-air cross-polarization
from
MXPD
−
Pxp = P0 × 10 10 (105)
where MXPD is the equivalent XPD margin for a reference BER given by:
⎧ C0 C0
⎪⎪ XPD0 + Q - I
= C -
I
without XPIC
MXPD = ⎨ (106)
⎪ XPD + Q - C0 C0
+ XPIF = C - + XPIF with XPIC
⎪⎩ 0
I I
where XPIF ( or XPIRF ) represents the improvement introduced by cross-polar
canceller.
XPIRF figures to be considered MUST be clearly defined
Furthermore it can not be the maximum (best phase) one because the
phase my vary fading event by fading event. We should consider an ‘av-
erage’ value for the total outage computation, but at the same time we
should take care how to compare with competitors (so not declaring
poor XPIRF figures).

Fig. 51 – X Polar Interference Reduction Factor
2.8.5. Some consideration about antenna XPD

In the previous paragraph we’ve shown that final XPI value depend on both the
transmission medium and antenna system.
Moreover we’ve seen that in the algorithm adopted for outage calculation the an-
tenna component is limited to 40 dB ( i.e. a link implemented with 35 dB XPD an-
tennas or with antennas with XPD >40 dB is characterized , ceteris paribus , by the
same outage). This is because the very high 42/44 dB XPD antennas start degrad-
ing XPD in multipath environment before a standard 35dB XPD one do. So the ad-
vantage to have very high XPD is limited to a more stable environment in case of
XPIC failure.
Obviously depolarization effects arising from transmission medium can’t

be removed by the antennas
Anyway attention must be paid to the difference exiting between dual polarized an-
tenna and High XPD antenna that represent a “standard” characterized by the fol-
lowing points

Ö > 40 dB XPD at antenna bore-

sight
Ö 40 dB XPD inside rectangle
Ö 36 dB XPD inside 1 dB CP con-
tour
Ö 30 dB XPD elsewhere
Tab. 6 – High XPD antenna definition
Moreover this particular kind of antennas , that aren’t only antenna with high
XPD , are characterized , on the contrary of standard dual polarized antenna also if
with high values of XPD , by symmetrical XPD values in the azimuth / zenith plane
as can be noticed looking at the next figure.
Fig. 52 – High XPD antenna (right) and antenna with high XPD (left) comparison
The antenna type to be used must be carefully evaluated case by case

avoiding improper use High XPD antennas that are more expensive.
2.9. Unprotected channel global outage
2.9.1. ITU approach

According to ITU standard the overall outage of the unprotected ( single channel
without any kind of diversity ) channel is the sum of the previously evaluated com-

ponents relevant to flat, selective and frequency re-use outages

O = Of + Os + Ox (107)
This number is expressed as probability and to obtain the number of monthly

seconds outage it’s enough multiply this probability to the number of seconds per
month = 2592000
2.9.2. Composite fade margin approach

Please refer to ANNEX 1
2.10. The protected channel

In order to improve link performance diversity scheme is adopted : using more
than one receiver the outage probability of outage can be significantly reduced.
The diversity techniques can be implemented by means of
Ö hitless switch ( best channel selection )
Ö combiner : a device that adds the two signals , coming form main and di-
versity antenna, elaborated with a suitable algorithm.
Ö switch and combiner.
Up to now there is no standardized method to be adopted in order to fo-

resee different improvement if a combiner ( BB or IF ) or a switch is
adopted .
There is no ‘official and standardized” difference depending of the

switching activation criteria (EW activation characteristics), and switch-
ing properties (delay for protect and restore). There is also no ‘official
and standardized’ way to compute the hysteresis effects (for example
resync time of the demodulators after a deep fading) on the protected
outage. Despite this absence it is known that the equipment implemen-
tation strongly affect the final quality performance.
As a conclusion we’ve also to emphasize that the real attainable im-

provement is STRONGLY related to the device implementation characte-
ristics. These effects may in fact easily transform a theoretical ‘0 SES’
into several tens of SES.
In a diversity configuration the probability that BER exceeds quality performance

objective depends on

Ö Single ( unprotected ) channel performance

Ö correlation between the bearers
Ö multipath fading probability
Finally diversity can be implemented by means of two ( space, frequency, Hybrid )
or four receivers ( space and frequency )
The considered diversity configurations are:
Diversity configuration Receivers
Frequency diversity 2
Space diversity 2
Hybrid diversity 2
Space ( RX side ) and Frequency diversity 2
Space ( RX side ) and frequency diversity 4
Space (RX and TX sides ) and frequency diversity 4
Tab. 7 – Diversity configurations
f1
Hitless
switch
f2
Fig. 53 – Frequency diversity configuration
f1
Baseband / IF combiner
OR hitless switch
f1
Fig. 54 – Space diversity configuration
f1
Hitless switch
f2
Fig. 55 – Hybrid diversity configuration
f1
Hitless switch
f2
Fig. 56 – Space and frequency diversity with 2 receivers

f1
Combiner
Combiner
Hitless
switch
f2
Fig. 57 – Space and frequency diversity with 4 receivers
Combiner
Combiner
Hitless
switch
Fig. 58 – Space and frequency diversity with 4 receivers and 4 antennas
Space (RX and TX sides ) and frequency diversity will be indicated as

four antennas and four receivers diversity and is not considered by ITU
Diversity configuration are needed to improve quality performance but

they are totally not applicable , by a propagation point of view , for un-
availability
Angle diversity is not implemented and normally should not suggested

by Alcatel as a preferred diversity solution (if not join with freq/space
one) because of its unreliability and poor environment dependencies
knowledge.
Fig. 59 - 3+1 Space and frequency diversity with 4 receivers and 4 antennas
2.10.1. Two receivers diversity

The diversity parameter m relevant to order two diversity is defined as:
m = η * ( 1 - k2 ) (108)

The outage probability for a protected channel is:
-n Oi . O j
ODIV (BER > 10 ) = (109)
m
where "i" and "j" are the single bearers with performance Oi, Oj.
The corresponding improvement is:
m
I = Oi = (110)
ODIV P (BER > 10-n )
The improvement factor I must always be I > 1: the upper limit of the
diversity configuration performance is given by the unprotected single
channel.
-n Of,i . Of,j
O f,div (BER > 10 ) = min ( , Of,i, Ofj)
m (111)
O . Os,j
Os,div (BER > 10-n) = min ( s,i , Os,i, Osj)
m
where i and j are the single bearers with performance Of,i, Of,j and Os,i, Os,j.
2.10.2. What is the improvement ?
Fig. 60 – Diversity improvement
In Fig. 60 the unprotected channel asymptotic behavior ( 10 dB / decade ) indi-

cated by a black dotted line is reported together with the theoretical 2 receiver pro-
tected channel asymptotic behavior of 5 dB/decade ( red dotted line) .
For each level or RX level , the diversity improvement is the ratio between the
probability of the unprotected channel ( black line ) and the probability of the pro-
tected channel ( red line ).

The crossing point between the two asymptotic lines is the value of the
above defined m parameter relevant to flat fading. Some method, in-
cluding Alcatel, assume the selective fading improvement have the same
parameter m. Some other method use a different specific parameter
2.11. Two diversity receiver : ITU method

ITU approach is different from ALCATEL one and the correlation factor for selective
fading ( K2s ) is computed performing the evaluation of the following parameters
Ö Improvement ( depending on diversity configuration )
Ö Flat fading correlation factor Kf
2 If ⋅ Of
kf = 1− (112)
η
Ö Of flat fading probability
Ö η multipath activity factor
Ö If flat fading improvement
Warning : sometimes ITU method is not applicable !
The general expression of If is C*10M/10 where C depends on diversity configuration

and M is the flat fading margin.
It means that we can write
M −M
2 I ⋅O C*10 10 *P0 *10 10 C*P0
kf = 1− f f = 1− = 1− (113)
η η η
But K2f MUST BE > 0 and this is true ONLY IF C*P0/η <1
If this method is selected , the tool must check this condition and gene-
rates a warning message reporting about the inconsistency of the re-
sults
2.11.1. ITU frequency diversity improvement

If is the flat fading improvement factor corresponding to:
80 ⎛ Δf ⎞ F/10
If = 10 (114)
f ⋅ d ⎜⎝ f ⎟⎠
where
Ö Δ f : frequency separation [GHz] : maximum value is 500 HMz

Ö f: carrier frequency [GHz]

Ö F: flat fade margin [dB].
Ö This equation applies only for the following ranges of parameters:
2 ≤ f ≤ 11 GHz
30 ≤ d ≤ 70 km
Δf / f ≤ 5%
Check on Ins >1 MUST be performed !
If f = 7 GHz, Δf = 30 MHz , d = 60 km and F=30 dB then Ins =0.816 !
IT MEANS THAT 1+1 CONFIGURATION QUALITY PERFORMANCE IS

WORST THAN 1+0 !
2.11.2. ITU space diversity improvement

The flat fading improvement factor If is given by:
A-| G1 - G2 |
If = [1 - exp (-3.34 x 10-4 S 0.87
f -0.12
d 0.48
P0-1.04 )] 10 10
(115)
Ö A: fade depth(dB) for the unprotected path

Ö P0: fading occurrence factor ( absolute value )
Ö S : vertical separation (center-to-center) of receiving antennas (m)
Ö f : frequency [GHz]
Ö d: path length [Km]
Ö G1,G2 : gains of the two antennas [dB].
This equation should be applicable for 43 ≤ d ≤ 240 km, 2 ≤ f ≤ 11 GHz, and 3 ≤ S ≤
23 m.
WARNING
Also in this case the improvement should be < 1
Antenna spacing effect on improvement is almost is wavelength inde-

pendent !
Looking at the improvement formula it’s quite clear that there is no

theoretical reason suggesting the adoption of different diameter anten-
nas for diversity branch. In fact the difference in gain is compensated by
the no protected channel performance of the lower gain antenna chan-
nel.

Antenna spacing wavelength independent !
2.11.3. ITU step-by-step procedure

The following parameters must be evaluated :
Ö If according to the selected diversity configuration
Ö Kf according to Eq. (113) or as Kf,f* Kf,s in case of space and frequency di-
versity
Ö rw ( as a function of Kf )
⎧⎪ 1 − 0.9746 (1 − k 2f ) 2..170 for k 2f ≤ 0.26

rw = ⎨ (116)
⎪⎩ 1 − 0.6921 (1 − k 2f ) for k 2f > 0.26
1. 034
Ö K2s ( as a function of rw )
⎧ 0.8238 for rw ≤ 0.5

⎪⎪
k s2 = ⎨ 1 − 0.195 (1 − rw )
0.109 − 0.13 log (1− rw )
for 0.5 < rw ≤ 0.9628 (117)
⎪
⎪⎩ 1 − 0.3957 (1 − rw )
0.5136
for rw > 0.9628
Ö Flat fading outage probability, Odf as

Of
Odf = (118)
If
Ö Selective outage probability Ods as
2
Os
Ods = (119)
η (1 − k s2)
Ö the total diversity outage probability, Od, as
2
⎛ α2 α
⎞α
Od = ⎜ Odsel + Odflat
2
⎟ with α =1.5 (120)
⎝ ⎠
2.12. Four receivers diversity

As shown in the figure, there are two possibilities for this configuration including, or
not, a space diversity also in transmission side : space diversity correlation in
transmission is generally given by Ks1 and its value will be 1 in the case in which
there is only one antenna.

2
S1 S2
3
4
.
Fig. 61 – Four receiver diversity configuration
Only with ALCATEL method it is possible to estimate the performance at-

tainable with space diversity in TX side too
This kind of configuration , that reaches the maximum efficiency when

the adopted frequencies belong to different bands , can’t be considered
as standard and must be implemented only after PM acknowledgement .
Moreover can be used only in 1+1 configuration (with combiner+switch)

or up to 3+1 (3+1 switch with 4 frequencies or 2 frequencies dual pol)
and it’s not further expansible.
2.12.1. ITU method

Ö Calculate η
Ö Calculate the diversity parameter, mflat , as
(
mflat = η 3 1 − k 2flat,s ) ( 1− k ) 2
flat,f (121)
Ö Calculate the flat outage probability, Odflat, from:

4
Oflat
Odflat = (122)
mflat
Ö Calculate the square of the equivalent flat fading correlation coefficient, kflat,
from:
k 2flat = 1 − η ( 1− k ) ( 1− k )
2
flat,s
2
flat,f (123)
Ö Calculate the equivalent selective correlation coefficient, ksel, using the

same procedure previously exposed
Ö The selective outage probability, Odsel, is given by:
2
⎡ O2 ⎤
Odsel =⎢ sel
⎥ (124)
(
⎣⎢η 1 − k sel ⎥⎦
2
)

where Osel is single channel selective fading outage probability

Ö The total outage probability, Od, is then found as
2
⎛ α2 α
⎞α
Od = ⎜ Odsel + Odflat
2
⎟ with α =1.5 (125)
⎝ ⎠
2.13. Diversity techniques for frequency re-used systems

When considering the diversity improvement in case of frequency reuse it should be
take into account that there is an high probability of having similar fadings on both
polarization when using the same Tx and Rx antennae.
In fact the correlation between the two cross polar channels is normally very high,
so the diversity parameter m will be very low and the outage after protection not
significantly improved. So this protection may be just considered as a unavailability
protection against equipment failures.
If a space diversity is added to the polarization diversity the situation change signif-
icantly. In fact you have to consider a degradation of the XPIC behavior as de-
scribed by Alcatel method in ITU-R rec.530 but you can take the advantage of
space div in outage computation so the final result may be improved. In fact one
must assume that the space div dual pol have at least the uncorrelation of the
space div co-pol configuration. In case of ground reflection the improvement may
be even higher, but there are not enough experimental data to propose a closed
formula.
2.14. MULTICHANNEL SYSTEMS

The probability of outage of a N+1 system is greater than that of a 1+1 system be-
cause it may happen that more than one channel requires, at the same time, to be
switched over to the stand-by one.
After 1+1 system performance calculation, the worsening factor, with respect to the
1+1 operation, is easily obtained if no channel has a priority.
This factor is the ratio between N+1 system performance and 1+1 system perfor-
mance.
Under the simplifying assumption that no more than two channels are simulta-
neously faded, the worsening factor for the K channel, which is different from p
stand-by position is:

1 1 N+1 1
Wk = + ∑ (126)
| k - p | 2 i=1 | i - k |
where the index values i=k and i=p must be excluded.
This worsening factor is multiplied by 1+1 system performance to obtain the K
channel performance in the N+1 system.
Configura- Worsening
tion
2+1 1.5
3+1 1.75
4+1 1.965
5+1 2.153
6+1 2.317
7+1 2.459
8+1 2.548
9+1 2.695
10+1 2.793
11+1 2.881
Tab. 8 – Worsening factor for N+1 configurations
This is a proprietary solution taking into account that no standard has

been defined for this worsening factor
Tab. 8 reports worsening factors for uniform frequency spacing and

spare in first position. The worsening factors for spare position different
from first frequency or with irregular frequency separations need specif-
ic computation in agreement with (126).
2.15. General guidelines

• Validity range for ITU P0 evaluation methods
As for general experience and comparison with available data, ITU 530-10
is more adjusted to the experimental P0 only in case of ‘standard’ average
link (about 50km, low height terminals) and for some oversea links (in Ca-
narias area its estimations seems more reliable that 530-7). By the oppo-
site in inland links with high profile elevations ITU 530-7 seems more relia-
ble giving that 530-10 seems to much optimistic (this is the case of inland
Spain with average heights above 700m for the links profiles). The worst
value between the two could be a viable solution.

• Whatever the adopted ITU method the applicability range must be veri-
fied
• Space diversity configurations with different antenna diameters must be
avoided.
• N+1 worsening factor
The application of (126) seems ok in general, but we must to include in
some way the equipment implementation impairments. An historic way to
consider that is to estimate the ‘number of events’ the switch need to oper-
ate (per month) Nsw, applying a degradation component at the diversity
outage; this will be represented as a probability of having SES per switching
operation as : Osw=Nsw* P(SES|SW).
This last probability if switch unsuccessful will increase increasing (and ob-
viously decrease decreasing):
Ö delay to activate the alarms
Ö delay to complete the switch ( in N+1 the delay is >> than in 1+1 ! )
Ö probability of simultaneous fading on both directions of the links affect-
ing the two terminals service channels switch protocol communication
Ö time to restore the demodulators output after a sync loss
Ö probability to produce a misalignment between channels exceeding
the limits of the dynamic alignment recovery
Ö MTTR and FIT (giving that a unrecovered failed channel make it im-
possible to protect fading on the other ones)
All this factors must be take into account to avoid to declare very good di-
versity outages that can not be met, giving that there is no difference for a
customer between outage due to simultaneous fading and outage due to
equipment limitations in applying the diversity strategy.
• If diversity improvement evaluated according to ITU-R P 530–X method
is < 1
A check about this event is MANDATORY .
In this case the solution should be set the improvement to 1 but we’ve to
consider that , in this way, unpredictable results can be obtained.
This issue can be avoided increasing the fading margin and verifying that
the calculations are included in the applicability validity range.
• The correlation factor evaluated according to ITU-R P 530–X method is
<0
A check about this event is MANDATORY .

We need to avoid mathematical absurd (it should be logical to impose im-

provement>1 and correlation not negative) but this topic is to be put in
evidence on ITU study groups for to emend it.
• How to limit , just in case, the amount of the diversity improvement de-
pending on the adopted configuration
If a equipment impairment outage is added all the unrealistic improvements
will be automatically reduced avoiding the introduction of questionable rules
( NOT included in ITU-R P 530–X method )
• How to evaluate 1+1 in 4 receivers diversity configuration
Standard ITU-R P 530–X method MUST be used without any modification.
• How to evaluate 1+1 with space diversity in TX and RX side.
This configuration is not considered by ITU : ALCATEL model must be used.
• How to evaluate out-of-band frequency diversity improvement
This configuration is not considered by ITU : case by case analysis must be
performed
• Refractivity database
Actually available ITU tables are the only solution : development of a much
more complete and detailed set of useful data is in progress and it will be
used instead of the actual one.
• Terrain database management
GIS approach must be used and a unique DEM based on SRTM3 data MUST
be implemented for profile management. P0 ( whatever the adopted calcu-
lation method ) evaluation must be based on GTOPO30 ( or equivalent ) da-
ta.
• Definition of proper equipment parameters to be used in the calculations
and SES thresholds
Typical figures must be used for each radio parameter , antenna and aerial
systems.
The threshold MUST be relevant to 30% of errored blocks ( ≈ BER 10-4 in
case of USY or LSY )
Tolerance of 1 dB for flat fading and 1 dB for XPD ( in case of frequency re-
used systems ) must be introduced.
The protected channel quality MUST be better than the required ITU target.
• Diversity improvements in case frequency re-used systems especially if
they’re not implemented as a double N+1 system
The improvement relevant to frequency diversity MUST be neglected and

used as a safety margin.

• MTBF contribution MUST be integrated with rain unavailability
• Correct implementation of ANSI procedure ( including frequency re-used
systems )
According to Annex
• Packet data network dimensioning criteria
In progress
• Selective fading evaluation methods
Signature method is MANDATORY.

3. Rain attenuation
3.1. Introduction
11
states how to estimate , if more detailed data are not available , the statistics of
rain intensity ( expressed in mm/h and based on 1 minute integration time ) start-
ing from the contents on some database tables containing different updated values
of parameter related to this kind of data.
Also if this is the ONLY official ITU method in the following also the oldest approach
will be mentioned.
When rain data are available , whatever the source , the rain attenuation can be
determined basing on12 rules and outage periods generated by rain attenuation
and/or depolarization effects can be evaluated following the step-by-step procedure
included in Rec. ITU-R P. 530.
3.2. ITU-R 837-2

In this case the rain zones definition is based on tabulated climatic regions accord-
ing to the following table defining different rain intensity values relevant to fixed
probability thresholds
Percentage
A B C D E F G H J K L M N P Q
of time (%)
1 < 0.1 0.5 0.7 2.1 0.6 1.7 3 2 8 1.5 2 4 5 12 24
.3 <0.8 2 2.8 4.5 2.4 4.5 7 4 13 4.2 7 11 15 34 49
.1 <2 3 5 8 6 8 12 10 20 12 15 22 35 65 72
.03 <5 6 9 13 12 15 20 18 28 23 33 40 65 105 96
.01 <8 12 15 19 22 28 30 32 35 42 60 63 95 145 115
0.003 14 21 26 29 41 54 45 55 45 70 105 95 140 200 142
0.001 22 32 42 42 70 78 65 83 55 100 150 120 180 250 170
Tab. 9 – Tabulated rain regions according to ITU-R 837-2
11
Rec. ITU-R P. 837-3 Characteristics of precipitation for propagation modelling
12
Rec. ITU-R P. 838-2 Specific attenuation model for rain for use in prediction methods

Rain intensity (mm/min)
1000
100
Rain intensity
10
0.1
A B C D E F G H J K L M N P Q
Rain zone
1% 0.3 % 0.1 % 0.03 %

0.01 % 0.003 % 0.001 %
3.2.1. ITU-R 837-4 Rain intensity statistics

These method has been developed on data derived from 15 years of data of the Eu-
ropean Centre of Medium-range Weather Forecast (ECMWF) and is based on the
following files downloadable from ITU WEB site.
The data files are
ESARAINPR6.TXT
ESARAIN_MC.TXT
ESARAIN_MS.TXT
ESARAINLAT.TXT
ESARAINLON.TXT
The procedure performs the following step-by-step procedure
Ö Extract Pr6, Mc and Ms for the four points closest in latitude and longitude to
the desired location in order to obtain the values Pr6(Lat,Lon), Mc(Lat,Lon)
and Ms(Lat,Lon) at the desired location by performing a bi-linear interpola-
tion.
Ö Derive the probability of rain, P, from:
(
P(Lat,Lon) = Pr6 (Lat,Lon) 1 − e −0.0117(Ms (Lat,Lon)/Pr6 (Lat,Lon) ) (127)
Ö If the result of this operation is undetermined , the rain probability

P(Lat,Lon) and the rainfall intensity are set to zero and the procedure is
stopped.

Ö Derive the rainfall rate, Rp, exceeded for p % of the average year from:
− B + B2 − 4AC
Rp (Lat,Lon) = (128)
2A
where
⎛ p ⎞
A=ab B = a + c*ln(C ) C = ln ⎜ ⎟
⎝ P0 (Lat,Lon) ⎠ (129)
M (Lat,Lon) + Ms (Lat,Lon)
a = 1.11 b= c c = 31.5*b
22932P0
3.3. Rain attenuation characteristics

ITU-R P.838-2 states that specific attenuation γR (dB/km) is obtained from the rain
rate R (mm/h) using the power law relationship:
γ R = kRα (130)
where K and α coeff. are given by
3 ⎛ ⎡ ⎛ log ( f ) − b ⎞
2
⎤⎞
l og (k ) = ∑ ⎜ aj exp ⎢ − ⎜ j
⎟⎟ ⎥ ⎟ + mk log ( f ) + c k
j=1 ⎜ ⎢ ⎜⎝ cj ⎠ ⎥⎟
⎝ ⎣ ⎦⎠
(131)
4⎛ ⎡ ⎛ log ( f ) − b ⎞ 2 ⎤ ⎞
α = ∑ ai exp ⎢ − ⎜
⎜ i
⎟ ⎥ ⎟ + mα log ( f ) + cα
i=1 ⎜ ⎢⎣ ⎝ ci ⎠ ⎥⎦ ⎟⎠
⎝
and f = frequency [GHz], k : either kH or kV and α either αH or αV.
Frequency kH kV αH αV
1 0.0000387 0.0000352 0.9122 0.8801
1.5 0.0000868 0.0000784 0.9341 0.8905
2 0.0001543 0.0001388 0.9629 0.9230
2.5 0.0002416 0.0002169 0.9873 0.9594
3 0.0003504 0.0003145 1.0185 0.9927
4 0.0006479 0.0005807 1.1212 1.0749
5 0.001103 0.0009829 1.2338 1.1805
6 0.001813 0.001603 1.3068 1.2662
7 0.002915 0.002560 1.3334 1.3086
8 0.004567 0.003996 1.3275 1.3129
9 0.006916 0.006056 1.3044 1.2937
10 0.01006 0.008853 1.2747 1.2636
12 0.01882 0.01680 1.2168 1.1994
15 0.03689 0.03362 1.1549 1.1275

20 0.07504 0.06898 1.0995 1.0663

25 0.1237 0.1125 1.0604 1.0308
30 0.1864 0.1673 1.0202 0.9974
35 0.2632 0.2341 0.9789 0.9630
40 0.3504 0.3104 0.9394 0.9293
Tab. 10 -
The remaining coefficients are given in the following tables

a b c mk ck mα cα
j=1 0.3364 1.1274 0.2916
–
2 0.7520 1.6644 0.5175 1.9925 – –
4.4123
3 –0.9466 2.8496 0.4315
i=1 0.5564 0.7741 0.4011
2 0.2237 1.4023 0.3475 – 0.899
– –
3 –0.1961 0.5769 0.2372 0.08016 3
4 –0.02219 2.2959 0.2801
Tab. 11 – Coefficients in equations (2) and (3) for horizontal polarization
a b c mk ck mα cα
j=1 0.3023 1.1402 0.2826
–
2 0.7790 1.6723 0.5694 1.9710 – –
4.4535
3 –1.0022 2.9400 0.4823
i=1 0.5463 0.8017 0.3657
2 0.2158 1.4080 0.3636 0.875
– – –0.07059
3 –0.1693 0.6353 0.2155 6
4 –0.01895 2.3105 0.2938
Tab. 12 - Coefficients in equations (2) and (3) for vertical polarization
1E+1 1.40
1.30
1E+0
1.20
1E-1
α coefficients
1.10
k coefficients
1E-2 1.00
0.90
1E-3
0.80
1E-4
0.70
1E-5 0.60
1 GHz 10 GHz 100 GHz 1 GHz 10 GHz 100 GHz

3.4. Rain outage

The following technique is suggested by ITU-R P. 530-10 for estimating the long-
term statistics of rain attenuation:
Ö Obtain the rain rate R0.01 exceeded for 0.01% of the time ( with an integra-
tion time of 1 min ).
Ö Compute the specific attenuation, γR (dB/km) for the frequency, polarization
and rain rate according to the relationships defined in 3.3
Ö Compute the effective path length deff of the link by multiplying the actual
path length d by a distance factor r :
1 −0.015*min(R 0.01, ,100)
r= with d0 = 35e (132)
d
1+
d0
Ö The path attenuation exceeded for 0.01% of the time is given by:
A 0.01 = γR deff =γR dr (133)
Ö Attenuation exceeded for other percentages of time p in the range

0.001% to 1% is obtained as
Ap
AR = = 0.12p−(0.546 + 0.043Log10p) if absolute latitude > 30°
A 0.01
(134)
Ap −(0.855 + 0.139Log10p)
AR = = 0.07p elsewhere
A 0.01
It means that if AR is known the outage probability p is given by
p = exp ⎡ −14.62 + 0.02326 * 395145.16 − 99011.16 * ln ( 8.33 * AR ) ⎤

⎣ ⎦
⎡ ⎛ 0.12 * A 0.01 ⎞ ⎤ (135)
log10 (P ) = 11.628 * ⎢ −0.546 + 0.29812 + 0.172 * log10 ⎜ ⎟⎟ ⎥
⎢ ⎜
⎣ ⎝ A p ⎠ ⎥⎦
if abs(Lat) > 30 ° and
p=exp ⎡ -7.0816+0.01798* 155032.5-51209.49*ln (14.28571*AR ) ⎤

⎣ ⎦
⎡ ⎛ 0.07 * A 0.01 ⎞ ⎤ (136)
log10 (p) = 3.59712 * ⎢ −0.855 + 0.731025 + 0.556 * log10 ⎜ ⎟⎟ ⎥
⎢ ⎜
⎣ ⎝ A p ⎠ ⎥⎦
if abs(Lat) < 30 °. The results of this method are included in Fig. 62

3.4.1. Some consideration about the “new” ITU model ap-

plicability
According to ITU 530 - 7/8/9/10 “ ... The attenuation exceeded for other percen-
tages of time p in the range 0.001% to 1% may be deduced from the following
power law ….” . It’s clear that the formula applicability is strongly related to Ap/Ap001
that approximately corresponds to 1.44 for tropical areas and 2.2 for all the other
regions as shown in Fig. 62
Fig. 62 - Rain outage % vs. AR depending on latitude
It means that , in tropical regions , if this ratio is > 1.44 the rain unavailability is <
0.001% ( 315 seconds/year ) ! About the validity of this method it’s very important
to notice , looking at Fig. 63, that all the data are in line with the new ITU method
except for those relevant to Rio de Janeiro that are best fitted by the old approach :
the reason should be found , neglecting the effects related to hop length and fre-
quency, in the different measurement periods reported in the following table.

Fig. 63 – ITU reference for tropical rain attenuation
It’s very important to underline that also if the validity range of the for-
mula is clearly fixed between 1% and 0.001% sometimes should be use-
ful to “stress” the formula outside the lower limit if , for instance, relia-
bility data are required .
Unfortunately its quite evident that the red curve ( relevant to -30° < latitudes <
30° ) can introduce ambiguities considering that two different outage % can cor-
respond to the same AR value : this trouble has been overcame with a modifying
(136) according to :
⎧ ⎛ A + 0.2853 ⎞ ⎫
p=max ⎨exp ⎡-7.0816+0.01798* 155032.5-51209.49*ln (14.28571*AR ) ⎤ , 10 ^ ⎜ R ⎟⎬
(137)
⎩ ⎣ ⎦ ⎝ -0.633211 ⎠ ⎭
obtaining the result reported in the next figure

Ar vs % of time
10
1
A/A0.01
0.1
0.01
1E-05 1E-04 1E-03 1E-02 1E-01 1E+00
% of time
0.07*x^(-(0.855+0.139*LGT(x))) -0.633211*LGT(X)-0.2853 < 30° ( global )
Fig. 64 –

3.5. General guidelines

• Procedure described in ITU-R P 530-X MUST be used not only for rain outage
definition but also for outage related to depolarization in case of frequency r4e-
used systems.
• At least 2 dB of additional losses must be introduced in order to consider wet
radome effects.
• All the calculation MUST be performed assuming a relative humidity > 90%.

4. Quality
4.1. International standards: from ITU-T G.821 to

ITU-T G.828
Ö 1980 – G.821 defines error performance parameters and objectives (media
independent) for ISDN connections operating below (64 Kbit/s) the primary
rate of the digital hierarchy and it's based upon bit error and bit error ratio
measurements.
Ö 1993 – Rec. G.826 is developed and adopted. G826 defines error perfor-
mance parameters and objectives for constant bit rate digital paths operat-
ing at bit rates at and above the primary rate and "... is the only Recom-
mendation specifying error performance parameters and objectives at bit
rates at or above the primary rate."
Ö 1996 – G.821 is revised
It’s applicable only to bit rates below the primary rate of the digital hie-
rarchy
the Degraded Minute parameter is deleted;
Annex D/G.821 (1988) is deleted;
Performance objectives are stated for each direction of a N*64 kbit/s
circuit-switched connection (1 < N< 24 or < 31 respectively) used for
voice traffic or as a "Bearer Channel" for data-type services.
Ö 1998 - A new Recommendation G.826.1 has been added, with more stringent
objectives than stated by G.826, for what concern BBER and ESR.
Ö 03/2000 – G.828 Rec. is developed ONLY for SDH systems
(SDH systems using equipments designed after the adoption of G.828

(03/2000) should be governed by this Rec. - instead of Rec. G.826).
Ö 12/2002 – G.826 Rec. is revised. Now it defines end-to-end error perfor-

mance parameters and objectives for international, constant bit-rate digital

paths and connections. Therefore this is the reference Rec. for all bit rates:
connections operating below the primary rate, PDH systems and SDH sys-
tems using equipments designed prior the adoption of G.828 - 03/2000
(specific for SDH systems).
Ö 12/2002 – G.821 Rec. is revised. It must be noted that the application of this
Rec. is only about the connections operating below the primary rate that are
using equipments designed prior the adoption of G.826 - 12/2002. The per-
formance of N x 64 kbit/s international digital connections using equipment
designed after 14 December 2002 shall be governed by revised Rec. G.826
For equipments designed after 12/2002 Rec. G.821 is not applicable
Ö 12/2002 – G.829 Rec. is introduced, defining the error performance events

for SDH multiplex and regenerator sections (only definitions, not objec-
tives).
Looking at this very short ITU Recs. history it’s clear that nowadays several ITU
standards are devoted to define quality requirements and in this plethora of rules
the correct customer requirements ( if any ! ) satisfaction becomes a very hard
task.
Finally, just for the sake of clarity, we’ve to take in account that general purpose
rules (media independent) are fixed by ITU-T (transmission branch) and “adapted”
to radio equipment by ITU-R (radio) committee.

4.2. Medium independent and radio specific Recs.

Ö Media independent (ITU-T) Recs. are
G.821 : International connection below the primary rate
G.826 (1999) : International path at or above the primary rate
G.826 (12/2002) : International path for every bit rate
G.828 : International path at or above the primary rate
G.829 : Events definition for SDH-MS and SDH-RS
1. 2. 2000 2001 2002 2003 2004 2005
4. 5. G82 6. G82
3. 2MBit/s 1 6
DH 7.
8. G826
2MBit/s
9. 11.
10. 12. G828
DH 826
Tab. 13 – ITU-T quality Recommendations
Ö Radio specific (ITU-R) are:

F.1092 (1997) International path at or above the primary rate ( super-
seded )
F.1189 (1997) National path at or above the primary rate (superseded)
F.1397 (2002) International portion real links (withdrawn)
F.1491 (2002) National portion real links (withdrawn)
F.1668 (2003) Real digital fixed wireless links (International and Na-
tional portions) equivalent to F. 1397 – F. 1491 extended to every bit
rate
QUALITY 2000 2001 2002 2003 2004 2005

F.634-4
<2
F.696-2
Mbit/s
PDH
F.697-2 F.1668
>2 F.1397 F.1397-1 F.1397-2
Mbit/s F.1491 F.1491-1 F.1491-2
F.1397 F.1397-1 F.1397-2

SDH F.1668
F.1491 F.1491-1 F.1491-2
Tab. 14 – ITU-R quality Recommendations

In the following a very short explanation of all these rules is reported in order to
clarify their applicability and to avoid some dangerous pitfalls.

4.3. ITU-T G.821

Error performance, to be evaluated during connection's availability periods, is bit
based.
4.3.1. Error performance events and parameters
4.3.1.1. Events
Ö Errored second (ES): one-second period in which there is at least one er-
rored bit or a LOS (Loss of Signal) or AIS (Alarm Indication Signal) is de-
tected. (2002)
Ö Severely errored second (SES): one-second period in which there is a B.E.R.
>= 10-3 or a LOS (Loss of Signal) or AIS ( Alarm Indication Signal) is de-
tected. (G.821 Cor.1 07/2001 - G.821 12/2002)
4.3.1.2. Parameters
Ö Errored second ratio (ESR): The ratio of ES to total seconds in available time
during a fixed measurement interval.
Ö Severely errored second ratio (SESR): The ratio of SES to total seconds in
available time during a fixed measurement interval.
4.3.2. Performance objectives

The performance objectives for an international ISDN connection (using equipments
designed before the adoption of the G. 826 – 12/2002) are:
Performance classification Objective
Severely Errored Second Ratio < 0.002
Errored Second Ratio < 0.08
Tab. 15 – ITU-T G.821 performance objectives
4.3.3. Basic apportionment principles

Apportionment is based on three distinct quality classifications of the Hypothetical
Digital Reference Path (independent of the adopted transmission system) as shown
in Fig. 65

Fig. 65 – Hypothetical Reference Digital Path (HRDP) ITU-T G. 821 apportionment
The remaining 0.001 SESR is a block allowance to the medium and high grade clas-
sifications introduced in order to take in account adverse propagation conditions oc-
casionally experienced on transmission systems.
Because of the statistical nature of the occurrence of worst month ef-

fects in a worldwide connection, it is considered that the following al-
lowances are consistent with the total 0.1% SESR figure:
0.05% SESR to a 2500 km HRDP for radio relay systems which can be
used in one high grade portion and one medium grade portion of the
connection;
0.1% SESR to a satellite HRDP.
4.3.4. G. 821 related specs.

High Medium Local grade
grade grade
HDRP Rec. 556 Rep. 1052 Rep. 1053
Performance objectives HDRP Rec. 594
Rec. 696 Rec. 697
Performance objectives Real link Rec. 634
Tab. 16 ITU-T G.821 related specs

4.3.4.1. HRPD (Rec. 594) and High Grade (Rec. 634) performance
objectives
According to G.821 statements:

Ö performance objectives are valid for each direction of the 64 Kbit/s HRDP
(2500 Km )
Ö fading , interference and all other sources of performance degradation are
taken in account
Quality parameters should not exceed the following values
Ö SESR < 0.054% ( 0.004% + 0.05% ) = 1400 SES -> 0.56 SES/km/month
Ö ESR < 0.32% = 8295 ES -> 3.3 ES/km/month
4.3.4.2. Rec. 696 – Quality and availability objectives for Medium-

Grade
Medium grade systems are divided in 4 quality classes with different objectives
Percentage of any month
Class 1 Class 2 Class 3 Class 4
Parameters
280 Km 280 Km 50 Km 50 Km
Unavailability 0.033 0.05 0.05 0.1
SESR 0.006 0.0075 0.002 0.005
ESR 0.036 0.16 0.16 0.4
Tab. 17 – ITU-R Rec. 696

4.4. ITU-T G.826 and G.828
G.826 - End-to-end error performance parameters and objectives for in-

ternational, constant bit-rate digital paths and connections. (formerly :
Error performance parameters and objectives for international constant
bit rate digital paths at or above the primary rate )
G.828 - Error performance parameters and objectives for international,

constant bit rate synchronous digital paths.
Ö Both are media-independent

Ö They fix the quality but not the availability objectives
Ö They are based upon the error performance measurement of blocks in case
of bit rate at or above primary rate using error detection and allowing ISM
(In Service Monitoring).
Ö In case of PDH bit rates below the primary rate G826 is bit based.
Ö In the previous version G.826 was applicable to PDH or SDH paths over a
27500 km HRDP: actually it is applicable to every PDH systems and to SDH
equipments ante G.828.
Ö G.828 is applicable just to SDH paths.
Ö Are the only recommendations required to design the error performance of
transport networks
Ö The performance objectives are applicable to each direction of the path
(multiplex and cross-connect are not included).
Ö Compliance with the performance specification of this Rec. will, in most cas-
es, meet the G.821 requirements.
4.4.1. The Definition and Measurement of the Block

Ö A block is a set of consecutive bits associated with the path
Ö Each bit belongs to one and only one block.
Ö Consecutive bits may not be contiguous in time.
The blocks are defined for
Ö Path by G.826 and G.828 for path based on SDH
Ö Multiplex Section (MS) and Regenerator Section (RS) by G.829

4.4.2. In Service Monitoring (ISM)

ISM quality evaluation (block based) is performed by means of EDC (Error Detec-
tion Code) (or BIP - Bit Interleaved Parity check) inserted in the SDH frame (frame
cells):
Ö B3 path
Ö B2 multiplex section
Ö B1 regenerator section
Ö The EDC bits are physically separated from the block to which they apply.
Ö OSM (Out of Service Monitoring) shall also be block-based.
Ö Obviously OSM error detection capability will be more reliable than ISM ca-
pability.
Ö Since ITU-T G. 826/828 define a block as consecutive bits associated with a
path, each BIP-n in the SDH path overhead pertains to a single defined
block.
Ö The BIP-n is NOT interpreted as checking "n" separate interleaved parity
check blocks.
Ö If any of the "n" separate parity checks fails, the block is assumed to be in
error.
Fig. 66 – SDH frame structure

4.4.3. Path block definition

Capacity Path type Block size EDC
2240 VC - 12 1120 BIP - 2
48960 VC - 3 6120 BIP – 8
150336 VC - 4 18792 BIP - 8
Tab. 18 – Path block definition (ITU-T Rec. G.707/Y.1322)
4.4.4. MS block definition

Capacity Path type Blocks/frame Block size
STM-0 VC - 3 8 801
STM-1 VC - 4 24 801
Tab. 19 – Multiplex section block definition (ITU-T G.829)
4.4.4.1. STM-1 definition
The 24 BIP-1s contained in the B2 bytes of an SDH-MS in a STM-1 pertain to 24 dif-

ferent blocks per frame corresponding to the i-th bit of the B2 bytes and the rele-
vant monitored bits in the same frame (Rec. G.707) : a SES is defined as one-
second period that contains more than 15% of EB or at least one defect.
The 8 BIP-1s contained in the B2 byte of an SDH-MS in a STM-0 pertain to 8 dif-

ferent blocks per frame corresponding to the i-th bit of the B2 byte and the relevant
monitored bits in the same frame (Rec. G.707) : a SES is defined as one-second
period that contains more than 15% of EB or at least one defect.
4.4.5. RS block definition

Capacity Path type Blocks/frame Block size
STM-0 VC - 3 1 6480
STM-1 VC - 4 1 19440
Tab. 20 – Regenerator section block parameters (ITU-T G.829)
The BIP-8 contained in the B1 byte of an STM1-RS pertain to 1 block : a SES is de-
fined as one-second period that contains more than 30% of EB or at least one de-

fect .
The BIP-8 contained in the B1 byte of an STM0-RS pertain to 1 block : a SES is de-
fined as one-second period that contains more than 10% of EB or at least one de-
fect.
4.4.6. ITU-T G.826/8 Error Performance Events and Para-

meters
4.4.6.1. Events
Ö Errored Block (EB) : 1 block with at least 1 errored bit

Ö Errored Second (ES) : 1 second period with at least one errored block or at
least one defect (as reported in Annex 1 )
Ö SES : 1 second containing more than 30% errored blocks or at least one de-
fect.
Ö Background Block Error (BBE) : 1 errored block not belonging to a SES
G.828 introduces an additional error performance event that is SEP (se-

verely errored period: a sequence of 3 to 9 (t.b.d.) consecutive SES ter-
minating by a second which is not a SES.
The parameter related to SEP it’s defined as SEPI (SEP intensity)
SEP and SEPI values are to be defined
4.4.6.2. Errored block – B.E.R. relationship
Hop calculations are performed assuming, for ALCATEL equipments, that

30% of errored blocks correspond to a 10-4 B.E.R.

Fig. 67 – BER vs. errored blocks relationship
4.4.6.3. Parameters
Ö Errored Second Ratio (ESR) The ratio of ES in available time to total seconds
in available time during a fixed measurement interval.
Ö Severely Errored Second Ratio (SESR) The ratio of SES in available time to
total seconds in available time during a fixed measurement interval.
Ö Background Block Error Ratio (BBER) The ratio of BBE in available time to
total blocks in available time during a fixed measurement interval excluding
all blocks affected by SES
The count of total blocks excludes all blocks during SES.
BBER is applicable ONLY for bit rate at or above the primary rate.
Ö SEP intensity (SEPI) the number of SEP events during available state, di-
vided the total available time
Error performance must be evaluated whilst the path is in the available

state.

Fig. 68 – Quality parameters flow chart
4.5. G826 – G 828 Error Performance Objectives

Mbit/s N * 0.064 1.5 - 5 5 - 15 15 - 55 55 - 160
ESR 0.04 0.05 0.075 0.16
G 826 SESR 0.002
BBER N. A. 2*10-4
Mbit/s N * 0.064 1.664-2.240 6.848 48.960 150.336
ESR 0.01 0.01 0.02 0.04
SESR 0.002
G 828 N. A. -5
BBER 5*10 1*10-4
SEP t.b.d.
Tab. 21 – Error Performance Objective (EPO) - absolute probability
Ö Compliance with the path performance specification of Rec. G.826 will, in

most cases, also ensure that a client 64 kbit/s connection will meet its re-
quirements.
Ö Compliance with the performance specification of Rec. G.828 will, in most
cases, meet the G.821 requirements.
Ö It must be noted that the G.821 and G.826 have the same SESR-objective
(0.002), but different ESR-objective
Ö Rec. G.828 based paths can carry ATM traffic. Synchronous digital paths
meeting the objectives of G.828 will enable the ATM traffic to meet I.356.
Ö The path fails to meet the error performance requirement if any of these ob-

jectives is not met (each direction of the path shall satisfy the allocated ob-
jectives).
The actually suggested evaluation period is 1 month: in cases where 1

month evaluation may not permit accurate statistical estimation, a long-
er period (up to 1 year) may be used.
Ö Systems submitted to this Recs. must meet their allocations of the end-to-
end objectives for the highest foreseen bit rate in such a way that meeting
the objectives for this highest bit rate path should be sufficient to ensure
that all paths through the system are achieving their objective.
Ö The adopted apportionment methodology specifies the levels of performance
expected from the national and international portions of an HRDP.
4.6. G.826/8 - National/International Portion of the

End-to-End Path
Ö The following allocation applies to each performance parameter previously
defined (ESR, SESR, BBER).
Ö For each national portion is allocated a fixed block allowance of 17.5 % of
the end-to-end objective.
Ö For the international portion is allocated a block allowance of 2 % per inter-
mediate country plus 1 % for each terminating country.
In both cases (national and international portion of path) a distance-

based allocation is added to the block allowance in terms of 0.2 % per
100 km.
For the G.826 (12/2002) the added distance-based allocation is rounded

up to the nearest 500 Km (i.e., the two national portions comprise at
least 500 km each).
For the G.828 (03/2000) the added distance-based allocation is rounded

up to the nearest 100 Km. Also in this case the two national portions are
allocated a minimum of 500 km (i.e. 1%) each.

Network De
esign Funda
amentals 1.0
0
Fig. 69 - ITU-T G. 826

6/828 apportion
nment criteria
Fig. 70 - ITU-T G. 826

6/828 apportion
nment criteria
4.7. ITU-T
T G. 821 vs. ITU
U-T G. 826/8
Ö G.821/8 6 kBit/s in terms of a BER > 10-3.
826 define a SES condition at N x 64
Ö G.826/8
8 define a SES
S conditio
on, at or above the prim i terms of >
mary rate, in
30% Errrored Block
ks (end-to-e e detection of certain defects.
end) and the
Ö It should be noted that the tw
wo criteria do ctly equate and the rela
d not direc a-
p between them
tionship t depends on block size, bit errror structure
e, etc.
Ö All Reco ons define a SES objective for a 27500 km HRDP of 0.00
ommendatio 02
dalmazio.ma
andich@alcate
r reserve
ed © 2008
but the apportionment strategy is different.

Ö G.821 assigns the first-half 0.001 and "the remaining 0.001 is allocated on a
block basis to medium and high grade classifications to accommodate the
occurrence of adverse network conditions occasionally experienced ".
Ö G.826-8 allocates the total 0.002 SESR allowance to the 27500 Km without
any special allowance for radio
SESR objective G. 821 G. 826 G.828
High Grade Internatio- Internatio-
nal nal
25000 km 40%*0.1%=0.04% 60% * 0.2%
(34 SES/day) (103 SES/day)
2500 km 10%*40%*0.1%=0.004
%
( 3.4 SES/day)
2500 km + 0.004% + 0.05% =
0.05% radio allo- 0.054%
wance ( 46 SES/day)
Tab. 22 – G.821 vs. G. 826/8 objectives comparison
4.8. ITU-T G. 826-8 Related Recommendations

International National
Rec. F.1092 ( superseded ) Rec. F.1189 (superseded)
Rec. F.1397 ( superseded ) Rec. F.1491 ( superseded )
Rec. F.1668
Tab. 23 – ITU-T G.826/8 radio related recommendations
4.8.1. ITU-R P.1092 (International portion) – Superseded!

Defines a rule in order to indicate the objectives based on real link length (with a
minimum length of 50 Km) and it should be used for path performances.
Assuming
Ö Lref reference length (1000 Km )
4.8.2. Lmin minimum radio link length (50 Km)

Ö L real radio link length ( Km )
Ö BR block allowance ratio (0 < BR <1 )

The Error Performance Objective (E.P.O.) is subdivided into:

The distance allocation factor: FL = 0.01 x L/500 L (Km)
The block allowance factor BL defined depending on the value of BR
Transit country Terminating country
L L L REF (138)
BL =BR *0.02* L min < L < L REF BL =BR *0.02* L min < L <
L REF L REF 2
2
L REF
BL =BR *0.02 L > L REF BL =BR *0.01 L>
2
A = F L + BL
Mbit/s 1.5-5 5-15 15-55 55-160

ESR .04*A .05*A .075*A .16*A
SESR .002*A
BBER .0002*A
Tab. 24 - ITU-R P.1092 objectives
4.8.3. ITU-R F. 1397 ( International portion )

ITU-R F. 1397 Recommendation identifies a rule to indicate the objectives based on
real link length and it should be used for path, multiplex and regenerator sections
performances according to the parameters defined in G. 826-828 for path and G.
829 for multiplex and regenerator sections.
4.8.3.1. EPO parameters for Intermediate countries (ITU-T G. 828)
⎧ L
⎪1.664/6.848 MBit/s 5*10-4 (1+BR )*
2500
⎪
⎪ L
ESR ⎨48.960 MBit/s 10-3(1+BR )*
⎪ 2500
⎪ L
⎪150.336 MBit/s 2*10-3 (1+BR )*
⎩ 2500
L
SESR 1.664/150.336 MBit/s 10-4 (1+BR )* (139)
2500
⎧ L
⎪⎪1.664/48.960 MBit/s 2.5*10-6 (1+BR )*
2500
BBER ⎨
⎪150.336 MBit/s L
5*10-6 (1+BR )*
⎪⎩ 2500
Tab. 25 – EPO parameters for Intermediate countries (ITU-T G. 828) 50 < L < 1000 km

⎧ L
⎪1.664/6.848 MBit/s 5*10-4 * + 2 * 10−4 * BR
2500
⎪
⎪ L
ESR ⎨48.960 MBit/s 10-3 * + 4 * 10−4 * BR
⎪ 2500
⎪ L
⎪150.336 MBit/s 2*10-3 * + 8 * 10−4 * BR
⎩ 2500
L
SESR 1.664/150.336 MBit/s 10-4 * + 4 * 10−5 * BR (140)
2500
⎧ L
⎪⎪1.664/48.960 MBit/s 2.5*10-6 * + 1 * 10−6 * BR
2500
BBER ⎨
⎪150.336 MBit/s L
5*10-6 * + 2 * 10−6 * BR
⎪⎩ 2500
Tab. 26 - EPO parameters for Intermediate countries (ITU-T G. 828) L > 1000 km
4.8.3.2. Example 1
ESR, SESR and BBER evaluation for a link length of 105 km.
STM-1 capacity ( G. 828 ) intermediate country is assumed.
BR=1
–3
ESR 2 *10 (1+1)*105/2500 1.68*10–4 ES/month = 435
–4 –6
SESR 10 (1+1)*105/2500 8.4*10 SES/month = 22
–6 –7
BBER 5 *10 (1+1)*105/2500 4.2 *10 BBE/month = 8709
BR =0
–3
ESR 2 *10 (1+0)*105/2500 8.4*10–5 ES/month = 218
–4 –6
SESR 10 (1+0)*105/2500 4.2*10 SES/month = 11
–6 –7
BBER 5 *10 (1+0)*105/2500 2.1 *10 BBE/month = 4355
Tab. 27 – ITU-R F. 1397 example (1)
4.8.3.3. EPO parameters for Terminating countries ( ITU-T G.828)
⎧ L
⎪1.664/6.848 MBit/s 5*10-4 (1+BR )*
2500
⎪
⎪ L
ESR ⎨48.960 MBit/s 10-3(1+BR )*
⎪ 2500
⎪ L
⎪150.336 MBit/s 2*10-3 (1+BR )*
⎩ 2500
L
SESR 1.664/150.336 MBit/s 10-4 (1+BR )* (141)
2500
⎧ L
⎪⎪1.664/48.960 MBit/s 2.5*10-6 (1+BR )*
2500
BBER ⎨
⎪150.336 MBit/s L
5*10-6 (1+BR )*
⎪⎩ 2500
Tab. 28 - EPO parameters for Terminating countries ( ITU-T G.828) 50 < L < 1000 km

⎧ L
⎪1.664/6.848 MBit/s 5*10-4 * + 10−4 * BR
2500
⎪
⎪ L
ESR ⎨48.960 MBit/s 10-3 * + 2 * 10−4 * BR
⎪ 2500
⎪ L
⎪150.336 MBit/s 2*10-3 * + 4 * 10−4 * BR
⎩ 2500
L
SESR 1.664/150.336 MBit/s 10-4 * + 2 * 10−5 * BR (142)
2500
⎧ L
⎪⎪1.664/48.960 MBit/s 2.5*10-6 * + 5 * 10−7 * BR
2500
BBER ⎨
⎪150.336 MBit/s L
5*10-6 * + 10−6 * BR
⎪⎩ 2500
Tab. 29 - EPO parameters for Terminating countries ( ITU-T G.828) L > 1000 km
4.8.3.4. EPO parameters for Intermediate countries (ITU-T G.826)
⎧ L
⎪1.5/5 MBit/s 2*10-3 (1+BR )*
2500
⎪
⎪5/15 MBit/s L
⎪⎪ 2.5*10-3 (1+BR )*
2500
ESR ⎨
⎪15/55 MBit/s L
3.75*10-3 (1+BR )*
⎪ 2500
⎪
⎪55 /160 MBit/s L
8*10-3(1+BR )*
⎪⎩ 2500
L
SESR 1.5/160 MBit/s 10-4 (1+BR )* (143)
2500
L
BBER 1.5/160 MBit/s 10-5 (1+BR )*
2500
Tab. 30 - EPO parameters for Intermediate countries (ITU-T G.826) 50 < L < 1000 km
⎧ L
⎪1.5/5 MBit/s 2*10-3 * + 8 * 10-4 *BR
2500
⎪
⎪5/15 MBit/s L
⎪⎪ 2.5*10-3 * + 10-3 *BR
2500
ESR ⎨
⎪15/55 MBit/s L
3.75*10-3 * + 1.5 * 10-3 *BR
⎪ 2500
⎪
⎪55 /160 MBit/s L
8*10-3 * + 3.2 * 10-3 *BR
⎪⎩ 2500
L
SESR 1.5/160 MBit/s 10-4 * + 4 * 10-5 *BR (144)
2500
L
BBER 1.5/160 MBit/s 10-5 * + 4 * 10-6 *BR
2500
Tab. 31 - EPO parameters for Intermediate countries (ITU-T G.826) L > 1000 km
4.8.3.5. Example 2:
ESR, SESR and BBER evaluation for a link length of 105 km.
140 Mbit/s (G. 826) capacity intermediate country is assumed.
The objectives are calculated from previous formula
BR=1

ESR 8 *10–3 (1+1)*105/2500 6.72*10–4 ES/month = 1742

–4 –6
SESR 10 (1+1)*105/2500 8.4*10 SES/month = 22
–5 –7
BBER 10 (1+1)*105/2500 8.4 *10 BBE/month = 17418
BR=0
–3
ESR 8 *10 (1+0)*105/2500 3.36*10–4 ES/month = 871
–4 –6
SESR 10 (1+0)*105/2500 4.2*10 SES/month = 11
–5 –7
BBER 10 (1+0)*105/2500 4.2 *10 BBE/month = 8709
Tab. 32 – ITU-R F. 1397 example (2)
4.8.3.6. EPO parameters for Terminating countries (ITU-T G.826)
⎧ L
⎪1.5/5 MBit/s 2*10-3 (1+BR )*
2500
⎪
⎪5/15 MBit/s L
⎪⎪ 2.5*10-3 (1+BR )*
2500
ESR ⎨
⎪15/55 MBit/s L
3.75*10-3 (1+BR )*
⎪ 2500
⎪
⎪55 /160 MBit/s L
8*10-3(1+BR )*
⎪⎩ 2500
L
SESR 1.5/160 MBit/s 10-4 (1+BR )* (145)
2500
L
BBER 1.5/160 MBit/s 10-5 (1+BR )*
2500
Tab. 33 - EPO parameters for Terminating countries (ITU-T G.826) 50 < L < 500 km
⎧ L
⎪1.5/5 MBit/s 2*10-3 * + 4 * 10-4 *BR
2500
⎪
⎪5/15 MBit/s L
⎪⎪ 2.5*10-3 * + 5 * 10-4 *BR
2500
ESR ⎨
⎪15/55 MBit/s L
3.75*10-3 * + 7.5 * 10-4 *BR
⎪ 2500
⎪
⎪55 /160 MBit/s L
8*10-3 * + 1.6 * 10-3 *BR
⎪⎩ 2500
L
SESR 1.5/160 MBit/s 10-4 * + 2 * 10-5 *BR (146)
2500
L
BBER 1.5/160 MBit/s 10-5 * + 2 * 10-6 *BR
2500
Tab. 34 - EPO parameters for Terminating countries (ITU-T G.826) L > 500 km
4.8.4. ITU-R P.1189 (National portion) – superseded

It concerns the national portion of the HRDP that is subdivided into three basic sec-
tions:
Ö Access
Ö Short haul
Ö Long haul

PEP LE P C /S C /T C IG
S ho rt Lo ng
A c ce ss
H aul H aul
Fig. 71 – HRDP national portion
Performance objectives are fixed for each of the three types of link, just for path
level, according to the following table where the values for the B parameter are
fixed as
Ö A1+.001*[L]/500 (1%<A1<2%) long haul [L] is the route length
rounded up to the next multiple of 500 km
Ö 7.5%<B<8.5% short haul
Ö 7.5%<B<8.5% access
MBit/s 1.5 - 5 5 - 15 15 - 55 55 - 160
ESR .04*B .05*B .075*B .16*B
SESR .002*B
BBER .0002*B
Tab. 35 – ITU-R P.1189 National portion objectives
The values indicated can be reallocated in different way within the na-
tional portion of the network taking into account that
• The sum of 3 block allowance factors shall not exceed 17.5% (accord-
ing to ITU-T G.826/8)
• The sum resulting from short and long haul contributions are in the
range 15.5% to 16.5%
4.8.5. ITU-R F. 1491 (National portion) - superseded

It’s referring to the national portion of the HRDP considering that ITU-R F. 1189 has
been deleted during 2001/2002.
It defines a rule in order to indicate the objectives based on real link length and it
should be used for path, multiplex and regenerator sections performances.
The parameters used for the performance objectives are defined in
Ö G. 826-828 for path
Ö G. 829 for multiplex and regenerator sections
Assuming

⎧ L
⎪( A1 + 0.002 ) 50 < L < 100 km
B= ⎨ 100
⎪A +2*10-5 *L (147)
⎩ 1
1% < A1 < 2%
Mbit/s 1.5 - 5 5 - 15 15 -55 55 -160

ESR 0.04*B 0.05*B 0.075*B 0.16*B
SESR 0.002*B
BBER 2*B*10-4 B*10-4
Tab. 36 - ITU-R F. 1491 objectives for G.826
Mbit/s 1664 6848 48960 150336

2240
ESR 0.01*B 0.02*B 0.04*B
SESR 0.002*B
BBER 5*B*10-5 1*10-4*B
Tab. 37 - ITU-R F. 1491 objectives for G.828
4.8.5.1. ITU-R F. 1491 - Example 1
ACCESS portion of the network 2 Mbit/s (2000 blocks/second) is 50 Km long

formed by a single link -> the objective is length independent and assuming
B=.075 we have :
G. 826
ESR 0.04*B 3*10-3 7776 ES/month
-4
SESR .002*B 1.5*10 389 SES/month
-4 -5
BBER 2*10 *B 1.5*10 77760 EB/month
G. 828
ESR 0.01*B 7.5*10-4 1944 ES/month
-4
SESR 0.002*B 1.5*10 389 SES/month
-5 -6
BBER 5*10 *B 3.75*10 19440 EB/month
Tab. 38 - ITU-R F. 1491 - Calculation examples (1)
4.8.6. ITU-R F. 1491 - Example 2

ACCESS portion of the network 2 Mbit/s (2000 blocks/second) is 50 Km long
formed by 2 links ( 20 Km + 30 Km ) -> the objective is length independent but
must be apportioned for each link .
Assuming B=.075 we have ( G. 826 )
20 Km 20 Km 30 Km 30 Km 50 Km
Seconds Seconds Seconds

ESR 0.04*B*20/50 3110 0.04*B*30/50 4666 7776

SESR 0.002*B*20/50 156 0.002*B*30/50 233 389
-4 -4
BBER 2*10 *B*20/50 31104 2*10 *B*30/50 46656 77760
4.8.6.1. ITU-R F. 1491 - Example 3
SHORT HAUL portion of the network 34 Mbit/s (8000 blocks/second) is 50 Km long

formed by a single link -> the objective is length independent and assuming
B=.075 we have :
G. 826
ESR 0.075*B 5.63*10-3 14580 ES/month
-4
SESR 0.002*B 1.5*10 389 SES/month
-4 -5
BBER 2*10 *B 1.5*10 311040 EB/month
4.8.6.2. ITU-R F. 1491 - Example 4
LONG HAUL portion of the network 155 Mbit/s (8000 blocks/second) is 105 Km long
formed by a single link ( .01 < A < .02 )
G. 826
ESR 0.16*B 0.16*(A+.002)*105/100 5225< ES/month < 9580
SESR 0.002*B 0.002*(A+.002)*105/100 65 < SES/month < 119
-4 -5
BBER 2*10 *B 0.0002*(A+2*10 *105) 17418<EB/month < 26127
G828
ESR 0.04 * B 0.04*(A+.002)*105/100 1306 < ES/month < 2395
SESR 0.002*B 0.002*(A+.002)*105/100 65 < SES/month < 119
-4 -5
BBER B*10 0.0001*(A+2*10 *105) 26127<EB/month < 45619
4.8.7. ITU-R F. 1668 (National and International portion)

Ö This is the only Rec. actually in force defining EPO for real digital fixed wire-
less links.
Ö It deals with both national and international portion.
Ö It defines the objectives according to G.826 and G.828 respectively.
Ö It deals with all bit rates (up to 400 Mbit/s for G.826; up to 150 Mbit/s for
G.828).
Ö Concerning rates at or above the primary, the objectives defined in this Rec.

are the same fixed in superseded F.1397, F.1491.

Ö According to G.826 (12/2002), this Rec. introduced the objectives also for
the rates below the primary
4.8.7.1. EPO parameters for Intermediate countries (ITU-T G. 828)
⎧ L
⎪1.664/6.848 MBit/s 5*10-4 (1+BR )*
2500
⎪
⎪ L
ESR ⎨48.960 MBit/s 10-3 (1+BR )*
⎪ 2500
⎪ L
⎪150.336 MBit/s 2*10-3 (1+BR )*
⎩ 2500
L
SESR 1.664/150.336 MBit/s 10-4 (1+BR )* (148)
2500
⎧ L
⎪⎪1.664/48.960 MBit/s 2.5*10-6 (1+BR )*
2500
BBER ⎨
⎪150.336 MBit/s L
5*10-6 (1+BR )*
⎪⎩ 2500
Tab. 42 – ITU-R F. 1668 EPO for Intermediate countries (ITU-T G. 828) 50 < L < 1000 km
⎧ L
⎪1.664/6.848 MBit/s 5*10-4 * + 2 * 10−4 * BR
2500
⎪
⎪ L
ESR ⎨48.960 MBit/s 10-3 * + 4 * 10−4 * BR
⎪ 2500
⎪ L
⎪150.336 MBit/s 2*10-3 * + 8 * 10−4 * BR
⎩ 2500
L
SESR 1.664/150.336 MBit/s 10-4 * + 4 * 10−5 * BR (149)
2500
⎧ L
⎪⎪1.664/48.960 MBit/s 2.5*10-6 * + 1 * 10−6 * BR
2500
BBER ⎨
⎪150.336 MBit/s L
5*10-6 * + 2 * 10−6 * BR
⎪⎩ 2500
Tab. 43 - ITU-R F. 1668 EPO for Intermediate countries (ITU-T G. 828) L > 1000 km
4.8.7.2. EPO parameters for Terminating countries ( ITU-T G.828)
⎧ L
⎪1.664/6.848 MBit/s 5*10-4 (1+BR )*
2500
⎪
⎪ L
ESR ⎨48.960 MBit/s 10-3 (1+BR )*
⎪ 2500
⎪ L
⎪150.336 MBit/s 2*10-3 (1+BR )*
⎩ 2500
L
SESR 1.664/150.336 MBit/s 10-4 (1+BR )* (150)
2500
⎧ L
⎪⎪1.664/48.960 MBit/s 2.5*10-6 (1+BR )*
2500
BBER ⎨
⎪150.336 MBit/s L
5*10-6 (1+BR )*
⎪⎩ 2500
Tab. 44 - ITU-R F. 1668 EPO parameters for Terminating countries ( ITU-T G.828) 50 < L < 500 km

⎧ L
⎪1.664/6.848 MBit/s 5*10-4 * + 10−4 * BR
2500
⎪
⎪ L
ESR ⎨48.960 MBit/s 10-3 * + 2 * 10−4 * BR
⎪ 2500
⎪ L
⎪150.336 MBit/s 2*10-3 * + 4 * 10−4 * BR
⎩ 2500
L
SESR 1.664/150.336 MBit/s 10-4 * + 2 * 10−5 * BR (151)
2500
⎧ L
⎪⎪1.664/48.960 MBit/s 2.5*10-6 * + 5 * 10−7 * BR
2500
BBER ⎨
⎪150.336 MBit/s L
5*10-6 * + 10−6 * BR
⎪⎩ 2500
Tab. 45 - ITU-R F. 1668 EPO parameters for Terminating countries ( ITU-T G.828) L > 500 km
4.8.7.3. EPO parameters for Intermediate countries (ITU-T G.826)
⎧ L
⎪< 5 MBit/s 2*10-3 (1+BR )*
2500
⎪
⎪5/15 MBit/s L
⎪⎪ 2.5*10-3 (1+BR )*
2500
ESR ⎨
⎪15/55 MBit/s L
3.75*10-3 (1+BR )*
⎪ 2500
⎪
⎪55 /160 MBit/s L
8*10-3 (1+BR )*
⎪⎩ 2500
L
SESR ≤ 400 MBit/s 10-4 (1+BR )* (152)
2500
L
BBER 1.5/400 MBit/s 10-5 (1+BR )*
2500
Tab. 46 - ITU-R F. 1668 EPO for Intermediate countries (ITU-T G.826) 50 < L < 1000 km
⎧ L
⎪<5 MBit/s 2*10-3 * + 8 * 10-4 *BR
2500
⎪
⎪5/15 MBit/s L
⎪⎪ 2.5*10-3 * + 10-3 *BR
2500
ESR ⎨
⎪15/55 MBit/s L
3.75*10-3 * + 1.5 * 10-3 *BR
⎪ 2500
⎪
⎪55 /160 MBit/s L
8*10-3 * + 3.2 * 10-3 *BR
⎪⎩ 2500
L
SESR ≤ 400 MBit/s 10-4 * + 4 * 10-5 *BR (153)
2500
L
BBER 1.5/400 MBit/s 10-5 * + 4 * 10-6 *BR
2500
Tab. 47 - ITU-R F. 1668 EPO parameters for Intermediate countries (ITU-T G.826) L >1000 km

4.8.7.4. EPO parameters for Terminating countries (ITU-T G.826)
⎧ L
⎪<5 MBit/s 2*10-3 (1+BR )*
2500
⎪
⎪5/15 MBit/s L
⎪⎪ 2.5*10-3 (1+BR )*
2500
ESR ⎨
⎪15/55 MBit/s L
3.75*10-3 (1+BR )*
⎪ 2500
⎪
⎪55 /160 MBit/s L
8*10-3 (1+BR )*
⎪⎩ 2500
L
SESR ≤ 400 MBit/s 10-4 (1+BR )* (154)
2500
L
BBER 1.5/400 MBit/s 10-5 (1+BR )*
2500
Tab. 48 - ITU-R F. 1668 EPO parameters for Terminating countries (ITU-T G.826) 50 < L < 500 km
⎧ L
⎪<5 MBit/s 2*10-3 * + 4 * 10-4 *BR
2500
⎪
⎪5/15 MBit/s L
⎪⎪ 2.5*10-3 * + 5 * 10-4 *BR
2500
ESR ⎨
⎪15/55 MBit/s L
3.75*10-3 * + 7.5 * 10-4 *BR
⎪ 2500
⎪
⎪55 /160 MBit/s L
8*10-3 * + 1.6 * 10-3 *BR
⎪⎩ 2500
L
SESR ≤ 400 MBit/s 10-4 * + 2 * 10-5 *BR (155)
2500
L
BBER 1.5/400 MBit/s 10-5 * + 2 * 10-6 *BR
2500
Tab. 49 - ITU-R F. 1668 EPO parameters for Terminating countries (ITU-T G.826) L > 500 km
4.8.7.5. EPO parameters for Long haul
Assuming
⎧ L
⎪( A1 + 0.002 ) 50 < L < 100 km
A= ⎨ 100
⎪A +2*10-5 *L (156)
⎩ 1
1% < A1 < 2%
Mbit/s 1664- 2240 6848 48960 150336

ESR 0.01*A 0.02*A 0.04*A
SESR 0.002*A
BBER 5*A*10-5 1*10-4*A
Tab. 50 - ITU-R F. 1668 EPO parameters for Long Haul (ITU-T G.828)
Mbit/s <5 5 - 15 15 -55 55 -160
ESR 0.04*A 0.05*A 0.075*A 0.16*A
SESR 0.002*A
BBER 2*A*10-4 A*10-4
Tab. 51 - ITU-R F. 1668 EPO parameters for Long Haul (ITU-T G.826)

4.8.7.6. EPO parameters for Short haul
Assuming 7.5% < B < 8.5 % ( provisional value )

Mbit/s 1664 - 2240 6848 48960 150336
ESR 0.01*B 0.02*B 0.04*B
SESR 0.002*B
BBER 5*B*10-5 1*10-4*B
Tab. 52 - ITU-R F. 1668 EPO parameters for Short Haul (ITU-T G.828)
Mbit/s <5 5 - 15 15 -55 55 -160

ESR 0.04*B 0.05*B 0.075*B 0.16*B
SESR 0.002*B
BBER 2*B*10-4 B*10-4
Tab. 53 - ITU-R F. 1668 EPO parameters for Short Haul (ITU-T G.826)
4.8.7.7. EPO parameters for Access
Assuming 7.5% < C < 8.5 % ( provisional value )

Mbit/s 1664 - 2240 6848 48960 150336
ESR 0.01*C 0.02*C 0.04*C
SESR 0.002*C
BBER 5*C*10-5 1*10-4*C
Tab. 54 - ITU-R F. 1668 EPO parameters for Access (ITU-T G.828)
Mbit/s <5 5 - 15 15 -55 55 -160

ESR 0.04*C 0.05*C 0.075*C 0.16*C
SESR 0.002*C
BBER 2*C*10-4 C*10-4
Tab. 55 - ITU-R F. 1668 EPO parameters for Access (ITU-T G.826)
4.8.7.8. NOTES
Ö EPOs ARE applicable to each direction of any real fixed wireless link of length
> 50 km only when the system is considered to be available.
Ö The objectives given in this Recommendation are understood to be long-
term objectives to be met over an evaluation period of a month, which is
understood to be any period of 28 to 31 (typically 30) consecutive 24 h in-
tervals. These objectives should be respected for any month (see Recom-
mendation ITU R P.581).

Ö The sum of the percentages A1%  B%  C% should not exceed 17.5%, in ac-
cordance with the allocations to the national portion of an international CBR
path or connection given in ITU-T Recommendations G.826 and G.828.
Ö The provisional values agreed for B% + C% are in the range 15.5% to
16.5%.
Ö Depending on national network configurations administrations may reallo-
cate the A%, B% and C% block allowances among the sections of the na-
tional portion of a radio path.
Ö In the case of multi-hop links the objectives derived according to this Rec-
ommendation apply to the overall links (irrespective of the date when each
hop was brought into service and of the number of independent operators
involved); the allocation of the objectives to each hop is under the responsi-
bility of the network operators.
4.8.8. ITU-R F. 1668 - Example 1

General assumptions: international portion, intermediate country, L = 105 km, BR =
1.
4.8.8.1. Bit rate: 150 336 kbit/s (ITU T G.828)
ESR = 2 x 10–3 (1 + 1) x 105/2500 = 168 x 10–6 ES/month = 435

–4 –7
SESR = 1 × 10 (1 + 1) × 105/2500 = 84 × 10 SES/month = 22
BBER = 5 × 10–6 (1 + 1) × 105/2500 = 4.2 × 10–7 BBE/month = 8709
4.8.8.2. Bit rate: 140 Mbit/s (ITU T G.826)
ESR = 8 × 10–3 (1 + 1) × 105/2500 = 672 × 10–6

ES/month = 1 741
–4
SESR = 10 (1 + 1) × 105/2500 = 84 × 10–7 SES/month = 22
–5 –7
BBER = 10 (1 + 1) × 105/2500 = 8.4 × 10 BBE/month = 17 418
4.8.8.3. Bit rate: 64 kbit/s (ITU T G.826)
ESR = 2 × 10–3 (1 + 1) × 105/2500 = 168 × 10–6 ES/month = 436

–4 –7
SESR = 10 (1 + 1) × 105/2500 = 84 × 10 SES/month = 22

4.8.9. ITU-R F. 1668 - Example 2 – Access

The Access portion of the network is 20 km long and it is formed by a single link
operating at 2 MBit/s
In this case the objectives are length independent; if C = 7.5% is assumed we
have:
ESR = 0.04 C = 3 × 10–3 ES/month
= 7 776
SESR = 0.002 C = 1.5 × 10–4 SES/month = 389
–4 –5
BBER = 2 × 10 × C = 1.5 × 10 EB/month = 77 760
4.8.10. ITU-R F. 1668 - Example 3: Short haul

Short haul portion of the network is 80 km long and it is formed by a single link op-
erating at 34 MBit/s
In this case the objectives are length independent; if B = 7.5% we have:
ESR = 0.075 B = 5.625 × 10–3 ES/month = 14 580
–4
SESR = 0.002 B = 1.5 × 10 SES/month = 389
–4 –5
BBER = 2 × 10 × B = 1.5 × 10 EB/month = 311 040
4.8.11. ITU-R F. 1668 - Example 4: Long Haul

Real link in long haul portion ( 75 km ) of the network, designed according to ITU-T
G.826 (before the adoption of ITU-T G.828 in March 2000) with SDH transmis-
sion rate STM-1 (155.52 Mbit/s):
ESR = 0.16 A = 0.16 (A1 + 0.002) × 75/100
SESR = 0.002 A = 0.002 (A1 + 0.002) × 75/100
BBER = 0.0002 A = 0.0002 (A1 + 0.002) × 75/100
In this case the objectives are length dependent and in the next table the minimum
and the maximum limits (A1 = 0.01 and A1 = 0.02) are shown:
A1 value ESR SESR BBER
–3 –5
0.01 1.44 × 10 1.8 × 10 1.8 × 10–6
(= 3 733 ES/month) (= 47 SES/month) (= 37 324 EB/month)
–3 –5
0.02 2.64 × 10 3.3 × 10 3.3 × 10–6
(= 6 843 ES/month) (= 86 SES/month) (= 68 429 EB/month)

4.8.12. ITU-R F. 1668 - Example 5: Long Haul

Real link in long haul portion ( 75 km ) of the network, designed after the adoption
of ITU-T G.828 in March 2000, with SDH transmission rate STM-1 (155.52
Mbit/s):
ESR = 0.04 A = 0.04 (A1 + 0.002) × 75/100
SESR = 0.002 A = 0.002 (A1 + 0.002) × 75/100
BBER = 0.0002 A = 0.0001 (A1 + 0.002) × 75/100
In this case the objectives are length dependent and in the next table the minimum
and the maximum limits (A1 = 0.01 and A1 = 0.02) are shown:
A1 value ESR SESR BBER
–3 –5
0.01 3.6 × 10 1.8 × 10 9 × 10–7
(= 933 ES/month) (= 47 SES/month) (= 18662 EB/month)
–3 –5
0.02 6.6 × 10 3.3 × 10 1.65 × 10–6
(= 1711 ES/month) (= 86 SES/month) (= 34214 EB/month)
4.8.13. ITU-R F. 1668 - Example 6 : National portion at 64

KBit/s
The connection consists on access portion of the network formed by a single link L1
= 10 km and short haul portion of the network L2 = 100 km

In this case the objectives are length independent; if B + C = 15% is assumed we
have:
ESR = 0.04 (B + C) = 6 × 10–3 ES/month = 15552
SESR = 0.002 (B + C) = 3 × 10–4 SES/month = 778

4.9. Annex 1: Relationship between PDH path perfor-

mance monitoring and the block-based parameters
(G.826 / Annex B - extract)
4.9.1. Anomalies
In-service anomaly conditions are used to determine the error performance of a
PDH path when the path is not in a defect state. The two following categories of
anomalies related to the incoming signal are defined:
Ö a1 an errored frame alignment signal;
Ö a2 an EB as indicated by an EDC.
4.9.2. Defects
In-service defect conditions are used in the G.730 to G.750 series of Recommenda-
tions relevant to PDH multiplex equipment to determine the change of performance
state which may occur on a path. The three following categories of defects related
to the incoming signal are defined:
Ö loss of signal;
Ö alarm indication signal;
Ö loss of frame alignment.
For the 2 Mbit/s hierarchy, the definition of the LOF defect condition is given in the
G.730 to G.750 series of Recommendations.
For some formats of the 1.5 Mbit/s hierarchy, the definition of the LOF defect condi-
tion requires further study
For both hierarchies, the definitions of LOS and AIS defect detection criteria are
given in ITU-T Rec. G.775.

5. Unavailability
5.1. Introduction
As already stated in the previous chapter we’ve to take in account that general pur-
pose rules ( media independent ) are fixed by ITU-T ( transmission branch ) and
“adapted” to radio equipment by ITU-R ( radio ) committee.
It must be stressed that definition of unavailability is yearly based and

that no rule can be applied to scale this value on a different time base.
5.1.1. Media independent (ITU-T)

Ö G.821 International connection below the primary rate (It contains the avail-
ability definitions in terms of events and network components) in case of
connections operating below the primary rate and it doesn’t fix any objec-
tive.
Ö G.827 International constant bit rate (CBR) digital paths
While the present edition of the Rec. deals with all the bit rates, the pre-
vious one (03/2000) was only for bit rates at or above the primary rate
5.1.2. Radio specific (ITU-R)

Ö F. 695, F. 557 (HRDP), F. 696 (HRDP), F. 697, F. 1053 ( G. 821)
Ö F.1493 ( G. 826/8 national portion ) : 05/2000
Ö F.1492 ( G. 826/8 international portion ) : 05/2000
Ö F. 1703 : DRAFT 01/2005
In the following a very short explanation of all these rules is reported in order to
clarify their applicability and to avoid some dangerous pitfalls.

UNAVAILABILITY 2000 2001 2002 2003 2004 2005

< 2 Mbit/s F. 695 F.696-2 F.1053 DRAFT
PDH
> 2 Mbit/s F.1492 F.1493 F. 1703
SDH F.1492 F.1493
Tab. 56 – ITU-R availability Recs
5.2. Unavailability definition

Entry and exit definitions relevant to unavailable period are specified in G. 826, G.
828 or Annex A/G. 821.
5.2.1. Criteria for a single direction

Ö A period of unavailable time begins at the onset of ten consecutive SES
events that are considered to be part of unavailable time.
Ö A new period of available time begins at the onset of ten consecutive non-
SES events and these ten seconds are considered to be part of available
time.
Fig. 72 – Mono-directional unavailability
5.2.2. Criterion for a Bi-directional Path

A bi-directional path is in the unavailable state if either one or both directions are in
the unavailable state.
Fig. 73 – Bi-directional unavailability

5.2.3. Some additional considerations
This objective includes all causes which are statistically predictable, un-
intentional and resulting from
• all equipment within a digital radio section ( MTBF and MTTR )
• power supplies
• propagation and /or interference from auxiliary equipment
• human activity
5.3. ITU-T G. 821 Unavailability evaluation
5.3.1. Rec. ITU-R F. 695 : High grade portion

Availability objective appropriate to a real digital radio included in a high grade por-
tion for a link with length, L, of between 280 km and 2500 km, should be :
A = 100 – (0.3 × L /2500) %
the percentage being considered over a period of time sufficiently long to be statis-
tically valid ( > 1 year )
It must be noted that if the connection length is < 280 Km no scaling

rule is applicable.
Obviously this limit must be applied for a path and NOT to each link be-
longing to the same connection!
5.3.2. Rec. ITU-R F. 696 : Medium grade portion

Class Unav Length
[%] [km]
1 0.033 280
2 0.05 280
3 0.05 50
4 0.1 50
Tab. 57 – ITU-R F. 696 : Availability objectives for Medium Grade
Four classes are defined for the purpose of partitioning the path (it must be decided
if by step of 50 or 280 km) and choosing between two different allocation values.
The indicated class lengths are those identified in Rec. G.801 and they should be

understood to correspond to maximum lengths.
5.3.3. Rep. ITU-R F. 1053 : Local grade portion
Report ITU-R F. 1053 (Please note that it’s a Report and not a Recom-
mendation !) states that a typical value of 1% can be used as unavaila-
bility target in case of local grade.
5.4. ITU-T G.827 (03/2000) superseded

Performance objectives are defined for specific elements of the international Path
Elements (PEs ) that are categorized according to their position in the path and
their length
These paths may be based on the PDH, SDH or some other transport network such
as cell based
Two types of paths are considered
Ö Paths between International Switching Centers (ISCs) which consist of an in-
ternational portion only ( Type a )
Fig. 74 - ITU-T G.827 : type a Path
Ö Paths which extend beyond the ISC and consist of national and international
portions (Type b ).

Fig. 75- ITU-T G.827 : type b Path
This Rec. specifies objectives for the availability performance of each of these por-
tions
5.4.1. Definitions
Ö Path: is a transport entity responsible for the integrity of client network in-
formation transfer
Ö Path element: PE is a portion of a path resulting from partitioning for the
purpose of availability management.
Ö Network location. Paths may traverse different portions of networks with dif-
ferent availability performance characteristics
Ö Inter-Country Path Core Element (ICPCE)
Ö International Path Core Element (IPCE)
Ö National Path Element (NPE, applicable only to paths of type “b”).
Ö Availability state is defined as in the Rec. G.821
Ö Availability Ratio (AR) : it is defined as the portion of time that a PE is in the
available state and it’s calculated by dividing the total available time during
the observation period by the duration of the observation period -> AR + UR
=1
Ö The observation period is recommended to be one year
Ö Mean time between digital path outages Mo: it is the average duration of
any continuous interval during which the portion is available.
Ö Outage Intensity (OI) : OI = 1/Mo
5.4.2. Availability performance objectives

Ö Performance objectives for PEs are derived using a fixed block allocation plus

a distance-based allocation
Ö Two types of objectives are specified:
mean values, i.e. the ensemble average of all PEs of a given category in
a country
worst-case values, i.e. the minimum acceptable value for individual PEs
The objectives apply to observation periods of 1 year ( 365 days ) , us-

ing a sliding window with 24-hour granularity.
Ö The objectives are intended for the following purposes

Network design/planning
Both mean and worst-case objectives are appropriate
Operational
Worst case objectives are appropriate for operational purposes.
To meet these requirements, it may be necessary for design purposes to
use more stringent values
Ö Meeting the allocated objectives for the highest bit-rate path should be suffi-
cient to ensure that all paths through the system achieve their objectives.
5.4.3. Availability ratio

j ARj Length [Km]
1 1 – ( bjn + i * xjn ) L < 2500
2 1 – ( bjn + ( i – 5 ) * xjn ) 2500 < L < 7500
3 1 – ( bjn + ( i – 15 ) * xjn ) 7500 < L < 10000
4 1 – ( bjn + 21 * xjn ) L > 10000
Tab. 58 – ITU-T G.827 availability ratio for path elements operating at primary rate
Where
Ö J NPE, IPCE, ICPCE
Ö bjn block allowance for PE type j and length range n
Ö i length category, such that 500⋅(i -1) km ≤ L < 500⋅i km
Ö xjn distance-based allowance for PE type j and length range n
Proper values of these coefficients are reported in the next tables
Previous figures are applicable to primary bit rate ONLY considering that
no figures are available for higher bit rates

5.4.3.1. Availability ratio
PE type bj1 xj1 bj2 xj2 bj3 xj3

NPE .002 .01
IPCE 0 .0015 .0075 t.b.d.
ICPCE .002 .01
Tab. 59 – ITU-T G.827 objectives for availability ratio : mean values
NPE .0052 .0047 .0287
IPCE .004 .0035 .0215 t.b.d.
ICPCE .0052 .0047 .0287
Tab. 60 - ITU-T G.827 objectives for availability ratio : worst values
ITU-T G.827
3.5%
3.0%
2.5%
1- Availability ratio
2.0%
1.5%
1.0%
0.5%
0.0%
0 500 1000 1500 2000 2500
Connection length [km]
IPCE (mean) ICPCE (mean) IPCE (worst) ICPCE (worst)
Fig. 76 – ITU-T G.827 : availability ratio depending on PCE
5.4.4. Outage intensity

J OIj Length [Km]
1 1 / ( bjn + i * xjn ) L < 2500
2 1 / ( bjn + ( i – 5 ) * xjn ) 2500 < L < 7500
3 1 / ( bjn + ( i – 15 ) * xjn ) 7500 < L < 10000
4 1 / ( bjn + 21 * xjn ) L > 10000
Tab. 61 - ITU-T G.827 outage intensity for path elements operating at primary rate
Where
Ö J NPE, IPCE, ICPCE

Ö bjn block allowance for PE type j and length range n

Ö i length category, such that 500⋅(i -1) ≤ L < 500⋅i
Ö xjn distance-based allowance for PE type j and length range n
ITU-T G.827
800
700
600
NPE (mean)
500 IPCE (mean)
NPE (w orst)
IPCE (w orst)
Mo
400
ICPCE (mean)
300
ICPCE (w orst)
200
100
0
0 500 1000 1500 2000 2500
C onnection length [km]
Fig. 77 – ITU-T G.827 : Values for mean time between digital path outage calculations

NPE 57 42 267
IPCE 30 20 130 t.b.d.
ICPCE 18 13 83
Tab. 62 - Values for mean time between digital path outage objectives

NPE 443 58 733
IPCE 222 27 357 t.b.d.
ICPCE 130 20 230
Tab. 63 - Values for mean time between digital path outage objectives
5.4.5. ITU-R F. 1492 – Unavailability objectives for interna-

tional portion
Ö Availability objectives for any real digital radio relay link forming part of the
international portion of a CBR digital path at or above the primary rate
should be fixed block plus distance base allocated considering a minimum
length of 50 km;

5.4.5.1. ITU-R F. 1492 – Availability ratio
⎧ ⎛ -3 L ⎞
⎪1- ⎜1.9*10 * +1.1*10-4 ⎟ L < 250 km
⎪ ⎝ 2500 ⎠
AR = ⎨
⎪1- ⎛ 3*10-3 * L ⎞ 250 < L < 7500 km
⎪⎩ ⎜⎝ 2500 ⎟⎠
ITU-T G.827
100.00%
10.00%
1- Availability ratio
1.00%
0.10%
0.01%
0 500 1000 1500 2000 2500
IPCE (mean) ICPCE (mean) IPCE (worst) ICPCE (worst) ITU-R F. 1492
Fig. 78 – A comparison between ITU-T G.827 and ITU-R F. 1492 availability ratio parameters
5.4.5.2. ITU-R F. 1492 – Outage Intensity
⎧ L
⎪⎪150* 2500 + 50 50 < L < 250 km
OI = ⎨ (157)
⎪100* L + 55 250 < L < 2500 km
⎪⎩ 2500
Tab. 64 – ITU-R F. 1492 outage intensity for path elements operating at primary rate

ITU-R F.1492 vs. ITU-T G.827

800
700
600
NPE (mean)
IPCE (mean)
500
NPE (worst)
IPCE (w orst)
Mo
400 ICPCE (mean)

ICPCE (w orst)
300 ITU-R F.1492
200
100
0
0 500 1000 1500 2000 2500
Fig. 79 – A comparison between ITU-T G.827 and ITU-R F.1492 for mean time between digital path outage
5.4.6. ITU-R F. 1492 – Calculation examples
5.4.6.1. Example 1 - hop length = 20 Km
The length is shorter than L=50 km, so the value of L=50 should be used.
⎛ 50 ⎞
AR = 1- ⎜ 1.9*10-3 * +1.1*10-4 ⎟ = 0.99985
⎝ 2500 ⎠
1 1 (158)
MO = = =18.86*10-3
50 53
150* +50
2500
Ö unavailability of 78 min/year
Ö number of events per year OI=53
Ö mean time between unavailability events MO=9917 min
The length is in the range 50-250 Km so :

⎛ 80 ⎞
AR = 1- ⎜ 1.9*10-3 * +1.1*10-4 ⎟ = 0.99983
⎝ 2500 ⎠
1 1 (159)
MO = = =18.25*10-3
80 54.8
150* +50
2500

Ö mean time between unavailability events MO=9591 min.
The length is in the range 250-2500 Km so :

⎛ 1056 ⎞
AR = 1- ⎜ 3*10-3 * = 0.998732
⎝ 2500 ⎟⎠
1 1 (160)
MO = = =10.28*10-3
1056 53
100* +55
2500

Ö mean time between unavailability events MO=5405 min.
5.4.7. ITU-R F. 1493 – Unavailability objectives for national

portion
Ö availability objectives applicable to any real digital radio link belonging to the
long haul network portion of the national portion of an international constant
bit rate digital path at or above the primary rate should be fixed block plus
distance-base allocated;
Ö availability objectives applicable to any real digital radio link belonging to the
access and short haul network portions of the national portion of an interna-
tional constant bit rate digital path at or above the primary rate should be
fixed block-based (i.e. length independent);
Ö availability objectives applicable to each direction of a radio-relay link of
length L in the national portion at or above the primary rate can be derived
as
Access AR =1-5*10-4 50 < L < 2500 km
Short haul AR =1-4*10-4 50 < L < 2500 km (161)
⎧ -3 L
⎪⎪AR =1-3*10 * 2500 50 < L < 250 km
Long haul ⎨
⎪A =1- ⎛1.9*10-3 * L +1.1*10-4 ⎞ 250 < L < 2500 km
⎪⎩ R ⎜ 2500 ⎟
⎝ ⎠
Tab. 65 – ITU-R F. 1493 : Availability ratio

Access OI=100 50 < L < 2500 km

Short haul OI=120 50 < L < 2500 km (162)
⎧ L
⎪⎪OI=100* 2500 +55 50 < L < 250 km
Long haul ⎨
⎪O =150* L +50 250 < L < 2500 km
⎪⎩ I 2500
Tab. 66 – ITU-R F. 1493 : Outage Intensity
5.4.8. ITU-R F. 1493 – Calculation examples

Km Avail. Ratio Unavail. OI MO
Access 40 99.950% 263 m/y 100 5256
Short H. 130 99.960% 210 m/y 120 4380
Long H. 950 99.886% 599 m/y 93 5651
Tab. 67 – ITU-R F.1493 calculation example
For the whole link consisting of 40 + 130 + 950 Km (ITU-R F.1493) :

UR=URA+URSH+URLH=0.00204 Î 0.00204*60*24*365 = 1072 m/y
AR=1–UR=0.99796 (99.796%)
OI = OIA+OISH+OILH = 100+120+93 = 313 events/y
MO = 1/OI = 1/ 313 = 3.19*10-3 = 1677 m/y
G827 objectives for 1120 Km NPE are:
UR Î worst = 1.93%= 0.0193; mean = 0.6%=0.006
OI Î worst = 617 events/year; mean = 183 e/y (warning)
MO Î worst = 852 min; mean = 2873 min (warning)

5.5. ITU-T G.827 (09-2003)

Performance objectives are defined for
Ö End-to-end international CBR digital paths
Ö Specific elements of the international path (PATH ELEMENTS - PEs) defined
according to their position in the path, their length and - this is the differ-
ence introduced by the new edition of this Rec. - to their performance level
(indicated as priority).
The paths are based on the PDH, SDH or the Optical Transport Network - OTN (fu-
ture revisions of this Recommendation may consider the availability specification of
other transmission technologies).
Fig. 80 – International PCE
The Rec. specifies objectives for the availability performance of each path element:
NPE, IPCE, ICPCE (defined in the next paragraph).
5.5.1. Definitions
Ö End-to-end path: is a transport entity responsible for the integrity of client
information transfer between path end points.
Ö Path element: PE is a portion of an end-to-end path resulting for the purpose
of availability specifications.
Ö Network location. Paths may traverse different portions of networks with dif-
ferent availability performance characteristics. Depending on the network lo-
cation, three different geographical categories are distinguished.
Ö Inter-Country Path Core Element (ICPCE)
Ö International Path Core Element (IPCE)

Ö National Path Element (NPE, applicable only to national portion of the path).
Ö Each path element can carry data at three performance priority levels:
standard
high
pre-emptible
The "pre-emptible" priority level does not have a defined availability

performance but it might be interrupted to provide protection capacity
when required for another path. The "standard" performance category is
the minimum requirement. The "high" performance category is a level of
performance above the "standard" level.
Ö Availability state is defined as in G. 826, G. 828 or Annex A/G. 821.

Ö Availability Ratio (AR) : it is defined as the fraction of time such that a PE is
in the available state and it’s calculated by dividing the total available time
during the observation period by the duration of the observation period ->
AR + UR = 1
Ö Mean time between digital path outages Mo: it is the average duration of any
continuous interval during which the portion is available.
Ö Outage Intensity (OI) : OI = 1/Mo
5.5.2. Availability performance objectives

Ö Performance objectives for PEs are derived using a fixed block allocation plus
a distance-based allocation.
Ö Actually, two types of objectives are specified, respectively for high and
standard priority (pre-emptible case is until under study)
In the previous edition of this Rec. the objectives were specified consi-
dering two cases: mean values and worst-case values.
Ö To meet these requirements, it may be necessary - for design purposes - to

use more stringent values.
Ö Meeting the allocated objectives for the highest bit-rate path should be suffi-
cient to ensure that all paths through the system achieve their objectives.
Ö Tab. 68 specifies the end-to-end objectives for a 27500 km HRDP at or
above the primary rate.
Ö An international digital path at or above the primary rate shall meet the ob-

jectives of Tab. 68 for all parameters concurrently.
Values for OI are based on an typical MTTR (Mean Time To Restore) val-
ue of 4 hours.
Rate 1.5 Mbit/s to 40 Gbit/s

AR OI
High priority 98% 70
Standard priority 91% 250
Pre-emptible priority F.F.S. F.F.S.
Tab. 68 - End-to-end availability objectives for a 27 500 km HRDP at or above the primary rate
The objectives apply to observation periods of 1 year (365 days), using

a sliding window with 24-hour granularity.
Ö For the calculation of AR/OI objectives of a PE, it must be considered: which

kind of PE we are dealing with, its associated priority level and its length
factor K depending on L rounded up to the next multiple of 100 km
Ö The relationships between K and L is
K = L \ 100 +1 L < 2500 km
(163)
if L = 99 km ⇒ K=1, if L=101 km ⇒ K=2 ....
5.5.3. Availability Ratio

1- AR vs. connection length
10.000%
1.000%
Unavailability ratio
0.100%
0.010%
0.001%
0 500 1000 1500 2000 2500
Connection lenght [km]
IPCE S IPCE H NPE S NPE H

Fig. 81 – ITU-T G. 827 : Unavailability Ratio ( 1 – AR ) Objectives ( S=standard, H=high)
The Path Elements availability ratio and outage intensity objectives are determined
by the following equations:
⎧standard k * 0.03% ⎧standard k * 0.04%
IPCE ⎨ NPE ⎨
⎩high k *0.006% ⎩high k *0.008% (164)
⎧standard k * 0.04%
ICPCE ⎨ k=(L\100)+1
⎩high k *0.008%
Tab. 69 Values for PEs availability ratio calculations (L < 2500 km)
5.5.4. ITU-R F. 1703 ( DRAFT New Recommendation ) –

Availability objectives for real digital fixed wireless links
used in 27 500 km hypothetical reference paths and
connections
Ö This Rec. provides updated information on availability objectives for real dig-
ital fixed wireless links used in 27500 km HRP taking into account ITU-T
G.827 approved in 2003.
Ö It is the only Recommendation defining availability objectives for all real
digital fixed wireless links.
Ö Rec. ITU R F.1492 and ITU R F.1493 are superseded by this Rec also if the
final targets are exactly the same.
Ö The applicability of ITU-R F.695, ITU-R F.696 and ITU-R F.1053 is limited to
systems designed prior to the approval of this Recommendation.
Ö The lower limit of Llink used to scale the objectives is Lmin equal to 50 km.
5.5.4.1. ITU-R F. 1703 – Parameters for AR objectives for links

forming part of an international portion of constant bit-rate digi-
tal path
⎧ ⎛ -3 L ⎞
⎪1- ⎜1.9*10 * + 1.1 * 10−4 ⎟ L< 250 km
⎪ ⎝ 2500 ⎠
AR= ⎨ (165)
⎪1-3*10-3 * L
⎪⎩ 250 < L < 10000 km
2500

Unavailability objectives for international portion ( F 1703)
0.35%
0.30%
Unavailability [%]
0.25%
0.20%
0.15%
0.10%
0.05%
0.00%
0 500 1000 1500 2000 2500
Length [km]
Fig. 82 - F. 1703 : AR objectives for international portion
5.5.4.2. ITU-R F. 1703 – Parameters for AR objectives for links

forming part of a national portion of constant bit-rate digital path
⎧99.95 % Access
⎪
⎪ 99.96 % Short haul
⎪⎪ ⎛ L ⎞
AR= ⎨1- ⎜1.9*10-3 + 1.1 * 10−4 ⎟ Long haul : L < 250 km (166)
⎪ ⎝ 2500 ⎠
⎪ ⎛ L ⎞
-3
⎪1- ⎜ 3*10 ⎟ Long haul : L > 250 km
⎪⎩ ⎝ 2500 ⎠
Unavailability objectives for national portion ( F 1703)
0.30%
0.25%
Unavailability [%]
0.20%
0.15%
0.10%
0.05%
0.00%
0 500 1000 1500 2000 2500
Length [km]
Fig. 83 - F. 1703 : AR objectives for national portion : Red (long haul), green (short haul) and
blue (Access)

5.5.4.3. ITU-R F. 1703 – Parameters for OI objectives for links

forming part of an international portion of constant bit-rate digi-
tal path
⎧ L
⎪⎪150* 2500 + 50 L < 250 km
OI = ⎨ (167)
⎪100* L + 55 L > 250 km
⎪⎩ 2500
Outage intensity objectives for international portion ( F 1703)
180
160
140
Outage intensity
120
100
80
60
40
0 500 1000 1500 2000 2500
Length [km]
Fig. 84 – F. 1703 : OI objectives for international portion
5.5.4.4. ITU-R F. 1703 – Parameters for OI objectives for links

forming part of a national portion of constant bit-rate digital path
⎧100 Access
⎪
⎪120 Short haul
⎪
OI = ⎨150* L + 50 Long haul : L < 250 km (168)
⎪ 2500
⎪ L
⎪100* + 55 Long haul : L > 250 km
⎩ 2500

Outage intensity objectives for national portion ( F 1703)
180
160
140
Outage intensity 120
100
80
60
40
0 500 1000 1500 2000 2500
Length [km]
Fig. 85 – F. 1703 OI objectives for national portion : Red (long haul), green (short haul) and
blue (Access)
• Large yearly variability of meteorological events can significantly

modify foreseen outage periods
• The amount of other unavailability sources ( interferences, faults and

so on ) are not taken into account
• The safety margin , that is the difference between the foreseen un-
availability and the target , must be carefully considered
• SDH urban systems unavailability: a 0.01% limit ( ≈ 53 min/year),

hop length independent, should be applied. Different limits can be even-
tually adopted in agreement with the customer.
• In case of urban networks custom calculations MUST be suggested to

customers and must be considered as a default
• Adopting custom targets try to fix the limits for each link and NOT for
the entire connection.
5.6. Some considerations about path topologies

For the calculation of the end-to-end performance of a path from the performances
of path elements (PEs), 2 basic topologies (linear and redundant) are distinguished
:

IB IB
Country 1 Country 2 Country 3
PEP TIC FS FS FS FS TIC PEP
NPE1 IPCE1 ICPCE1 IPCE2 ICPCE2 IPCE3 NPE2
PEP Path End Point ICPCE Inter-Country Path Core Element

TIC Terminal International Centre IPCE International Path Core Element
FS Frontier Station NPE National Path Element
IB International Border
Fig. 86 – Linear topology network

Country 1 Country 2 Country 3
TIC FS FS FS FS TIC
NPCE1 IPCE1 ICPCE1 IPCE2 ICPCE2 IPCE3 NPCE2

PEP PEP
PSE PSE
PAE1 PAE2
NPCE1 TIC FS FS FS FS TIC NPCE2
IPCE1 ICPCE1 IPCE2 ICPCE2 IPCE3
PEP Path End Point ICPCE Inter-Country Path Core Element

TIC Terminal International Centre IPCE International Path Core Element
FS Frontier Station NPCE National Path Core Element
IB International Border PAE Path Access Element
PSE Path Switching Element
Fig. 87 – Redundant topology network
For the purposes of end-to-end calculations, it is more convenient to use the un-
availability ratio than the availability ratio and the outage intensity than the mean
time between service outage.
5.6.1. End-to-end unavailability ratio

Definition:
Ö Ur unavailability ratio of a PE
Ö UR unavailability ratio of an end-to-end path
(In the previous edition of the Rec. ur and UR were defined for both the mean and
worst case.)
It will be assumed that the unavailability performance of the constituent PEs follows
a normal or Gaussian distribution and that it is small so that the contribution of
each PE can, therefore, be considered to be independent, i.e., no overlap in the du-
ration of the periods of unavailability.
UR values are calculated for each topology.
5.6.1.1. Linear Topology
If a path is made of N path elements used in series, as indicated in the figure, then

the following approximations can be used for small values of unavailability ratios:
UR=∑ (uri ) (169)

i
5.6.1.2. Redundant Topology
In a redundant configuration using two parallel paths and a protection switch at one
end (for each direction of transmission), the availability of the protected end-to-end
path in the figure is:
UR end-to-end = ( UR 1×UR 2 ) +UR A +UR B (170)
where UR1, UR2 are the unavailability ratios of each parallel path, URA and URB are
the unavailability ratios of protection switches A and B (in the previous edition of
the Rec. the unavailability of only one protection switch was considered).
Given that the two connections are independent of each other, this would result in a
large improvement in the availability of the end-to-end path because typical values
of the unavailability should became negligible.
5.6.2. End-to-end outage intensity

Definition:
Ö oi: outage intensity of a PE
Ö OI: outage intensity of an end-to-end path
OI values are calculated for each topology.
5.6.2.1. Linear Topology
If a path is made of N PEs used in series, as indicated in the figure, then the follow-
ing formulae can be used, assuming that the outage intensity in each PE is small
and the outages are, therefore, independent and do not overlap:
OI = ∑ ( oi j ) (171)
j
5.6.2.2. Redundant Topology
The outage intensity of the end-to-end protected path in the figure is:
OIend-to-end = ( OI1 × UR 2 ) + ( OI2 × UR1 ) + OIA + OIB (172)

Where OI1 and OI2 are the outage intensities of each parallel path and OIA and OIB
are the outage intensity of the protection switches A and B (in the previous edition
of the Rec. the outage intensity of only one protection switch was considered).

5.7. Reliability
5.7.1. What is reliability

ITU performance are evaluated in terms of quality and unavailability according to
the previous statements but a different availability criterion is sometimes adopted
in terms of annual percentage (99.xxx %) : it consists of a weighted sum of ITU
quality and unavailability .
This quantity, deriving from Bellcore standard , will be defined as reliability in order
to avoid confusion with the previous ones.
5.7.2. How reliability is evaluated

Reliability is evaluated according to the following procedure:
1) Mono-directional ITU annual unavailability is evaluated
2) Equipments failure annual unavailability ( as the sum of contribution of each
terminal ) is considered
3) The contributions 1) and 2) are summed ( yearly based )
4) Bi-directional link quality is evaluated ( single side quality is doubled )
5) The result of point 4) is multiplied for the number of fading months that de-
pends on average temperature according to the relationship:
⎛ 9 t°+ 32 ⎞
⎜ ⎟
months = 0.253 ⋅ ⎝ ⎠ ⋅ 12
5
(173)
50
( t° is the temperature in Celsius )
6) The results of points 3) and 5 are summed obtaining the annual value
7) The result of point 6) is complemented to 1

6. References
Ö ITU-T Rec. E.801, Framework for service quality agreement.

Ö ITU-T Rec. G.705, Characteristics of PDH equipment functional blocks.
Ö ITU-T Rec. G.707 (1996), Network node interface for SDH.
Ö ITU-T Rec. G.709, Interface for the optical transport network (OTN).
Ö ITU-T Rec. G.783 (1997), Characteristics of synchronous digital hierarchy (SDH)
equipment functional blocks.
Ö ITU-T Rec. G.784 (1999), Synchronous digital hierarchy (SDH) management.
Ö ITU-T Rec. G.803 (2000), Architecture of transport networks based on the syn-
chronous digital hierarchy (SDH).
Ö ITU-T Rec. G.821 (1996), Error performance of an international digital connec-
tion operating at a bit rate below the primary rate and forming part of an inte-
grated services digital network.
Ö ITU-T Rec. G.821 (2002), Error performance of an international digital connec-
tion operat-ing at a bit rate below the primary rate and forming part of an Inte-
grated Services Digital Network.
Ö ITU-T Rec. G.826 (1999), Error performance parameters and objectives for in-
ternational, constant bit rate digital paths at or above the primary rate.
ternational constant bit rate digital paths.
Ö ITU-T Rec. G.827 (2000), Availability parameters and objectives for path ele-
ments of international constant bit rate digital paths at or above the primary
rate.
Ö ITU-T Rec. G.827 (2003), Availability performance parameters and objectives for
end–to-end international constant bit rate digital paths
ternational CBR synchronous digital paths.
Ö ITU-T Rec. G.841, Types and characteristics of SDH network protection architec-
tures.
Ö ITU-T Rec. G.842, Interworking of SDH network protection architectures.
Ö ITU-T Rec. I.355, ISDN 64 kbit/s connection type availability performance.
Ö ITU-T Rec. I.356 (2000), B-ISDN ATM layer cell transfer performance.
Ö ITU-T Rec. I.357, B-ISDN semi-permanent connection availability.

Ö ITU-T Rec. X.137, Availability performance values for public data networks when
providing international packet switched services.
Ö CCITT Rec. I.321 (1991), B-ISDN Protocol reference model and its application.
Ö ITU-T Rec. M.2101 (2000), Performance limits for bringing-into-service and
maintenance of international SDH paths and multiplex sections.
Ö ITU-R Rec. F.1492 (2000), Availability objectives for digital radio-relay links
forming part of the international portion of a CBR path at or above the primary
rate.
Ö ITU-R Rec. F.1493 (2000), Availability objectives for digital radio-relay links
forming part of the national portion of a CBR digital path at or above the prima-
ry rate.
ANNEX 1
This annex is based on the procedure received from F. Vigneron ( 19/03/2001 )
7. US Path Availability Calculation Method
The total path outage time for digital radios is:

TTOTAL = TTHERMAL + TDIGITAL + TRAIN + TUPFADE
Where:
TTOTAL = Total 1-way path outage in seconds, assuming adequate path clear-
ance.
TTHERMAL = Thermal multipath outage. This is the outage caused by multipath
fading below the radio thermal noise threshold (typically for a 10-6 BER). It also
includes the effects of co-channel interference from other microwave systems and
adjacent channel interference from other radio channels on the same microwave
path.

TDIGITAL = Dispersive multipath outage. This is the outage caused by

in-band distortion of the digital signal.
TRAIN = Rain outage.
TUPFADE = Outage caused by up fading above the maximum received
signal level.
To calculate 2-way outage, each factor except rain is calculated in both directions
and added. Since rain outage occurs simultaneously in both directions of transmis-
sion, the rain outage is the same for a 1-way or 2-way calculation for equal fade
margins. If the fade margins are different in the 2 directions of transmission, the
rain outage is calculated for the smaller fade margin.
To calculate availability:
A = 100 x (1 - TTOTAL / TREF)
Where:
A = Availability in percent
TREF = 3.15 x 107 seconds for yearly availability
= 2.63 x 106 seconds for monthly availability
7.1.1. Multipath Outage

To calculate the outage time due to multipath fading:
TTHERMAL = (R To x 10-FFM/10) / ITHERMAL
TDIGITAL =(R To x 10-DFM/10) / IDIGITAL
Where:
R = Fade occurrence factor
To = Number of seconds in the heavy fading season
= 8 x 106 x (t/50) for yearly availability
6
= 2.63 x 10 for monthly availability
t = Average annual temperature in degrees Fahrenheit (used to ad-
just the length of the fading season from region to region)
FFM = Flat fade margin in dB (composite fade margin not including
the dispersive fade margin)
DFM = Dispersive fade margin in dB (from radio specifications)
ITHERMAL = Improvement factor for thermal outages from frequency and
space diversity
IDIGITAL = Improvement factor for dispersive outages from frequency and
space diversity.

The fade occurrence factor is calculated using the basic outage equation for atmos-
pheric multipath fading:
R = c (50/w)1.3 (f/4) D3 x 10-5

Where:
c = Climate and humidity factor
• 0.5 for dry climate
• 1 for average conditions
• 2 for coastal and over water paths
w = Terrain roughness factor in feet (20 < w < 140)
f = Frequency in GHz
D = Path length in miles
The flat fade margin (FFM) is derived from the power addition of up to three indivi-
dually contributing fade margins, using the following equation:
FFM = -10 LOG10 [ 10-TFM/10 + 10-EIFM/10 + 10-AIFM/10 ]
Where:
TFM = Thermal fade margin in dB
EIFM = External interference fade margin for all exposures in dB, due to
interference from transmitters on other microwave paths = EIFM - 10 LOG10 N
EIFM = External interference fade margin for 1 exposure in dB
N = Number of interference exposures
AIFM = Adjacent channel interference fade margin in dB, due to interfe-
rence from other transmitters on this microwave path
The Thermal Fade Margin (TFM) is determined as follows:
TFM = G - L - Th
Where:
G = Antenna gains + Radio transmit power
L = Free space loss + Feeder losses + Circulator losses + Radome losses
+ Attenuators + Field Margin + Atmospheric Absorption Loss
Th = Radio thermal noise threshold (typically for a 10-6 bit error rate)
The External Interference Fade Margin (EIFM) and Adjacent Channel Interference
Fade Margin (AIFM) are determined as follows:
EIFM =C/I1 - C/I2 AIFM =C/I1 - C/I2
C/I1 =Carrier-to-Interference ratio of an interfering signal, in dB
C/I2 =Carrier-to-Interference ratio where the radio runs continuous errors
at a specified bit error rate, in dB (from radio specifications, typically for a 10-6

BER, same BER as the thermal noise threshold defined above)

To calculate free space loss in dB:
FSL = 96.6 + 20 LOG10 (f* D)
Where:
f = Frequency in Ghz
Atmospheric absorption loss is calculated as follows:
⎡ 6.09 4.81 ⎤
L 0 = ⎢7.19 * 10−3 + 2 + ⎥ * f 2 * 10−3
⎢⎣ f + 0.227 ( f − 57 )2 + 1.5 ⎥
⎦
⎡ 3.6 10.6 ⎤
L w = ⎢0.05 + 0.0021 * p + + ⎥ * f 2 * 10−4 (174)
( f − 22.2 ) + 8.5 ( f.183.3) + 9 ⎥⎦
2 2
⎢⎣
A TOTAL = (L 0 + L w ) * D * 1.609
Where:
f = Frequency in GHz (f < 57 GHz)
p = Absolute humidity of the region in g/m
Lo = Oxygen attenuation in dB/kilometer
Lw = Water vapor attenuation in dB/kilometer
ATOTAL = Total atmospheric absorption in dB for the path
For more accurate computation of atmospheric absorption, temperature corrections
of -1.0% per degree Centigrade for oxygen and -0.6% per degree Centigrade for
water vapor can be applied, for deviations from 15° C.
The improvement due to space diversity is calculated using the following equa-
tions:
ISD-THERMAL = 7 x 10-5 s2 (f/D) x 10FFM/10
ISD-DIGITAL = 0.09 (f/D) x 10DFM/10
Where:
ISD-THERMAL = Space diversity improvement factor for thermal outages
(1 < ISD-THERMAL < 200)
ISD-DIGITAL = Space diversity improvement factor for dispersive outages
(1 < ISD-DIGITAL < 200)
s = Vertical separation of receiving antennas in feet, center-to-center


FFM = Flat fade margin in dB
DFM = Dispersive fade margin in dB
When the two diversity antennas have different flat fade margins (FFM), the multi-
path outage time TTHERMAL is calculated using the higher of the two fade margins,
and the space diversity improvement factor ISD-THERMAL is calculated using the
lower of the two fade margins.
Angle diversity improvement may be calculated using the above equations for
space diversity, assuming a 30 foot equivalent space diversity spacing. The angle
diversity improvement factor for thermal fading is 1/10 of the calculated space di-
versity improvement factor ISD-THERMAL, up to a maximum of 20. The angle diversi-
ty improvement factor for dispersive fading is the same as the space diversity fac-
tor
ISD-DIGITAL, up to a maximum of 200.
The improvement factors for a 4-receiver system with space diversity and frequen-
cy diversity (i.e., quad diversity) are:
ITHERMAL = ISD-THERMAL x IFD-THERMAL
IDIGITAL = ISD-DIGITAL x IFD-DIGITAL
The improvement due to frequency diversity is calculated from the following equa-
tions:
IFD-THERMAL = 100 (f/DG) x 10FFM/10
IFD-DIGITAL = (468 / f2 D) x 10DFM/15 for 1+1 systems
IFD-DIGITAL = 9324 (f/DG) x 10DFM/15 for 1+N systems, N > 1
Where:
IFD-THERMAL= Frequency diversity improvement factor for thermal outages
(1 < IFD-THERMAL < 200)
IFD-DIGITAL = Frequency diversity improvement factor for dispersive outages
(1 < IFD-DIGITAL < 200)
G = Empirical factor, depending on the number of working channels,

transmitting frequencies, and flat fade margins
G = 0.25 (f1 + f2)3 / (f2 - f1) for a 1+1 frequency diversity system with
equal flat fade margins
f2 = Higher frequency in GHz
f1 = Lower frequency in GHz
FFM = Flat fade margin in dB

DFM = Dispersive fade margin in dB
The improvement factors for a 2-receiver system with space diversity and frequen-
cy diversity (i.e., hybrid space diversity, using a different frequency on each anten-
na) are:
ITHERMAL = maximum (ISD-THERMAL , IFD-THERMAL )
IDIGITAL = maximum (ISD-DIGITAL , IFD-DIGITAL )
Rain Outage - Crane Path Attenuation Model

Rain attenuation is calculated from the following equations:
AR =aRB[(euBD - 1)/uB ] for: 0 < D < d kilometers
AR =aRB[(euBd - 1)/uB - (bB ecBd)/cB + (bB ecBD)/cB] for: d < D < 22.5 km
Where:
u = LN [ b ecd ] / d
b = 2.3 R-0.17
c = 0.026 - 0.03 LN ( R )
d = 3.8 - 0.6 LN ( R )
AR = Path attenuation due to rain in dB
AR = TFM - WRL
TFM = Thermal fade margin in dB
WRL = Wet radome loss in dB
R = Point rain rate in mm/hr
D = Path distance in kilometers = DMILES x 1.609

DMILES = Path distance in miles
The coefficients "a" and "B" are a function of frequency and polarization and are
listed in Table I of ITU-R Recommendation P.838.
To calculate rain outage, an iterative computer program is used to vary the rain
rate R until the calculated path attenuation is equal to the rain attenuation margin,
AR = TFM-WRL. This rain rate is then applied to a distribution of rain rate vs. time
for a particular geographic area to find the rain outage time.

For paths longer than 22.5 kilometers, the attenuation AR is calculated for a 22.5
kilometer path and the resulting rain outage time is multiplied by a factor of
(D/22.5).
If a radome is not used, the rain outage may be substantially greater due to rainfall
collecting in the antenna feed.
Rain Outage – North American Rain Rate Data
To find the total time a rain rate R occurs at a specific geographic location, meas-
ured curves of rain rate vs. time are used for the following locations.
Continental U.S., Hawaii, and Puerto Rico
Rain curves are available for 212 locations in the continental U.S., Hawaii, and
Puerto Rico. Measured data is from the U.S. Climatic Center (Reference 17). The
curves are based on 20 years of data in the eastern U.S. and 50 years of data in
the western U.S.
Canada
Rain curves are available for 47 locations in Canada. Measured data is from CRC
Report No. 1329-E, Canadian Department of Communications (Reference 19).
Upfade Outage
Outage due to up-fading above the maximum received signal level is calculated
from the following equation:
⎡ ⎛ UFM ⎞ ⎤
TUPFADE = 0.3RT0 ⎢0.5 * erfc ⎜ ⎟⎥ for UFM< UMAX
⎣ ⎝ 3.5 2 ⎠ ⎦ (175)
TUPFADE = 0 for UFM > UMAX
Where:
TUPFADE = Up fade outage time in seconds
R = Fade occurrence factor for multipath outage

To = 8 x 106 x (t/50) for yearly availability
= 2.63 x 106 for monthly availability
t = Average annual temperature in degrees Fahrenheit
erfc()=complementary error function used to calculate the probability of up fades

> UFM.
UFM= Up fade margin in dB = RSLMAX - RSL - FM
RSLMAX = Maximum received signal level in dBm (typically for a 10-6 BER)
RSL = Calculated received signal level in dBm
FM = Field margin in dB (typically 1 dB)
UMAX = Maximum up fade in dB for this path
= 10 LOG10 d - 0.03 d , UMAX > 6 dB
d = Path length in kilometers
d = 1.609 x D
This calculation assumes that up fades are log normally distributed with a standard
deviation of 3.5 dB and that 9.5 dB up fades occur at the same rate as 30 dB fades.
For systems with two or more receivers, the up fade margin is calculated for each
receiver and the highest up fade margin is used.

Annex 2
Geopotential Height.
The altitude of the radiosonde is reported in units of scaled geopotential (geopoten-
tial height) above mean sea level (MSL). The elevation of the release point should
be a surveyed geometric position, and shall be converted to geopotential height.
The strata or layers (in pressure) resulting from the radiosonde telemetry rate shall
be employed to calculate the geopotential height of the instrument. The time in-
crements are usually the points sampled, but the values of the variables should be
the result of the signal processing algorithm in order to achieve a balance between
vertical resolution and the quality of the variables. Layer thicknesses shall be calcu-
lated from the measured pressures at the bounds of each layer and the mean vir-
tual temperature within the layer.
The definition of geopotential is the potential energy due to gravity of a unit mass
of air at some point above a standard position (i.e. zero energy), usually means
sea-level, and is measured in a positive sense vertically upward. Geopotential is:
Z
Φ = ∫ gdZ (176)
0
The geopotential meter is defined as a rescaled geopotential, and is given by:

n
1
9.80665 ∫0
hn = gdZ (177)
Thus, a discrete version of the hypsometric equation is

pi +1 gδ h
ln =− (178)
pi RTv
where
p = pressure in hPa
g = acceleration of gravity = 9.80665 according to WMO Technical Regulations
R = gas constant for dry air = 287.04m2s-2K-1
Tv = mean virtual temperature for the layer = [Tv(i+1) + Tv(i)]/2 in degrees K
dh = layer thickness (meters)i refers to lower boundary.
For simplicity, the equation above can also be expressed as:
⎛p ⎞
Δhi = −14.6355 (Ti −1 + Ti ) *ln ⎜ i −1 ⎟ (179)
⎝ pi ⎠
Finally, the thickness Δhi of each pressure layer above the surface in geopotential
meters may be summed to give the height hn of any pressure surface. Thus,

n
hn = ∑ Δhi (180)
i =0
Occasionally it is necessary to convert from geopotential to geometric height. The

conversion algorithm is:
hn Re
zn = (181)
Gr − hn
where
Gr = the gravity ratio = [gfRe/9.80665]
Re = radius of the earth at latitude f , and
gf = gravity at f = 9.80616[1-0.002637(cos(2f )) + 0.0000059(cos2(2f ))]

1
Rec. ITU-R P.453
2
Rec. ITU-R P.1510

Network Design Fundamentals

Загружено:

Сведения о документе

Авторское право

Доступные форматы

Поделиться этим документом

Поделиться или встроить документ

Параметры публикации

Этот документ был вам полезен?

Это неприемлемый материал?

Авторское право:

Доступные форматы

Network Design Fundamentals

Загружено:

Авторское право:

Доступные форматы

Network Design Fundamentals 1.

Network design fundamentals

dalmazio.mandich@alcatel-lucent.it -1- All rights reserved © 2008

2. DIGITAL RADIO LINKS DIMENSIONING CRITERIA .............................................................- 16 -

3. RAIN ATTENUATION ......................................................................................................- 95 -

4. QUALITY .................................................................................................................... - 103 -

dalmazio.mandich@alcatel-lucent.it -2- All rights reserved © 2008

5. UNAVAILABILITY ......................................................................................................... - 133 -

6. REFERENCES ............................................................................................................. - 156 -

7. US PATH AVAILABILITY CALCULATION METHOD .............................................................. - 157 -

dalmazio.mandich@alcatel-lucent.it -3- All rights reserved © 2008

dalmazio.mandich@alcatel-lucent.it -4- All rights reserved © 2008

dalmazio.mandich@alcatel-lucent.it -5- All rights reserved © 2008

Fig. 69 - ITU-T G. 826/828 apportionment criteria .................................................... - 116 -

dalmazio.mandich@alcatel-lucent.it -6- All rights reserved © 2008

1.1. ORTHODROMIC DISTANCE BETWEEN TWO POINTS

1.2. Haversine formula

The Haversine formula remains particularly well-conditioned for numerical computa-

travelling perpendicular to that is

dalmazio.mandich@alcatel-lucent.it -7- All rights reserved © 2008

Under this assumption the antenna bearings are

dalmazio.mandich@alcatel-lucent.it -8- All rights reserved © 2008

Although this formula is mathematically exact, it is unreliable for small distances

1.2.1.1. The radius of the Earth, R

dalmazio.mandich@alcatel-lucent.it -9- All rights reserved © 2008

1.3. Coordinates conversion

1.3.1. Coordinate systems

1.3.2. Reference Ellipsoids

dalmazio.mandich@alcatel-lucent.it - 10 - All rights reserved © 2008

Datum Ellipsoid Region of use

Adindan Clarke 1880 Burkina Faso; Cameroon; Ethiopia; Mali;

Afgooye Krassovsky 1940 Somalia

dalmazio.mandich@alcatel-lucent.it - 11 - All rights reserved © 2008

Astro DOS 71/4 St Helena Island

dalmazio.mandich@alcatel-lucent.it - 12 - All rights reserved © 2008

MEAN FOR Austria; Finland; Netherlands;

Fort Thomas 1955 Clarke 1880 Nevis; St. Kitts

dalmazio.mandich@alcatel-lucent.it - 13 - All rights reserved © 2008

Massawa Bessel 1841 Ethiopia (Eritrea)

dalmazio.mandich@alcatel-lucent.it - 14 - All rights reserved © 2008

Qornoq Greenland (South)

dalmazio.mandich@alcatel-lucent.it - 15 - All rights reserved © 2008

2. Digital radio links dimensioning criteria

Performance evaluation of digital microwave radio links must be imple-

As far as the propagation path characteristics are concerned, rain atten-

DRRS should be mainly designed in terms of unavailability at frequen-

All kind of quality and unavailability objectives MUST be referenced by

dalmazio.mandich@alcatel-lucent.it - 16 - All rights reserved © 2008

Fig. 1 – Parameters affecting quality and availability evaluations

ANALOG DIGITAL ANALOG DIGITAL

THERMAL THERMAL INTERMODULATION INTERSYMBOL

Fig. 2 – Flat and selective fading

dalmazio.mandich@alcatel-lucent.it - 17 - All rights reserved © 2008

Fig. 3 – Received spectrum affected by selective fading

Selective fading outage = limMargin→∞ (total fading outage)

dalmazio.mandich@alcatel-lucent.it - 18 - All rights reserved © 2008

2.2. The radio refractive index and its gradient

2.2.2. Radio refractivity1

where Nd is the dry term, Nw the wet term and

Temperature and humidity are the predominant factors for changes in

dalmazio.mandich@alcatel-lucent.it - 19 - All rights reserved © 2008

dalmazio.mandich@alcatel-lucent.it - 20 - All rights reserved © 2008

Fig. 5 – Surface water vapor density ( winter )

Fig. 6 - Surface water vapor density ( spring )