G Model

AEUE-51440; No. of Pages 5 ARTICLE IN PRESS

Int. J. Electron. Commun. (AEÜ) xxx (2015) xxx–xxx

Contents lists available at ScienceDirect

International Journal of Electronics and

Communications (AEÜ)
journal homepage: www.elsevier.com/locate/aeue

REGULAR PAPER

Performance of a subcarrier intensity modulated differential

phase-shift keying over generalized turbulence channel
Jagadeesh Vellakudiyan a,b,∗ , Palanivel Muthuchidambaranathan a ,
Francis Minhthang Bui b , Vineeth Palliyembil a
a
Department of Electronics and Communication Engineering, National Institute of Technology, Tiruchirappalli, India
b
Department of Electrical and Computer Engineering, University of Saskatchewan, Saskatoon, Canada

a r t i c l e i n f o a b s t r a c t

Article history: Optical wireless communication technologies are finding a greater interest and wider attention within
Received 23 January 2015 the research community of late. In this paper, we investigate the performance of a free space optical
Accepted 25 June 2015 communication system over a generic propagation model called M-distributed channel in the presence
of atmospheric turbulence. We analyzed a Subcarrier intensity-modulated free-space optical (SIM-FSO)
MSC[2010]: communication system using DPSK and closed form expressions are derived using Meijer G function for
00-01
bit error rate, channel capacity and outage probability for M-distribution.
99-00
© 2015 Elsevier GmbH. All rights reserved.
Keywords:
Free space optics
M-distribution
Bit error Analysis
Channel capacity
Outage probability

1. Introduction poor transmission rate. Hence pulse position modulation (PPM)

Recently, optical wireless communication (OWC) has gained
elevated research focus as an alternative to the conventional
wireless communication technologies because of the potential
improvement in delivering high speed data communication over
short distance. The key attraction of OWC system is their high
bandwidth, easy deployment, cost effectiveness and free from
license regulations [1]. OWC technology is found to be suitable
for both indoor as well as outdoor applications. The OWC used
for outdoor is commonly termed free space optics (FSO). Ran-
dom intensity fluctuation induced by atmospheric turbulence often
degrades the performance of FSO communication system. On off
keying (OOK) based intensity modulation/direct detection (IM/DD)
is widely employed in FSO system because of its simplicity [2].
However, in order to achieve optimal performance, OOK needs
adaptive detection threshold [3] to vary in accordance with the
underlying irradiance fluctuation and noise. Also OOK provides
∗ Corresponding author at: Department of Electronics and Communication Engi-
neering, National Institute of Technology, Tiruchirappalli, India.
Tel.: +91 8903983048.
E-mail address: vkjagadeesh@gmail.com (J. Vellakudiyan).
[4] has been projected as an alternative of OOK modulation as it
does not require any adaptive thresholds. There onwards, several
statistical models have been proposed to model the behavior of
atmospheric turbulence with different modulation schemes [5–11].
Log-normal [12], K [13] and the gamma–gamma (GG) distribution
[13,14] are the few successful approaches. The log-normal distri-
bution is valid only in weak turbulence and K for strong turbulence
condition. The gamma–gamma distribution is suitable to model
moderate to strong turbulence channels.
A new propagation channel model termed M (Malaga) distri-
bution [15] has been adopted recently which unifies most of the
previously proposed statistical models and is valid for a wide range
of turbulent conditions. Performance analysis of FSO communica-
tion systems with OOK modulation over M-distributed turbulence
channel [16,17] is already analyzed. In [18] error rate performance
of binary phase shift keying based subcarrier intensity modulation
(BPSK-SIM) is studied using Meijer G function. The performance of
a FSO communication system with differential phase-shift keying
(DPSK) for different atmospheric turbulence channels has already
been analyzed [19,20]. A DPSK-SIM FSO link is considered over
negative exponential atmospheric turbulence environment and
analyzed the performance in [21]. In [22], the bit error rate (BER)
and outage probability of DPSK based FSO communication system

http://dx.doi.org/10.1016/j.aeue.2015.06.015
1434-8411/© 2015 Elsevier GmbH. All rights reserved.

Please cite this article in press as: Vellakudiyan J, et al. Performance of a subcarrier intensity modulated differential phase-shift keying
over generalized turbulence channel. Int J Electron Commun (AEÜ) (2015), http://dx.doi.org/10.1016/j.aeue.2015.06.015
G Model
AEUE-51440; No. of Pages 5 ARTICLE IN PRESS
2 J. Vellakudiyan et al. / Int. J. Electron. Commun. (AEÜ) xxx (2015) xxx–xxx

over a M-distributed turbulent channel has been reported. The Here ˛ is a positive parameter related to the effective number of
author used series expansion to express the probability density large-scale cells of the scattering process and the parameter ˇ rep-
function (PDF). resents the amount of fading and are given by

−1
0.49R2
2. Paper contributions and organization ˛≈ exp −1 (4)
12/15 7/6
(1 + 0.18d2 + 0.56R )
In this paper, we generalize the results obtained in [21] for all
turbulence regimes and derive a novel and general closed-form 
−1
12/15 −5/6
analytical expression for the bit error rate, channel capacity and 0.51R2 (1 + 0.69R )
ˇ≈ exp −1 (5)
outage probability of the system. Since a generic distribution is 12/15 5/6
(1 + 0.9d2 + 0.62d2 R )
used to model the turbulence-induced fading, the derived expres-
sions can be used for analyzing the performance of a DPSK-SIM

kD2
based FSO system over most of the statistical models proposed so where d = 4L
, k = 2/ is the optical wave number,  is
far, including gamma–gamma, K, and negative exponential distri- the wavelength, L is the length of optical link and D is the
butions. To the best of our knowledge, this is the first work that receiver aperture diameter. The Rytov variance R2 is given as
investigate the performance of a DPSK-SIM based FSO system. The R2 = 0.5Cn2 k7/6 L11/6 , where Cn2 is the refractive index structure
remainder of the paper is organized as follows. In Section 2, the M- parameter. Atmospheric turbulence is well characterized by Rytov
distributed turbulence channel model is discussed. In Section 3, the variance. In [23], the authors have used some experimental data of
performance of a DPSK based SIM-FSO communication system over Cn2 provided by University of Waseda, Japan to find the typical value
turbulence channel is presented. Section 4 presents corresponding of R2 . At night (1 a.m.), Cn2 parameter registered its minimum value
numerical results. Finally, paper is concluded in Section 5. as 7 × 10−15 m2/3 . At sunrise (6.45 a.m.), the value is 1.2 × 10−14 m2/3
and at midday it is 2.8 × 10−14 m2/3 . The corresponding R2 are 0.32,
0.52 and 1.2, respectively for a length L = 1 km,  = 785 nm and
3. Channel model
D = 100 mm.
The average power from the coherent contributions is given as
The M-turbulence model is based on a physical model that
includes three terms as shown in Fig. 1 [16] (reproduced with the  =  + 2b0 + 2 2b0  cos( A − B ). For simplicity, we have
permission from authors): the first term is a line of sight (LOS) com- denoted g = E[|USG |2 ] = 2b0 (1 − ). In Eq. (1), Kv (.) represents the
ponent represented by UL ; the second term USC is a component that modified Bessel function of the second kind and order v. The exist-
is scattered by the eddies on the propagation axis and coupled to ing atmospheric distribution models for optical communications
the LOS contribution and the third term USG is the component due can be generated from the M-distribution model and is presented
to scattering by off-axis eddies. The average power of the LOS term in [16].
and the average power of the total scatter components are repre- The generalized form of Eq. (1) with ˇ being a real number is
sented by  = E[|UL |2 ] and 2b0 = E[|USC |2 + |USG |2 ], respectively. In given in [24] as
addition, the average power of individual scatter term is given by

∞   
E[|USC |2 ] = 2b0 and E[|USG |2 ] = (1 − )2b0 , for the coupled-to-LOS (G) ˛+k ˛I
fI (I) = A (G)
ak I 2 −1 K˛−k 2 (6)
scattering term and the classic scattering component received by g
k=1
off-axis eddies, respectively, where the parameter  represents the
amount of scattering power coupled to the LOS component and where
ranges from 0 to 1. ˛  ˇ+ ˛2
As derived in [16] the M-pdf of the irradiance I is given by 2˛ 2 g ˇ
A(G) = (7)
1+ ˛ (˛) g ˇ + 
  
2
g

ˇ
˛+k ˛ˇI
−1
fI (I) = A ak I 2 K˛−k 2 (1) k k−1
g ˇ +  (G) (ˇ)k−1 (˛g ) 2 ( )
k=1 ak = (8)
2 k−1
[(k − 1)!] gk−1 ( + g ˇ)
where
(ˇ)k represents the Pochhammer symbol.
˛  ˇ+ ˛2
2˛ 2 g ˇ
A= (2)
1+ ˛
2 g ˇ +  4. Performance analysis of DPSK based SIM-FSO
g (˛)

1− 2k
 k−1   k In DPSK, the information is encoded as phase differences
(g ˇ +  )  ˛ 2 between successive signal transmissions. In binary DPSK, corre-
ak = (3)
(k − 1)! g ˇ sponding to the information bits 0 and 1, the relative phase shifts
are 0◦ and 180◦ respectively. This indicates that the information bit
1 is transmitted by shifting the carrier phase by 180◦ relative to
the phase in the previous signaling interval, while the information
bit 0 by a zero phase shift relative to the previous carrier phase. At
the receiver side, demodulation is performed based on the phase
difference between the received signals in two consecutive inter-
vals. DPSK is an attractive alternative to ordinary PSK modulation.
It avoids the need for complex carrier recovery schemes to provide
an accurate phase estimate. Compared to PPM, SIM is bandwidth
efficient and does not require an adaptive decision threshold simi-
lar to OOK [3]. The performance of the DPSK based SIM in analyzed
Fig. 1. M-distributed turbulence channels [16]. with the help of BER, channel capacity and outage probability.

Please cite this article in press as: Vellakudiyan J, et al. Performance of a subcarrier intensity modulated differential phase-shift keying
over generalized turbulence channel. Int J Electron Commun (AEÜ) (2015), http://dx.doi.org/10.1016/j.aeue.2015.06.015
G Model
AEUE-51440; No. of Pages 5 ARTICLE IN PRESS
J. Vellakudiyan et al. / Int. J. Electron. Commun. (AEÜ) xxx (2015) xxx–xxx 3

The atmospheric channel is a slow fading channel because the channel without any error. The channel capacity is expressed by
atmospheric turbulence is slow varying compared with the symbol the expression
duration of the transmitted signal. So outage probability analysis
C(I) = Blog2 (1 +
◦ I 2 ) (15)
best describes the performance of the system. But here we carry
out the average bit error rate and ergodic channel capacity analysis Upon normalizing this expression with respect to the channel
as well even though these parameters are slow varying. bandwidth B, Eq. (15) becomes [27],
C(I) = log2 (1 +
◦ I 2 ) (16)
4.1. Average bit error rate
This value fluctuates with respect to the change of channel irra-
For a coherent DPSK demodulator, the conditional BER can be diance I, so we take the expected value of it which is called
derived as [20], ergodic capacity, Cerg = E[C(I)]. For a DPSK based SIM-FSO system,
  the ergodic capacity is given by
(RA0 I)2   
Pe|I (I) = 0.5 exp − (9) ∞
(RA0 I)2
4n2 Cerg = log2 1+ fI (I)dI (17)
0 2n2
where R is the photo-detector responsivity, is the variance of n2
the channel noise n(t). In most of the FSO systems, the received Substitute (1) in (17) and re arrange the equation as,
signal is limited by shot noise caused by background radiation from 
ˇ  ∞ ˛+k
−1
various optical sources and thermal noise due to electronic devices Cerg = A ak I 2
following the photodetector, which can be modeled as an additive k=1
0

(18)
white Gaussian noise (AWGN) and A0 is the area of photo detector.   
(RA0 I)
2
˛ˇI  −
The probability of bit error for DPSK-based SIM-FSO over an M- ×log2 1+
2n2
2,0
G0,2
g ˇ + 
 ˛−k ˛−k
dI
,−
distributed turbulence channel can be derived using 2 2
 ∞ 
ˇ  ∞    
2
A ˛+k
−1 (RA0 I) ˛ˇI
Pe = Pe|I (I)fI (I)dI (10) = ak I 2 ln 1+ K˛−k 2 dI (19)
ln(2) 2n2 g ˇ + 
0 k=1
0

By substituting Eqs. (1) and (9) in (10), we derive the closed form The natural logarithmic function can be expressed using
expression for the average BER of DPSK-SIM modulation in M-  G function
Meijer  [[26], Eq. 01.04.26.0003.01] as ln(1 + x) =
1,2
distributed turbulence channel as G2,2 x | 1,1
1,0
. So (19) can be written as

ˇ  ∞    ˛+k −1  
A (RA0 I)
2
2 ˛ˇI

ˇ  ∞ ˛+k
Pe = ak exp − I K˛−k 2 dI (11) A −1
2 4n2 g ˇ +  Cerg = ak I 2
0
k=1 ln(2)
k=1
0

The exp(·) and the modified Bessel functions can be expressed    

(RA0 I)
2
 1, 1 ˛ˇI  −
in terms of Meijer G functions by employing  [[25],− Eq. (11)] and
1,2
×G2,2
2n2
 1, 0
2,0
G0,2
g ˇ + 
 ˛−k ˛−k
dI
1,0
[[26], Eq. (03.04.26.0009.01)] as e−x = G0,1 x | 0 and Kv (x) = ,−
   2 2
1 2,0 x2  − (20)
G
2 0,2 4  v ,− v , respectively. Then
2 2

  ˛+k       
A  − 
ˇ ∞
R2 A20 h2 ˛ˇI −
Pe = ak I 2
−1 1,0
× G0,1  2,0 
4 4n2  0 G0,2 g ˇ +   ˛−k
, − ˛−k
dI (12)
0 2 2
k=1

Solving the integral of the product of two Meijer G functions

[25], and the closed form expression for the average BER (ABER) of By integrating the Meijer G functions using Eq. (21) [25], the Cerg
DPSK-SIM modulation is derived as can be solved in a closed form expression as,
    8R2 A2 B2  
˛

ˇ
4R2 A20 B2  1−˛ 2−˛ 1−k 2−k 
ˇ

Pe =
A2˛−1 B 2 k k 1,4  2
, 2 , 2 , 2 A2˛−1 B 2
˛
k  1, 1, 1−˛ 2−˛ 1−k 2−k
, 2 , 2 , 2
8
ak 2 B 2 G4,1
n2  0
Cerg =
4ln(2)
1,6
ak 2k B 2 G6,2
2
0
 2
1, 0
(21)
n
k=1 k=1
(13)
˛ 
ˇ   1−˛ 2−˛ 1−k 2−k

Cerg =
A2˛−1 B 2 1,6
ak 2k B 2 G6,2
k
16B2
◦  1, 1, 2
, 2 , 2 , 2
(22)
4ln(2) 1, 0

   k=1

A2˛−1 B 2
˛ ˇ
 1−˛ 2−˛ 1−k
, 2 , 2 , 2−k
Pe = 1,4
ak 2k B 2 G4,1
k
16B2
¯  2 2
8 0 4.3. Outage probability
k=1
(14)
Outage probability is defined as the probability of the instanta-
g ˇ+ neous received SNR falling below a particular threshold SNR value,
where B = ˛ˇ
which can be readily validated with computa-

th . It can be expressed as,
tional tools and the average SNR denoted as
¯ =
◦ /2, where
◦
  
is the is the electrical SNR received in the absence of atmospheric
th
(RA◦ ) 2 Pout = Pr [
≤
th ] = Pr [
0 I2 ≤
th ] = Pr I ≤
turbulence and is given by .
0
2n2
    (23)
1 1
4.2. Channel capacity = Pr I≤ √ = FI √

n
n

The channel capacity is the limiting information rate that where

0 is the electrical SNR in the absence of atmospheric tur-
can be achieved when the information is transmitted through a bulence,
th is the threshold SNR and
n is the normalized SNR,

Please cite this article in press as: Vellakudiyan J, et al. Performance of a subcarrier intensity modulated differential phase-shift keying
over generalized turbulence channel. Int J Electron Commun (AEÜ) (2015), http://dx.doi.org/10.1016/j.aeue.2015.06.015
G Model
AEUE-51440; No. of Pages 5 ARTICLE IN PRESS
4 J. Vellakudiyan et al. / Int. J. Electron. Commun. (AEÜ) xxx (2015) xxx–xxx

0 0
10 10

−1
10
−2 σ2I = 0.40
−2
10 10

−3
10
−4
10

Average BER
−4
10
Average BER

−5
10
−6
10 −6
10

−7
10 Ω= 0.5, b0= 0.25
−8
10
−8 α= 14.6, β= 10, ρ= 0.5
10 α= 8.57, β= 6, ρ= 0.65
−9
α= 6.71, β= 4, ρ= 0.8
−10 10 α= 5, β= 3, ρ= 1
10
−10
10
0 10 20 30 40 50 60
0 10 20 30 40 50 60 70 80
Average Electrical SNR, dB
Average Electrical SNR, dB

Fig. 3. ABER against SNR for a same intensity of turbulence (I2 ).

Fig. 2. ABER against SNR for different values of ˛, ˇ, .

20
ρ= 1

0

n =
th . FI (I) is the cumulative distribution function (CDF) and can 18 ρ= 0.75
ρ= 0.25
be expressed as
Ergodic channel capacity (bits/s/Hz)

α= 10, β= 5, Ω= 0.5 ρ= 0
   √1
16
1
n
FI √ = fI (I)dI (24) 14

n 0
12
substitute (1) into (24) and the integral can be solved using Meijer G
function [[25], Eq (26)]. For DPSK based SIM modulation the outage 10
probability can be expressed in a closed form as
8


˛ + k 6
A
ˇ 2
4n2
th
Pout = ak 4
2 R2 A20
k=1
⎛ ⎞ (25)
  ˛+k 2
4n2
th  1−
2,1
×G1,3 ⎝  2 ⎠
R2 A20 B2  ˛−k ˛+k k−˛ 0
0 5 10 15 20 25 30 35 40 45 50
,− ,
2 2 2 Electrical SNR in absence of turbulence, γ0, dB

5. Results and discussions Fig. 4. Ergodic channel capacity.

In this section, average BER (ABER) performance, channel capac-

ity and outage probability of DPSK based SIM-FSO system are
analyzed based on the derivation carried out in previous section.
Fig. 2 shows the ABER performance of the DPSK-SIM which is
evaluated using the Eq. (14) against average SNR for different tur-
bulence conditions. The behavior of gamma gamma distribution
has been obtained for  = 1, which is used to model moderate
to strong atmospheric turbulence and K- distribution for  = 0,
 = 0, is used for modeling the strong atmospheric turbulence.
Here we take different values for ˛, ˇ and  and the correspond-
ing values of irradiance scintillation index I2 is calculated using
equation in [[16], Eq. 26]. From the graph, we can see that the
higher scintillation index results in poorer BER performance for
the same value of  (same family of irradiance distribution). While
considering  = 1, the worst BER performance is obtained for the
case when I2 = 0.36, and the best performance is obtained when
I2 = 0.32. Definitely, this result shows good agreement with the
th ) for
previous literature for every single different statistical distribu-
tion. For all the cases the average optical power is normalized as
0 is taken as 10 dB.
 + 2b0 = 1.
In Fig. 3, ABER performance is analyzed for the same value of
th value till
th reaches 10 dB. For the
irradiance scintillation index, I2 . Here we consider I2 = 0.40. The
th above 10 dB, the Pout varies slowly and tends to approach a
ABER will vary depending on the values of . When  = 1, whole
scattering power travels through USC and coupled entirely to the
LOS component, gets better BER performance. When  decreases,
then the scattering power is distributed into USC and USG , with an
average scattering power of 2b0 and 2b0 (1 − ), respectively.
The more power from the USG component degrades the BER perfor-
mance and effects more harmfully in the performance of a wireless
optical communication system.
Fig. 4 shows the numerical results of ergodic channel capac-
ity obtained from Eq. (22) against SNR for different values of ,
i.e.,  = (0, 0.25, 0.75, 1). Here we take ˛ = 10, ˇ = 5 and  = 0.5.
Here also the transmitted power is assumed to be normalized, i.e.,
 + 2b0 = 1.
Outage probability of the system is described in Fig. 5. The
outage performance is analyzed against threshold SNR (
different values of  in Fig. 5. The electrical SNR in the absence
of atmospheric turbulence of received signal,
So for the threshold SNR below 10 dB, the outage probability is low
and it increases with the

Please cite this article in press as: Vellakudiyan J, et al. Performance of a subcarrier intensity modulated differential phase-shift keying
over generalized turbulence channel. Int J Electron Commun (AEÜ) (2015), http://dx.doi.org/10.1016/j.aeue.2015.06.015
G Model
AEUE-51440; No. of Pages 5 ARTICLE IN PRESS
J. Vellakudiyan et al. / Int. J. Electron. Commun. (AEÜ) xxx (2015) xxx–xxx 5

a 10
0
References

[1] Kedar D, Arnon S. Urban optical wireless communication networks: the main
challenges and possible solutions. IEEE Commun Mag 2004;42(5):S2–7.
[2] Zhu Xiaoming, Kahn JM. Free-space optical communication through atmo-
−1 α= 10, β= 5, Ω= 0.5
10 spheric turbulence channels. IEEE Trans Commun 2002;50(8):1293–300.
ρ= 1 [3] Li J, Liu JQ, Taylor DP. Optical communication using subcarrier PSK intensity
Outage probability, Pout

ρ= 0.75 modulation through atmospheric turbulence channels. IEEE Trans Commun

ρ= 0.25 2007;55(8):1598–605.
ρ= 0 [4] Xu F, Khalighi M-A, Bourennane S. Pulse position modulation for FSO systems:
−2
10 capacity and channel coding. In: 10th International conference on telecommu-
nications, ConTEL 2009. 2009. p. 31–8.
[5] Nistazakis HE, Assimakopoulos VD, Tombras GS. Performance estimation of free
space optical links over negative exponential atmospheric turbulence chan-
nels. Optik - Int J Light Electron Opt 2011;122(24):2191–4. ISSN 0030-4026,
−3 http://dx.doi.org/10.1016/j.ijleo.2011.01.013.
10 [6] Nistazakis HE, Tsiftsis TA, Tombras GS. Performance analysis of free-space opti-
cal communication systems over atmospheric turbulence channels. IEEE Trans
Commun 2009;3(8):1402–9.
[7] Samimi H, Azmi P. Subcarrier intensity modulated free-space optical commu-
nications in K-distributed turbulence channels. IEEE/OSA J Opt Commun Netw
−4
10 2010;2(8):625–32.
0 5 10 15 20 25 30
[8] Hassan MZ, Song X, Cheng J. Subcarrier intensity modulated wireless opti-
Threshold Electrical SNR, γth, dB cal communications with rectangular QAM. IEEE/OSA J Opt Commun Netw
2012;4(6):522–32.
0 [9] Jurado-Navas A, Garcia-Zambrana A, Puerta-Notario A. Efficient lognor-
10
b mal channel model for turbulent FSO communications. Electron Lett
2007;43(3):178–9.
−2 [10] Uysal M, Navidpour SM, Li J. Error rate performance of coded free-space optical
10
links over strong turbulence channels. IEEE Commun Lett 2004;8(10):635–7.
[11] Popoola WO, Ghassemlooy Z. BPSK subcarrier intensity modulated free-
−4 space optical communications in atmospheric turbulence. J Lightwave Technol
10
Outage probability, Pout

2009;27(8):967–73.
[12] Osche GR. Optical detection theory for laser applications. Wiley; 2002.
−6 [13] Andrews L, Phillips RL, Hopen CY. Laser beam scintillation with applications.
10
SPIE; 2001.
[14] Al-Habash MA, Andrews LC, Phillips RL. Mathematical model for the irradiance
−8 probability density function of a laser beam propagating through turbulent
10
media. Opt Eng 2001;40(8):1554–62.
[15] Jurado-Navas A, Garrido-Balsells JM, Paris JF, Puerta-Notario A. A unifying sta-
−10 α= 10, β= 5, Ω= 0.5 tistical model for atmospheric optical scintillation. In: Numerical simulations
10
of physical and engineering processes. Intech; 2011. p. 181–206.
ρ= 1
[16] Jurado-Navas A, Garrido-Balsells JM, Paris JF, Puerta-Notario A. General ana-
ρ= 0.75
−12 lytical expressions for the bit error rate of atmospheric optical communication
10 ρ= 0.25
systems. Opt Lett 2011;36(20):4095–7.
ρ= 0
[17] Garrido-Balsells JM, Jurado-Navas A, Paris JF, Castillo-Vzquez Miguel, Puerta-
−14 Notario A. On the capacity of M-distributed atmospheric optical channels. Opt
10 Lett 2013;38:3984–7.
0 10 20 30 40 50 60
[18] Samimi H. Optical communication using subcarrier intensity modulation
Normalized Electrical SNR in absence of turbulence, γ , dB through generalized turbulence channels. IEEE/OSA J Opt Commun Netw 2012
n
May;4(5):378–81.
Fig. 5. (a) Outage probability. (b) Outage probability. [19] Popoola WO, Ghassemlooy Z, Leitgeb E. BER performance of DPSK subcar-
rier modulated free space optics in fully developed speckle, communication
systems, networks and digital signal processing, 2008. In: 6th International
constant value. The outage performance is analyzed against nor- symposium on CNSDSP 2008. 2008. p. 273–7.
[20] Kiasaleh K. Performance of coherent DPSK free-space optical communica-
malized SNR in Fig. 5. It is shown that the outage probability gets
tion systems in K-distributed turbulence. IEEE Trans Communications 2006
reduced with increase in normalized SNR. Here we take ˛ = 10, ˇ = 5 April;54(4):604–7.
and  = 0.5. [21] Popoola WO, Ghassemlooy Z, Leitgeb E. BER and outage probability of dpsk
subcarrier intensity modulated free space optics in fully developed Speckle. N
Am J Commun 2009 sep;4.
6. Conclusion [22] Samimi H, Uysal M. Performance of coherent differential phase-shift key-
ing free-space optical communication systems in M-distributed turbulence.
IEEE/OSA J Opt Commun Netw 2013;5(7):704–10.
In this paper, the performance of a subcarrier intensity modu-
[23] Jurado-Navas A, Garrido-Balsells JM, Paris JF, Castillo-Vazquez M, Puerta-
lated differential phase-shift keying for FSO communication system Notario A. Impact of pointing errors on the performance of generalized
was presented by considering atmospheric turbulence. A new atmospheric optical channels. Opt Express 2012;20(11):12550–62.
generic model called M-distribution is used to model the fad- [24] Garrido-Balsells JM, Jurado-Navas A, Paris JF, Castillo-Vzquez M, Puerta-Notario
A. General analytical expressions for the bit error rate of atmospheric optical
ing channel. All the previous conventional channel model can be communication systems. Opt Lett 2014;39:5896 [erratum].
expressed from this by varying various parameters. An exact closed [25] Adamchik VS, Marichev OI. The algorithm for calculating integrals of hyperge-
form analytical expression for average BER, channel capacity and ometric type functions and its realization in REDUCE system. In: Watanabe
S, Nagata M, editors. Proceedings of the international symposium on Sym-
outage probability was derived for the proposed system which can bolic and algebraic computation (ISSAC’90). New York, NY, USA: ACM; 1990. p.
be applied for the analysis of the conventional previously proposed 212–24.
channel model like gamma–gamma and K-distribution channel [26] Wolfram, http://functions.wolfram.com
[27] Goldsmith A. Wireless communications. Cambridge University; 2005.
models.

Please cite this article in press as: Vellakudiyan J, et al. Performance of a subcarrier intensity modulated differential phase-shift keying
over generalized turbulence channel. Int J Electron Commun (AEÜ) (2015), http://dx.doi.org/10.1016/j.aeue.2015.06.015