Академический Документы
Профессиональный Документы
Культура Документы
fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TVT.2017.2765004, IEEE
Transactions on Vehicular Technology
Abstract: The next generation of mobile communication The basic idea of UDN is to make users as close to BS as
have progressed towards the deployment of small cells to possible. In order to fulfill such criteria dense deployment
meet the growing demands of increased capacity and QoS. of small cells are done resulting in the generation of high
As compared to conventional method where all the amount of traffic. This huge amount of traffic generation
computations were performed by the BS’s only. To reduce by users creates a challenging task of backhauling during
burden at the BS side, small cells deployment have been the deployment of UDN. The problem of interference
stressed upon. Thus resulting in the formation of Ultra
from the neighboring cells is also predominant factor in
Dense Networks (UDN) having high concentration of small
cells (eg. Picocell, femtocell, hotspots) and as a counter effect limiting the performance of network due to close
of which distance between two small cells is reduced to a proximity of small cells among themselves. Therefore, a
greater extent. Such a close association of cells in UDN, the strict interference mitigation strategies need to be
probability of frequent handovers also increases abruptly. employed to stop hindrance from the neighboring cells
As a result, the chances for attacker to spoof the bandwidth BS. The users who are experiencing sudden arrival and
also increases many folds. In this paper, we have analyzed departure for them interference is a limiting factor for
SNR, SINR and CQI channel parameters. Such that the achieving a considerable higher level of performance
position of moving users at which handover is initiated in goals. The power consumption is also a limiting factor
Picocell deployment of UDN can be traced out successfully
and hence, the motivation for idle model concept should
or region where the probability of attackers presence is high.
Through this paper, we have tried to introduce the attacking be done such that unused small cells can be turned off
scenario for the high mobility conditions in UDN. partially (or fully) or they should be equipped with traffic
offload capability in order to make network efficient in
Keywords: UDN, Small cell, SNR, SINR, Handover, Spoofing, terms of energy or spectrum.
SIT, security issues. The nature of the users inside a UDN is highly mobile
and they are subjected to frequent handover as the user
I. INTRODUCTION switches its location from one cell region to another. This
The increase in demand for high QoS for next generation has attracted many potential attackers to explore the
cellular communication has gathered the interest of many weaknesses present in the UDN architecture. During the
researchers and industries to meet with growing demands handover phase, the attacker might try to steal the
of subscribers. It is expected that the overall traffic information and use for its own benefit known as
requirement of these devices will grow exponentially and bandwidth spoofing attack. However, the eavesdropper
grow up to 30.6 exabytes per month which is eight times action is related with its nature. The monitoring of the
more than the requirement in 2015 [1]. This network behavior by the attacker can be done either
transformation begins in terms of data rate, mobility, actively or passively. But the anomalous behavior of the
coverage area and the spectrum utilization of network. In attacker is very difficult to detect as the information about
order to provide seamless coverage to subscribers with the state of the channel is not readily available.
high data rate capability, small cells are deployed in There are variety of limiting constraints which would
greater density which serves as the best alternative. These greatly affect the propagation of RF signals in dense
small cells deployment under a macrocell results in the environment: 1. firstly, RF signal in the range of 3~300
formation of Ultra Dense Network. From the basic GHz experience high path loss with small size of antenna
definition of UDN, the density of small cell BS will be experiencing more spreading loss, 2. Molecular
very large as compared to UE (User Equipment) density absorption in which a part of energy of propagating signal
such that ideally one UE is entertained by only one BS is converted into kinetic energy. The network
with the capability to support connection for high speed densification techniques incorporated with the help of
vehicles having speed up to 500 km/h [2]. pico/femtocells, relays installation and device-to-device
communication may also give boost or add-on to
interference power even if directional antennas are used
Garima Chopra is a PhD Research Scholar in Department of Electronics at BSs [3]. But Picocells are low transmit BSs with are
and Communication Engineering, Shri Mata Vaishno Devi University,
J&K, India. (E-mail: garimachopra100@gmail.com).
used for indoor/outdoor deployment and often equipped
Sanjeev Jain is a Professor in Department of Computer Science with omnidirectional antenna such that it can act as
Engineering, Shri Mata Vaishno Devi University, J&K, India. (E-mail: Hotspot [3].
dr_sanjeevjain@yahoo.com). In such a case, SINR is a fundamental performance
Rakesh Kumar Jha, is Assistant Professor in School of Electronics and
Communication Engineering, Shri Mata Vaishno Devi University, J&K,
metrics for the analysis of wireless communication
India. (E-mail: jharakesh.45@gmail.com).
1
Copyright (c) 2015 IEEE. Personal use of this material is permitted. However, permission to use this material for any other purposes must be obtained
from the IEEE by sending a request to pubs-permissions@ieee.org.
0018-9545 (c) 2017 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TVT.2017.2765004, IEEE
Transactions on Vehicular Technology
40
security. With the increase in density of UE, the chances
for jamming a larger number of users also increases many 30
folds and getting the accurate knowledge of the channel
state of the eavesdropper also becomes a difficult task. 20
For proper protection, the time taken to detect the
presence of any malicious activity ought to be less than 10
the transmission rate. Such that prevention can be
mitigated before severe damage is caused to the network. 0
1 2 3
With the increase in handover percentage, it is expected Speed of UE(30km/h,60km/h and 120km/h)
that during the handover phase the attacker might try to Fig 2. Handover Failure Rate vs Speed of UE.
1
Copyright (c) 2015 IEEE. Personal use of this material is permitted. However, permission to use this material for any other purposes must be obtained
from the IEEE by sending a request to pubs-permissions@ieee.org.
0018-9545 (c) 2017 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TVT.2017.2765004, IEEE
Transactions on Vehicular Technology
1
Copyright (c) 2015 IEEE. Personal use of this material is permitted. However, permission to use this material for any other purposes must be obtained
from the IEEE by sending a request to pubs-permissions@ieee.org.
0018-9545 (c) 2017 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TVT.2017.2765004, IEEE
Transactions on Vehicular Technology
II. STATE OF THE ART AND CONTRIBUTIONS Once the real time deployment have been performed
It is already known by now that in UDN cell size reduces in the third stage , we have performed the analysis of
in order to provide seamless connectivity (i.e. for Picocell the results on the basis of SINR, path loss, real time
in our case) to the users. As far as practical deployment is positions of attack, handover mobility and we are
concerned, user handover probability is very high and the also able to trace the sensitive time (SIT) and
rate at which handovers occur depends on the speed with spoofing distance with the considerations of all
which user is traveling from one region to another. possible conditions.
Therefore, the channel conditions for high mobile UE The optimal goal of this paper is to find the flaws/issues
varies frequently like from favorable conditions to that will be encountered during handover phase when one
interference conditions between or among different cells. user switches its location from one small cell to another
Hence, the probability of spoofing attack increases in the Ultra dense environment. However, the main
because large number of overlapping regions have been contribution of this paper is to find the possibilities of
appeared during the journey of UE from one BS to other. security attacks on the basis of mathematical analysis.
In this paper, we have proposed an architecture and Also, in the case of handover conditions SNR value is
system model for possible spoofing attack in case of high examined and the SIT (in seconds) is calculated on the
speed handover management. basis SINR values obtained for signals received from
adjacent cellular areas. We have also performed the
Below are the few points listed to provide readers with real-time analysis for actual spoofing attack in UDN. This
the clear understanding about the work performed in this paper also comprises of algorithms which will be helpful
paper: in the development of small cell scenario for observing
In the first phase, we have considered real time the security related issues. Table 1 shows the comparison
deployment of UDN with two possibilities i.e. of existing work with the work presented in this paper.
straight and random path deployment. On the basis However, main purpose is to show the novelty in the
of assumptions, we have derived the SINR and CQI work because comparisons cannot be made due to highly
for real-time system model deployment. diversified topics covered in various areas.
In the second phase, we have proposed a system
model in order to calculate the handover percentage
in different cell approaches.
TABLE 1
COMPARISON TABLE WITH EXISTING WORK
Reference Physical Security Application Technique adopted Results/Conclusion
No. Layer Analysis in covered
Security UDN
[27] X UDN General Analysis Analysis of SR vs UE &
Eavesdropper density for idle mode
BS and LOS path of UE.
[28] X WPT AN assisted IA Drop in Eavesdropper rate as the
scheme noise is increased
[29] X Cognitive Radio Joint information and Minimizing transmission power of
Networks AN Beamforming ST.
Technique in Security for SU and PU.
Spectrum Shared
Networks
[30] X D2D pairs AN scheme AN is used for achieving non-zero SR
1
Copyright (c) 2015 IEEE. Personal use of this material is permitted. However, permission to use this material for any other purposes must be obtained
from the IEEE by sending a request to pubs-permissions@ieee.org.
0018-9545 (c) 2017 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TVT.2017.2765004, IEEE
Transactions on Vehicular Technology
have been computed based on the deployment scenarios Remark 1: In the UDN scenario, the coverage range of
with each user is having different values for cell reduces and they are closely associated, thus
urban,sub-urban and rural areas. The channel conditions increasing the interfering regions many times than the
of each user can be seen from their path loss values and conventional scheme. Here, we are taking two
their corresponding channel gain parameters. Here, we trajectories of user i.e. random as well as straight. We
are checking for the probable regions of handover for a have tried to compute received signal power for both
moving user where the presence of attacker could be felt movements of user.
from its action (see lemma 2). In this paper, we are The main objective of this paper is to check for the
observing signal strengths i.e. the RSS parameters for possibility of security breach in dense conditions which is
checking handover conditions. enabled with frequent handovers for high mobile user,
The Received signal power for the moving client is given hence increasing the probability/chance for the attacker.
by : The attack is initiated by checking the values of SNR (or
Pr ,i | H i , j |2 Ptx 2 (2) SINR) at the interfering regions.
Definition 1: SINR can be defined as the ratio of signal
The received power at the BS can be computed by:
power to noise power and is given as:
Pr ,BS 1 d 2 2
(3) SINR
Signal power
Interference power Noise power
The Total power received at the BS from mobile at a
distance ‘d’ is given by: which can be further rewritten into a mathematical
expression as:
2
PBSt 1 x 2 dx Ptx | H i , j |2
(4) SINR 2
(8)
On further modification, the above integration can be I
reduced to [21]: Pp ,tx Giu , p
SINR
( 1 ) 1 n N 1
N o B Pp ,txa ,iu Giu , p a Pj ,tx G j ,iu
PBSt where 1
( 2) 2 (5)
t 1 j 1
(9)
and the background noise is given by where I represents the summation of interference from
N b KTB other base station (BS) and the adjacent node interference
(6) power. In equation 9, j≠iu for calculating adjacent nodes
and the log-normal distribution R.V ‘yi’ is given by with interference and ‘a’ represents the symbol for adjacent
( i' , i' 2 ) : BS . However, the value of interference due to adjacent
(ln yi i' )2 nodes are very less. Thus, the interference power is
1 2( ' i )2 considered from the adjacent BSs only.
f ( yi ) '
e
2 yii (7)
1
Copyright (c) 2015 IEEE. Personal use of this material is permitted. However, permission to use this material for any other purposes must be obtained
from the IEEE by sending a request to pubs-permissions@ieee.org.
0018-9545 (c) 2017 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TVT.2017.2765004, IEEE
Transactions on Vehicular Technology
Let D { d 1 ,d 2 ,...d n } be the distance vector and for distance ‘x’ is ‘x/v’ where x can be calculated by
each di D SINR and SNR are computed for BS1 and assuming coordinates of user at distance ‘di’. During that
time SINR1 and SINR2 is calculated or signal strength for
BS2 for straight and random trajectory of user as shown in both the BSs. For our analysis purpose, we have taken 3
figure 5. SINR for BS1 is given by positions which shows the maximum possibility of attack
SINR BS1 st / rand { SINR BS1 1 ,SINR BS1 2 ,...,SINR BS1 n } during the movement consisting of P1 which is present at
Similarly, the boundary of interfering region, P2 at the middle of
SINR BS2 st / rand { SINR BS2 1 ,SINR BS2 2 ,...,SINR BS2 n } overlapping region and last P3 at the boundary but close
to cell 2 as shown in Figure 4. All these positions shows
The main purpose of computing the SINR values is to
the random mobility. Our main aim is to check the
extract the colliding point where both BSs have the same positions where handover would take place and hence
values (see lemma 1). such critical regions can act as a sensitive zone.
Additional comments: The interference is due to close
For the real time analysis, we require to find out the
association of BSs i.e nearer the user more is the received
maximum number of Picocells that could fit into a one
power. The channel gain G is predominantly affected by
large macrocell of radius 500m. To calculate the actual
path loss value which varies with indoor or outdoor
number of cells deployed is very important as for real
deployment conditions. time environment. The calculate for number of Picocells
is based on mathematical analysis. Here, we are taking
Definition 2: Channel Quality indicator is defined as
interfering regions to be between 30-50%. The reason for
quantity to measure the performance of RF links
taking such a value is that signal strength degrades only
based on the signal strength values.
around the boundary and we cannot take such cases in
CQI puts the channel conditions on a scale ranging from
which one cell cover almost another cell such case wont
0~30 in which 0 depicts bad channel conditions and it be acceptable. Moreover, only after a break-point distance
should not be used for channel transfer and a highest our strength degrades drastically with inverse of forth
value of 30 depicts very good channel state and UE can
power of distance of node from the BS. So taking such a
transmit the packets with larger block size. The CQI is
range would serve our purpose appropriately and tends to
defined for 3GPP standard by [22]:
provide feasible results. For calculating the number of
0 ; SINR 16dB Picocell BS that can be accommodated in a large cell, we
SINR have divided macrocell into 8 sectors such that maximum
CQI 16.62 ; 16dB SINR 14dB (10) number can be seen during extreme dense conditions. The
1.02 number of sectors(S1, S2, S3,.. so on) varies or can be
30 ; SINR 14dB
varied depending on different approach considered for the
On the basis of the SINR value, handovers may take place sectorization. However, this number may vary according
if the user is not provided with the desired threshold value to the user density as some area may experience
and capacity. Here, capacity is referred to as Shannon degradation in performance due to fading or shadowing
capacity theorem having maximum achievable rate effect but installing more BSs than required could only
required for our application like audio, video or files increase the installation cost and handover probability
which satisfies the desired barrier and it is represented as: when signal strength is sufficient to provide good QoS.
B* log 2 ( 1 SINRi ) . In Figure 5, Picocells deployment is considered in
As far as interference is concerned in handover conditions which a particular user is concentrated and its movement
in the case of Picocells under Macrocell deployment. The is checked. However the user is present at a specific
numerical value as per the Mathematical, Theoretical and location from base station at a distance ' d ' and angle ‘θ’
Actual model implementations, the value of interference with d ( c x )2 ( f y )2 where ( c, f ) and ( x, y ) are
is very low because considering practical deployment
scenario i.e Handover conditions between Picocells, the the coordinates of the Base Station and user respectively
maximum number of interfering points are appeared to be [25]:
only 2-3 points. Throughout the simulation work, the d ;i 0
receivers are equipped with a single-antenna. For ri 2 2
Multi-antenna communication could serve as a possible ( d d i - 2dd i cos( - i ) ;i 1,2,..., pn (11)
extension to the the work for attack analysis. In the The security flaws in the ultra dense network scenario can
analysis purpose random positions for mobile user is be explored by considering a small network consisting of
taken in the overlapping region and distance from a group of small cells which depicts a dense environment.
previous location is computed and receiver signal The typical range of group of small cells are taken to be
strength. 150-200m having the configuration of a Picocells. These
cluster of small cells are embedded within a macrocell
B. Real time deployment in case of Ultra Dense Network having base station NB. Here the concentration on users of
Here, we have taken the BS1 coordinates to be (0,0) only one Picocell with highlighted background. The
and (C1, C2,…) values represent the cell numbers. In case deployment of users, N is purely random within a cell
of UDN, the user is assumed to have high mobility with a region and N≥6. But the stress on mobile users are here
speed of v km/hr. So the time elapsed to cover a certain laid upon which is subjected to frequent change in its
1
Copyright (c) 2015 IEEE. Personal use of this material is permitted. However, permission to use this material for any other purposes must be obtained
from the IEEE by sending a request to pubs-permissions@ieee.org.
0018-9545 (c) 2017 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TVT.2017.2765004, IEEE
Transactions on Vehicular Technology
Fig 4(a) & 4(b). Handover probability in UDN and Probability of interference regions respectively.
1
Copyright (c) 2015 IEEE. Personal use of this material is permitted. However, permission to use this material for any other purposes must be obtained
from the IEEE by sending a request to pubs-permissions@ieee.org.
0018-9545 (c) 2017 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TVT.2017.2765004, IEEE
Transactions on Vehicular Technology
Picocell(100m) and Femtocell(40m) having 1, 4, 12, 30 14: assign new location s.t. above condition satisfies
count respectively. 15: end if
The method used for getting the values are as follows: 16: end for
1.For Macrocell, HO(%)=1/(1+4+12+30)=2%(approx.)
2.Similarly, for Picocell having 200 m radius Algorithm2:Calculation of channel path loss and other
HO(%)=4/(1+4+12+30)=9%(approx.). Using the same performance parameters for users in picocells
procedure, the other two values can be computed in the Initialization:
same manner. 1: assign frequency to users
2: target a mobile user
III. REALIZATION AND REPRESENTATION OF OBJECTIVES LOOP Process:
3: for (distance d i from BS1 and d i1 from BS2 wit
This section comprises of the objectives that is fulfilled
respect to user %for straight trajectory
through the work. However, algorithms best describes the
4: compute path loss from equations (table 4)
method for implementing and realization of objectives.
5: compute received signal power
Hence, we have divided algorithms into three parts: 1)
6: SINR(BS1)=Received power - Pl (BS1 & user)
Algorithm 1 describes the basic deployment scenarios of
7: SINR(BS2)=Received power - Pl (BS2 & user)
the cells and the various parameters associated along with
8: end for
it as shown in Figure 8, and 2) Algorithm 2 describes the
further approach towards the computation of parameters
9: for distance d i' from BS1 and d i1' from BS2
of interest as shown in Figure 9, and 3) Algorithm 3 10: compute path loss from equations (table 4)
describes the loop hole when the attack will occur in a 11: compute received signal power
cell. The main objectives have been listed as follows: 12: SINR(BS1)=Received power - Pl(BS1 & user)
1. Firstly, to determine the spoofing attack in UDN. 13: SINR(BS2)=Received power - Pl (BS2 & user)
2. To determine the handover possibilities(%age) in the 14: end for
UDN. CONDITIONAL statement:
3. Considering the effect of interference in UDN for 15: if (‘A’< SINR(BS1) <‘B’) then %simultaneously
security concerns. calculate CQI for BS2
From fig. 5(a), as per the literature survey, the Handover 16: CQI(BS1)= ‘x’; %based on eq. 11
probability is increased by 10% from 3G to 4G 17: else CQI(BS1)=0;
migrations because we know from the concept as the 18: end if
operating frequency increase resulting in the cell size 19: Plot numerical results
reduction (for end-to-end delivery). As far as 5G is 20: end for
concerned, the frequency of operation is 28-32 GHz
(licensed band) and cell size is up to 500 m(macrocell). Algorithm3:Method for attacking scenario
So, for the UDN large number of Picocells/Femtocells CONDITIONAL statement:
recites under this scenario (Fig. 5(a)) i.e probability of 1: if (SINR(BS1)<SINR(BS2)) then
interference between 2 or more than 2 cells can be 2: Initiate handover
increased by 20-30%. Hence, this is the prime objective 3: Send control messages
of researchers and for our paper also to do the analysis for 4: Compute the time ‘ta’ seconds such that above
spoofing in the case of handover scenarios. condition satisfies
5: Initiate attack s.t Bandwidth is spoofed and user gets
Algorithm1: Deployment of users in picocells for UDN disconnect from region
scenario 6: end if
Initialization:
1: enter Ni ith picocell TABLE 3
2: input transmit power Ptx,p of picocell BS PARAMETERS USED IN PSEUDO-CODE
Notation Description
3: input transmit power Ptx,m of macrocell BS
CONDITIONAL statements: N No. of users in a Picocell
4: if (Ni >max(no. of user in a cell)) then NBS,p No. of Picocell BSs
5: reassign users
CQI(BS1) & Channel Quality information between mobile user
6: end if CQI(BS2) with respect to BS1 and BS2 respectively.
Initialization SINR(BS1) & Signal to noise and interference ratio between user
7: enter Nt & Nr SINR(BS2) with respect to BS1 and BS2 respectively.
8: input cell size (r, θ2 ) and (R, θ1 ) Nt Number of transmit antennas
LOOP Process Nr Number of receiver antennas
9: for i=1 to N do Gr Gain of receivers antenna
10: assign random values (xi,yi)
11: if (distance between((xi,yi) & BS) < r) then Gt Gain of transmit antenna
12: deploy user and mark with ‘*’ NBS,m No. of macrocell BS
13: else ri Distance of User from ith Picocell BS (i=1,2…)
1
Copyright (c) 2015 IEEE. Personal use of this material is permitted. However, permission to use this material for any other purposes must be obtained
from the IEEE by sending a request to pubs-permissions@ieee.org.
0018-9545 (c) 2017 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TVT.2017.2765004, IEEE
Transactions on Vehicular Technology
d Distance of user from pico BS with which it has are having mobility or some may have static orientation.
already established connection. Here, we are talking about the cellular users but they may
ta Time of attack when SINR variation is observed.
be other wireless systems connected through the base
R,θ1 Radius and angle of macrocell deployment station like wireless printers, laptops etc. In this scenario,
r,θ2 Radius and angle of Picocell deployment the various channel models have been applied keeping in
Ptx,M, Pp,tx Power transmitted by macro and Picocell BS mind the area in which it is the deployed typically for
Urban or sub-urban regions. Each user experiences
Pl , σm Penetration losses for indoor user and Standard
deviation of shadow fading for all users in macro
BS
1
Copyright (c) 2015 IEEE. Personal use of this material is permitted. However, permission to use this material for any other purposes must be obtained
from the IEEE by sending a request to pubs-permissions@ieee.org.
0018-9545 (c) 2017 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TVT.2017.2765004, IEEE
Transactions on Vehicular Technology
10
TABLE 5
150
MACRO-PATH LOSS MODEL
Path loss in db
and
d)-3+20log10(fc)-(3.2(
5log10(h)-(24.37-3.7(h/
161.04-7.1log10(W)+7.
140
log10(11.75hUT))2-4.97
2-3.1log10(hBS))(log10(
hBS)2)log10(hBS)+(43.4
1. 10m < d < 5000m
3. W=20m, h=20m
130
Urban (NLOS)
hBS=25m
hUT=1.5m
120
110
2.
100
and
61.04-7.1log10(W)+7.5lo
g10(h)-(24.37-3.7(h/hBS)2
0(fc)-(3.2(log 10(11.75hUT
)log10(hBS)+(43.42-3.1lo
(hBS))(log10(d)-3+20log1
3. W=20m, h=10m
Fig 10. Path loss for different deployment scenarios as shown in table 4.
hBS=25m
))2-4.97)
(log10(d)-3+20log10(fc)-(3.2(l
161.04-7.1log10(W)+7.5log10
BS)+(43.42-3.1log10(hBS))
1
Copyright (c) 2015 IEEE. Personal use of this material is permitted. However, permission to use this material for any other purposes must be obtained
from the IEEE by sending a request to pubs-permissions@ieee.org.
0018-9545 (c) 2017 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TVT.2017.2765004, IEEE
Transactions on Vehicular Technology
11
250
200
150
100
y coordinate(m)
50
-50
-100
-150
-200
-250
30
20
10
-10
0 50 100 150 200 250 300
Fig 12. SNR Variation with respect to distance for scenario represented
in Figure 4.
Case 2: We have applied the same concept for random Fig 14. Possible Spoofing distance for interference of different cells.
mobility. In this case, we have again measured the
handover point for random mobility points. On the basis In Figure 15, we have observed an abrupt decrease in CQI
of the above approach, we are able to compute the SIT with respect to distance. This shows the presence of
and the handover distance as shown in Figure 13 and 14. faulty region where bad channel condition will indicate
Here, we have developed the prototype study for the the probable region of attack.
UDN scenarios and the observations are performed on the
case study basis. Now, the SINR variation is used for a
user for random mobility. The modulation scheme used in
5G is the OFDM and we are assuming that the
sub-carriers are orthogonal to each other such that the
interference between the two user is negligible.
1
Copyright (c) 2015 IEEE. Personal use of this material is permitted. However, permission to use this material for any other purposes must be obtained
from the IEEE by sending a request to pubs-permissions@ieee.org.
0018-9545 (c) 2017 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TVT.2017.2765004, IEEE
Transactions on Vehicular Technology
12
Fig 15. CQI vs distance. Fig. 17. Tracing pattern for Spoofing attack vs different mobility of UE.
On comparing both graphs in Figure 13, 14 & 15, we In our proposed work, we have done pattern analysis of
have observed that at a distance around 150 m, the SIR versus speed and SNR versus SIT with considering
channel conditions of the user receiving signals from both speed. On comparing the results from fig. 16 & 17, we
the BS’s are not very appropriate for further transmission have found that both results are justified to each other and
to continue. at the same time minima and maxima point have been for
the same value of SNR having different value of UE
40 speed corresponding to different speeds. However, the
evaluated results can be used to predict the positions of
35 user irrespective of where the BS’s are deployed. But the
estimation of exact position of user is a complex task but
30 the actual region of attack lies nearer to the computed
values.
SIT (seconds)
25 VII. CONCLUSION
The need for security the systems arises with the increase
20 in capacity demand by the users with considerably good
QoS in future generations of mobile communication . The
15 efforts for securing the transmissions is needed to make
our information secure with the evolution of wireless
10 communication.
In this paper, we have tried to explore the weaknesses
5
20 40 60 80 100 120 140 associated with handover conditions in dense
Speed of UE(km/hr) environment which is expected to continuously switch
Fig 16. Time instant for spoofing attack with respect to mobility of UE. from one cell area to another cell. Thereby increasing the
chances between 20-30% of attack. This paper also
We have performed the above analysis for predicting the concludes that using the trace method approach, once the
position of attack for a UE speed of 60 km/hr. As the SIT(session initiation time) gets synchronized with the
speed of UE increases, time taken to reach the attacking attackers time. Thus, we have created a scenario which
region decreases as shown in Figure 16. When we talk will continuously monitors the signal strength resulting in
about the UDN architecture, we generally assumes the the increase of successful execution of attack. The
speed of moving user to be high. Hence, the above protection against this will serve as a major challenge for
analysis shows the exponential decay trend as the speed the researchers with the increase in connection make and
increases. By analyzing the SIT vs Speed graph, we have break.
concluded that handover time varies with UE speed. In One approach that we have used in this paper is by
Figure 17, we have tried to analyze the trend that would tracing the victim by constantly checking the values of
be followed by the user for different speeds for SNR SNR and SINR. Another possible solution by seeing the
values in the case of random motion. The square filled CQI parameters where extremely low numerical values
with blue color is the values for SNR’s when user is indicates the region where the attacker is more
having high mobility of 120 km/hr. Whereas, triangle susceptible to be present in that region and the channel
shaped values marked are the values for SNR’s for cell1 conditions deteriorates.
and cell2 for both when user is moving with 60 km/h. In this paper, we have concentrated on only one scenario.
Similarly, circle having green color marked values are for The other challenges which would be present in UDN are
speed of 30 km/h. as follows:
1
Copyright (c) 2015 IEEE. Personal use of this material is permitted. However, permission to use this material for any other purposes must be obtained
from the IEEE by sending a request to pubs-permissions@ieee.org.
0018-9545 (c) 2017 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TVT.2017.2765004, IEEE
Transactions on Vehicular Technology
13
1
Copyright (c) 2015 IEEE. Personal use of this material is permitted. However, permission to use this material for any other purposes must be obtained
from the IEEE by sending a request to pubs-permissions@ieee.org.
0018-9545 (c) 2017 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TVT.2017.2765004, IEEE
Transactions on Vehicular Technology
14
1
Copyright (c) 2015 IEEE. Personal use of this material is permitted. However, permission to use this material for any other purposes must be obtained
from the IEEE by sending a request to pubs-permissions@ieee.org.
0018-9545 (c) 2017 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TVT.2017.2765004, IEEE
Transactions on Vehicular Technology
15
1
Copyright (c) 2015 IEEE. Personal use of this material is permitted. However, permission to use this material for any other purposes must be obtained
from the IEEE by sending a request to pubs-permissions@ieee.org.
0018-9545 (c) 2017 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.