DESIGN OF SECURE AND RELIABLE MU-MIMO TRANSCEIVER SYSTEM FOR VEHICULAR NETWORKS

International Journal of Computer Networks & Communications (IJCNC) Vol.11, No.2, March 2019
DOI: 10.5121/ijcnc.2019.11202 15
DESIGN OF SECURE AND RELIABLE MU-MIMO
TRANSCEIVER SYSTEM FOR VEHICULAR NETWORKS
C. Manikandan, P. Neelamegam, S.Rakesh Kumar, G. Venkat Babu, and Sai Siva
Satwik Kommi
School of Electrical and Electronics Engineering, SASTRA Deemed University,
Thanjavur, Tamil Nadu, India.
ABSTRACT
MU-MIMO (Multi-User MIMO) has been a promising technique for vehicular networks to achieve faster
communication. Conventional MU-MIMO transceiver is designed with One-dimensional (1-D) improper
modulation schemes such as Binary Phase Shift Keying (BPSK) and Multilevel Amplitude Shift Keying (M-
ASK) failed to yield standard ABER (average bit error rate). To achieve high reliability, a novel MU-
MIMO uplink transceiver system is designed under PAPC (Per-Antenna Power Constraint) by assuming
perfect and imperfect channel state information (CSI). MIMO communication channels are perceptible.
Hence, security of the proposed system is improved by novel pseudorandom key generation technique using
randomized synthetic colour image. Analytical design for proposed systems is carried and simulated for
various p-norm constraints. Simulation results show higher reliability and security than the existing system.
It also satisfies the linearity constraint of a power amplifier, which makes the system more suitable for
practical applications.
KEYWORDS
MU-MIMO, Uplink, PAPC, 1-D improper modulation, Perfect & imperfect CSI, Colour Image,
Pseudorandom key generation
1. INTRODUCTION
Vehicular networks are widely used for autonomous navigation, remotely operated vehicle, and
swarm robot [1]. It utilizes cellular and wireless local area networks and demands high-speed,
reliable communication without increasing channel bandwidth and power required for
transmission [2-4]. To meet such requirement, MU-MIMO communication technology has been
introduced in such applications [5]. Several MU-MIMO transceivers design has been reported to
increase throughput and minimize total mean square error (TMSE). In all the cases a common
transceiver design has been proposed for both proper and improper modulation [6-13]. The
outcome revealed that the Average Bit error rate (ABER) performance of existing MU-MIMO
transceiver systems was suboptimal for improper modulation. To overcome this problem, a
separate MU-MIMO transceiver system had been designed for improper modulations with novel
precoding strategy [14]. Joint optimal precoders and decoders were designed for improper
modulation to improve the ABER performance [15-18].
Predominately the exiting MU-MIMO transceiver systems were designed under TPC (β). In TPC
based power allocation techniques, linearity constraint of the individual power amplifier is not
considered. This makes the system fail to meet the practical requirements and makes its
challenging to be realized in real-time. An SU-MIMO transceiver has been proposed with 1-D
improper modulation and PAPC [19] to meet the practical requirements in point-to-point
communication. This has been extended for MU-MIMO uplink transceiver system. Additionally,
to provide security, a simple and efficient novel key generation algorithm has been developed

16
using randomized synthetic colour image. This makes the proposed system to have high data rate
and reliability with security, which are the primary requirements for vehicular network.
The remaining sections of this article are organized as follows: Section 2 illustrates the system
model used for uplink vehicular network and describes the design procedure of the proposed MU-
MIMO transceiver systems and its key generation techniques. In section 3 numerical results of the
proposed system are provided. Section 4 concludes the paper with future directions.
2. SYSTEM MODEL
Figure 1. MU-MIMO Vehicular system models
The vehicular network provides reliable communication between the individual vehicles and the
base station [1]. In the proposed work, uplink MU-MIMO transceiver is implemented in
infrastructure-based vehicular networks. It consists of a single Base Station (BS) and four
Vehicular Nodes (VNs) as shown in Figure 1. Both the BS and the VNs are equipped with
multiple antennas. Each VNs connects to the base station with an independent uplink channel
Hm
(UL)
where m = 1, 2…4. All the vehicles utilize the same frequency spectrum for signal
transmission. However, they are separated in the spatial domain. Vision sensor attached at front-
end of each VNs simultaneously transmits its information to BS through Hm
(UL)
. After receiving
the information, BS generates a command to navigate multiple vehicles on the road.

17
Figure 2. Proposed MU-MIMO uplink system
The block diagram of the proposed MU-MIMO uplink transceiver system with ‘m’ channels is
shown in Figure 2. For the encryption process, the analog data obtained from the sensor is
converted into binary and XORed with a key to get cipher text. This encrypted plain text is
converted to parallel form and modulated using 1-D improper modulator for transmission. Digital
modulation scheme yielding improper signal is called as improper modulation. In the proposed
linear MIMO transceiver system improper modulation such as BPSK and M-ary ASK was
implemented its pseudo-autocorrelation and cross-correlation function of complex envelope
signals are assumed as nonzero [18]. A linear precoder concerning PAPC (α) is used to encode
data stream of each VNs. Since the uplink channel matrices are complex, the coded data streams
are multiplied by complex channel matrix (Hm
(UL)
) and decoded by the linear decoders at the
receiver. As the modulator outputs consist of only real components. Hence the real part of
decoder output is considered for detection. The precoder and decoder matrix is updated based on
the assumption of perfect and imperfect channel state information known at the BS and VN.

18
Figure.3. Flow Chart for key generation algorithm
A novel image-based key generation technique proposed for the stream cipher encryption and
decryption algorithm is described in Figure.3. A 256×256 random pixel synthetic colour image is
chosen as shown in Figure. 4. The colour image is made up of three layers namely red, green and
blue. The pixel values in each layer vary from 0 to 255 representing the intensity level. Therefore,
8 bits are required to represent the value of one pixel. This facilitates the generation of 8 separate
keys with a size of 65536 bits by extracting one bit at a time from LSB to MSB as shown in
Figure.5.
Synthetic Colour
Image
Image Splitter
R-Plane
Colour Plane
Sequencer
Key Extraction
(LSB to MSB)
Store Keys in Data Base
G-Plane B-Plane
If
Key is
Random
Key Selector
No
Discard
Total
Key=8
Yes
No
Yes
Km

19
The key generation process is repeated for all the three layers, and a total of 24 keys are
generated. The keys fulfill the randomness property are used for encryption process, and others
are discarded. To get better protection against brute force attacker, a key selection process for
encryption and decryption also performed randomly.
Figure. 4 Random pixel synthetic colour image
Figure. 5 Method for Key Generation
2.1. Problem formulation for minimum TMSE design with perfect CSI
The mathematical model for uplink MU-MIMO system by assuming channel matrix H is
perfectly known to both transmitter and receiver is derived as follows. This model is used to
minimize the TMSE for the detected symbol. The notations and their meanings used in this paper
are given in Table 1.
01001101
11001100
01001101
01101100
10101101
01001000
01001101
11001100
01001101
LSBMSBLSBMSBLSBMSB
Pixel
C1 C2
R2
C256
R256
R1
K-1 K-2 K-3 K-4 K-5 K-7 K-8K-6

20
Table 1. Notations and their meanings used in the system model
Notation Meaning
Km Km is a randomly chosen key from the finite set of possible keys
)(UL^
ld Output data vectors form the lth
decoder
la Input data vectors to the lth
precoder
)(UL^
mH Uplink MIMO channel matrices of mth
vehicle
)(UL
ln Additive White Gaussian Noise (AWGN) vector with zero-mean and
variance 2)(
)( UL
n .whose entries are spatially, temporally and
independent and identically distributed (i.i.d)
mP Precoding matrix for the mth
Vehicle
lD Decoding matrix for the lth
Vehicle
Β TPC
Α PAPC
. p-norm norm
E(·) Expectation
R(·) Real part of a complex-valued vector
(·)-1
Matrix inverse
(·)T
Matrix transpose
(·)*
Matrix complex conjugate
(·)H
Matrix Hermitian
Tr(·) Trace of a matrix
IB Identity matrix,
NT Number of transmit antennas
NR Number of receive antennas
Bm Data stream for the mth
Vehicle
The TMSE matrix for uplink can be formulated as,
(1)
where )][(= )()(
1=
)(^ UL
llmm
UL
m
K
ml
UL
l R nDaPHDd 
Substituting the value of )(^ UL
ld in (1), we obtain
])][([=][
2
)()(
1=
2)(
l
UL
llmm
UL
m
K
m
l
UL
REE anDaPHDe 
(2)
The TMSE calculation is expanded as follows:
][=][
2)(^2)(
l
UL
l
UL
EE ade 

21
)}
))()(0.5()()()(0.5(
))(0.5())((0.5({=][
1=
)()()(
1=
)(
1=1=
*)(*)(***)(
1=
*)(
1=
2)(
H
TTHHTTTTHHHH
m
K
m
UL
ll
UL
llmm
UL
m
K
mlmm
UL
m
K
mlm
K
m
UL
ll
UL
llmm
UL
m
K
mlmm
UL
m
K
ml
UL
ETrE
a
nDnDaPHDaPHDa
nDnDaPHDaPHDe






(3)
Using assumptions made in [15-18], equation (3) is simplified for calculation of TMSE
})0.25()0.25(
)(0.5)(0.5)(0.50.5
)()(0.25)()(0.25
)(0.25)({0.25=][
*2)(
,
2)(
,
)(
1=
)(
1=
**)(
1=
*)(
1=
*)(*)(
1=
**)(*)(
1=
*
)()(
1=
)()(
1=
2)(
lBll
UL
lnll
UL
ln
m
UL
m
K
m
lm
UL
m
K
m
lm
UL
m
K
m
lm
UL
m
K
m
l
lmm
UL
m
UL
m
K
m
llmm
UL
m
UL
m
K
m
l
lmm
UL
m
UL
m
K
m
llmm
UL
m
UL
m
K
m
l
UL
TrE
IDDDD
PHDPHDPHDPHD
DPPHHDDPPHHD
DPPHHDDPPHHDe







TH
HHHTTT
HHHTTT
HHHTTT

(4)
To minimize the TMSE subject to PAPC,
)][(min
2)(UL
Etr e
mD,mP
αTrts pp
mm
K
m
 1/
1=
))((.. H
PP
(5)
Where α is the PAPC.
Using Lagrangian to get the solution for the defined problem
)]))((([][= 1/
1=
)(2)()(
αTrμE pp
mm
K
m
ULULUL
  H
PPe
(6)
where )(UL
μ is the Lagrange multiplier.
Substitute equation (4) in (6) to get an expanded version of (6). The uplink one-dimensional
improper modulation based transceiver design problem in (5) is non-convex and also
continuously differentiable. Therefore by taking the derivatives of )(UL
 with respect to lD and
lP , Then the related Karush-Kuhn-Tucker(KKT) conditions can be obtained and given in the
following
0=
)(
l
UL
η
D

HHHHHH
)(2=})()({})({ )()(**)(
1=
*2)()()
1=
UL
Zz
UL
Zzz
UL
Z
K
z
z
UL
n
UL
zzz
UL(
Z
K
z
z HPHPPHDHPPHD   
(7)
0=
)(
l
UL
η
P


22
HHHHHH
z
UL
Zz
K
z
)UL(
z
UL
Z
UL
Z
K
z
zzz
UL
Z
UL
z
K
z
zz k( DHPPHHDDPHHDD )2(=2}){})()({ )(
1=
)()(
1=
*)(*)(
1=
*
  
(8)
where 1*
1=
1)(1/
1= )(= 

p
zz
K
z
pp
zz
K
z ][]Tr[k PPPP TH
The expression of Lagrange multiplier is obtained by equating the equations (7) and (8).
)]([
2
)(
= 1=
2)(
)( H
zz
K
z
UL
nUL
Tr
α
μ DD

(9)
An iterative procedure shown in Figure 2 is used to find the optimum solution for zD and zP
ImzRezz j ,,= DDD  ,
}{= )()
1=
)(
,
)
,
HH UL
zzz
UL(
Z
K
z
z
UL
Imz
UL(
Rez j HPPHDEE  ,
})()({= )(**)(
1=
)(
,
)(
,
HH UL
Zzz
UL
Z
K
z
UL
Imz
UL
Rez j HPPHFF  ,
HH
)(2= )()(
,
)(
,
UL
Zz
UL
Imz
UL
Rez j HPGG  .
Then Rez,D and Imz,D can be expressed as
1
2)(
,,,,
,,
2)(
,,
,,,,
)(
)(
=





























RN
UL
n
)UL(
Rez
)UL(
Rez
)UL(
Imz
)UL(
Imz
)UL(
Imz
)UL(
ImzRN
UL
n
)UL(
Rez
)UL(
Rez
)UL(
Imz
)UL(
RezImzRez
IFEEF
FEIFE
GGDD


(10)
Likewise, define
ImzRezz j ,,= PPP 
,
}){= )()(
1=
)(
,
)(
,
HH UL
Z
UL
Z
K
z
zz
UL
Imz
UL
Rez (j HHDDRR  ,
*)(*)(
1=
)(
,
)(
, )()(= UL
zzz
UL
Z
K
m
UL
Imz
UL
Rez j HDDHSS HH

HH
z
UL
z
UL
Imz
UL
Rez j DHTT )2(= )()(
,
)(
, 
.
Then Rez,P and Imz,P can be expressed as

































)(
,
)(
,
1
,
)(
,
)(
,
)(
,
)(
,
)(
,
)(
,,
)(
,
)(
,
,
,
2
2(
= UL
Imz
UL
Rez
TN
)UL(UL
Rez
UL
Rez
UL
Imz
UL
Imz
UL
Imz
UL
ImzTN
)UL(UL
Rez
UL
Rez
Imz
Rez
k
k
T
T
ISRSR
RSISR
P
P


(11)

23
2.2. Problem formulation for minimum TMSE design with imperfect CSI
For imperfect case, the th
m user channel is modeled as
)()()(
= UL
m
UL^
m
UL
m EHH  (12)
To estimate the channel matrix, training sequences were used.
1/2)(1/2
,
)(
= T
UL^
wmmR
UL^
m RHRH and
1/2)(
,
1/2
,,
)(
= T
UL
mwmRe
UL
m RERE
, m=1...K (13)
where 1/2
TR & 1/2
,mRR are the correlation matrices and )(UL
mE is the error model. The entries of
)(UL^
wmH and )(
,
UL
mwE are independent and identically distributed.
Substitute equation (13) in (12) to get an expanded version of (12).
1/2)(
,
1/2
,,
1/2)(1/2
,
)(
= T
UL
mwmReT
UL^
wmmR
UL
m RERRHRH  (14)
where 11
,
2
,,,, ][= 
 mRmcemRNmRe RIR  and mtrnTmce Tr ,
212
, /)(= PR  
. In mRe ,,R ,
2
m,ce is a
channel estimation error variance and in 2
,mce , 2
n is a noise variance and mtr,P is a training
power of the user m.
The TMSE calculation for the improper modulation based uplink MU-MIMO system for
imperfect case defined as follows:
][=][
2)(^2)(
l
UL
l
UL
EE ade 
(15)
where )])([(= )()()(
1=
)( UL
llmm
UL
m
UL^
m
K
ml
UL^
l R nDaPEHDd  is the vector after decoder.
Substituting the value of )(UL^
ld in (15), we obtain
])])([([=][
2
)()()(
1=
2)(
l
UL
llmm
UL
m
UL^
m
K
m
l
UL
REE anDaPEHDe  (16)
The TMSE calculation is expanded as follows:
)}))((0.5())(
)(0.5())()((0.5(
)))(0.5())(0.5(
))((0.5({=][
1=
)())(
1=
)(
1=
)()(
1=
)(
1=
1=
*)(*)(**
1=
*)(*
1=
)(
***)(
1=
*)(
1=
2)(
HTTHHTTTT
HHHHTTTTHHHH
m
K
m
UL
ll
UL
llmm
K
m
UL
ll
mm
K
m
UL
llmm
UL^
m
K
mlmm
UL^
m
K
ml
m
K
m
UL
ll
UL
llmm
K
m
UL
llmm
K
m
UL
ll
mm
UL^
m
K
mlmm
UL^
m
K
ml
UL
ETrE
anDnDaPED
aPEDaPHDaPHD
anDnDaPEDaPED
aPHDaPHDe








(17)
Using the following assumptions made in [15-18] simplifies the equation (17). Then TMSE is
calculated as

24
2)(*
2
,
*
1=
*
,,
*)(**)(
1=
*
)(**)(
1=
*)(
1=
2)()(
1=
2
,
1=
,,
)(
1=
)()(
1=
**)(
1=
*)()(
1=
2)(
)(0.25
)]([0.25])()([0.25
])()([0.25)(0.5
)(0.250.5)]([0.25
)(0.5])([0.25
)(0.5])([0.25=][
UL
nll
lcemmT
K
m
llRell
UL^
mmm
UL^
m
K
m
l
l
UL^
mmm
UL^
m
K
m
llBl
UL^
m
K
m
l
UL
nllm
UL^
m
K
m
llcemmT
K
m
llRel
l
UL^
m
K
m
ll
UL^
mmm
UL^
m
K
m
l
m
UL^
m
K
m
ll
UL^
mmm
UL^
m
K
m
l
UL
Tr
Tr
{TrE



T
HTTTT
HHHTTT
HHH
HHHTTT
HHH
DD
PPRDRDDHPPHD
DHPPHDIPHD
DDPHDPPRDRD
PHDDHPPHD
PHDDHPPHDe











(18)
To minimize the TMSE subject to PAPC, hence
)][(min
2)(UL
Etr e
mD,mP
αTrts pp
mm
K
m 
1/
1= ))((.. H
PP (19)
Where α is the PAPC.
Using Lagrangian to arrive at the solution for the defined problem
)]))((([][= 1/
1=
)(2)()(
αTrμE pp
mm
K
m
ULULUL
  H
PPe
(20)
where )(UL
μ is the Lagrange multiplier.
By taking the derivatives of η with respect to Dl & lP , it can be shown that the Lagrangian can be
derived.
0=
)(
l
UL
η
D

HH
HHHH
)(2=)(
)()())()((
)(2)(
)(**)(
1=
*
1=
2
,,,
)()(
1=
UL^
Zzz
UL
n
UL^
Zzz
UL^
Z
K
z
z
H
zz
K
z
TzcezRe
UL^
zzz
UL^
Z
K
z
z Tr
HPD
HPPHDPPRRHPPHD



 
(21)
0=
)(
l
UL
η
P

z
UL^
Zz
K
z
)UL(
UL^
zzz
K
z
UL^
Zzz
H
zz
K
z
zRezceT
UL^
Zzz
K
z
UL^
Zz
k
Tr
DHP
HDDHPPDDRRHDDHP
H
HHHH
)2(=2
)()())()((
)(
1=
*)(*
1=
)(*
1=
,,
2
,
)(
1=
)(






(22)
where 1*
1=
1)(1/
1= )(= 

p
zz
K
z
pp
zz
K
z ][]Tr[k PPPP TH
The expression of Lagrange multiplier is obtained by equating the equations (21) and (22).

25
)]([
2
)(
= 1=
2)(
)( H
zz
K
z
UL
nUL
Tr
α
μ DD

(23)
An iterative procedure is shown in Figure.6. is used to find the optimum solution for zD and zP
ImzRezz j ,,= DDD  ,
)()(= 1=
2
,,,
)(
1=
)()(
,
)
,
HHH
zz
K
zTzcezRe
UL^
Zzz
K
z
UL^
Z
UL
Imz
UL(
Rez Trj PPRRHPPHEE    ,
HH
)()(= )(*
1=
*)()(
,
)(
,
UL^
Zzz
K
z
UL^
Z
UL
Imz
UL
Rez j HPPHFF  ,
HH
)(2= )()(
,
)(
,
UL^
Zz
UL
Imz
UL
Rez j HPGG  .
Then Rez,D and Imz,D can be expressed as
1
2)(
,,,,
,,
2)(
,,
,,,,
)(
)(
=





























RN
UL
n
)UL(
Rez
)UL(
Rez
)UL(
Imz
)UL(
Imz
)UL(
Imz
)UL(
ImzRN
UL
n
)UL(
Rez
)UL(
Rez
)UL(
Imz
)UL(
RezImzRez
IFEEF
FEIFE
GGDD


(24)
Likewise, define
ImzRezz j ,,= PPP  ,
)()(= 1=,,
2
,
)()(
1=
)(
,
)(
,
HHH
zz
K
zzRezceT
UL
zzz
UL^
z
K
z
UL
Imz
UL
Rez Tr^j DDRRHDDHRR    ,
*)(*)(
1=
)(
,
)(
, )()(= UL^
Zmm
UL^
Z
K
m
UL
Imz
UL
Rez j HDDHSS HH
 ,
z
UL^
Z
UL
Imz
UL
Rez j DHTT H
)2(= )()(
,
)(
,  .
Then Rez,P and Imz,P can be expressed as

































)(
,
)(
,
1
,
)(
,
)(
,
)(
,
)(
,
)(
,
)(
,,
)(
,
)(
,
,
,
2
2(
= UL
Imz
UL
Rez
TN
)UL(UL
Rez
UL
Rez
UL
Imz
UL
Imz
UL
Imz
UL
ImzTN
)UL(UL
Rez
UL
Rez
Imz
Rez
k
k
T
T
ISRSR
RSISR
P
P


(25)
With the help of duality theory, the above design of the Uplink MU-MIMO transceiver system
can be implemented for the downlink design.

26
Figure 6. Iterative procedure to get optimum precoder and decoder
3. RESULTS &DISCUSSION
The proposed work aims to improve the security and reliability of MU-MIMO transceiver system.
The security of the systems is evaluated by using randomness test, and the reliability of the
system is tested in terms of ABER. The proposed cryptographic key generation technique used
for security is validated using the National Institute of Standards and Technology (NIST) test
suite, and its probability values of the tests are given in Table.2. It is observed that the probability
values resulting from all the tests are greater than 0.01. [20] These results show that the generated
keys are random in nature. Similarly, the reliability of the proposed MU-MIMO uplink system
mathematical model is tested by evaluating the ABER concerning SNR using MATLAB in a
simulation environment with parameters described in Table. 3.
Initialize Pz
Update Dz Using (10) & (24)
Update µ(UL)
Using (9) & (23)
If
αTr pp
zz
K
z 
1/
1= ))(( H
PP
Update Pz Using (11) & (25)
Stop
Yes
No
Yes
Scale Pz such that
αTr pp
zz
K
z 
1/
1= ))(( H
PP
If
410
/p1
)
p
))1m
zz
)(1m
zz
((Tr(K
1z=


HPPPP
No

27
Table.2 NIST Parameter analysis
Test P-Key1 P- Key2 P- Key3 P- Key4 P- Key5 Result
Frequency 0.534146 0.122325 0.035174 0.350485 0.911413 Pass
Block Frequency 0.534146 0.213309 0.739918 0.739918 0.739918 Pass
Cumulative Sums 0.739918 0.122325 0.213309 0.213309 0.534146 Pass
Runs 0.911413 0.122325 0.350485 0.350485 0.350485 Pass
Longest Run 0.350485 0.739918 0.017912 0.534146 0.739918 Pass
Rank 0.066882 0.213309 0.534146 0.213309 0.739918 Pass
FFT 0.739918 0.350485 0.534146 0.534146 0.911413 Pass
Non-Overlapping
Template
0.991468 0.911413 0.739918 0.911413 0.911413 Pass
Approximate Entropy 0.122325 0.122325 0.035174 0.739918 0.739918 Pass
Serial 0.350485 0.534146 0.017912 0.350485 0.739918 Pass
Linear Complexity 0.350485 0.534146 0.534146 0.534146 0.066882 Pass
Table. 3. Simulation parameters and its values
Parameter Values
Vehicular stations 4
Base station 1
No of transmitting antennas(NT) at each
vehicle
4
No of receiving antennas(NR) at base
station
16
No of the data stream (B) from each vehicle 4
p-values 4.12 2.36 1.76
α β α β α β
1.1W 3.16W 2.8W 6.31W 5.5W 10W
The ABER is compared for both proposed and conventional transceiver system, at the output of
decryption algorithm by concerning proposed key generation technique. Two improper
modulation techniques namely BPSK and 4-ASK modulation are used for evaluation. Standard p-
values such as 4.12, 2.36, and 1.76 are also considered and simulated for both perfect and
imperfect CSI condition.
Figure 7. Performance comparison results of MU-MIMO uplink system with respect to perfect CSI for
p=4.12

28
Figure 7 shows the performance of the proposed MU-MIMO systems for BPSK and 4-ASK
modulation for the p-value of 4.12 with a perfect CSI at the transmitter and receiver. It is
observed that SNR is increased from 7.5dB to 9.5dB for BPSK and from 11.5dB to 13.5dB in the
case of 4-ASK. This indicates that the proposed system demand an additional 2 dB of SNR at the
ABER level of 10-2
. This notifying increase in SNR is worth as the design is more realistic.
p=2.36
The p-value is decreased and simulated for p=2.36 as illustrated in Figure 8. The SNR is
increased from 4.5dB to5.5dB for BPSK and from 8.5dB to 9.5dB in the case of 4-ASK. Here an
SNR increase of 1 dB is observed at the same ABER level of 10-2
for both the improper
modulation schemes. A lesser increase in SNR showcases that the system is nearing optimal as
the p-value is reduced from 4.12 to 2.36.
p=1.76
The p-value is further reduced close to unity (p=1.76), and its ABER is evaluated as in Figure 9.
It is observed that for BPSK modulation, both TPC and PAPC require an SNR value of 2.5dB and
4-ASK modulation requires an SNR of 6.5dB at 10-2
ABER. This makes the SNR requirement for
PAPC is same as TPC. From these results, the proposed system found to be optimal for practical
realization.

29
The proposed system is also tested for imperfect CS condition for various p-values (p=4.12 and
p=2.36). The values of transmitting and receive correlation matrices are set to be ρT=0.5, ρR=0.5
respectively and the channel estimation error is set as σce
2
=0.015 to simulate the imperfect
channel state condition.
Figure 10a. Performance comparison results of MU-MIMO uplink system with respect to imperfect CSI for
p=4.12.
Figure.10a&10b.shows the comparisons of MU-MIMO uplink system with TPC and PAPC for
the imperfect CSI. It illustrates BPSK and 4-ASK requires an increase of 2 dB (BPSK: 10dB to12
dB & 4-ASK: 14dB to 16dB) and 1 dB (BPSK: 7 to8dB & 4-AKS: 11dB to 12dB) for p-value of
4.12 and 2.36 respectively at 10-2
ABER level. A difference of 2.5dB in the SNR is observed
between the perfect and imperfect case, which makes the system suitable for realization.
Figure 10b. Performance comparison results of MU-MIMO uplink system with respect to imperfect CSI for
p=2.36.

30
4. CONCLUSIONS
In this paper, 1-D improper modulation based uplink MU-MIMO transceiver employing PAPC
with TMSE criterion is proposed for the vehicular network to achieve reliability. In addition to
that to achieve security, a novel pseudo-random key generation technique was proposed for its
stream cipher technique. The proposed key generation technique uses random pixel value based
synthetic colour image to generate keys. Five out of 24 generated keys satisfy the randomness
test, and its probability value is found to be greater than 0.01. The proposed key generation
techniques make the RF-based vehicular network secure and also the overall transceiver design to
be less complex.
Reliability is improved by jointly optimizing the precoder and decoder of the transceiver system
using the iterative algorithm. It uses the inverse water-filling technique with respect to PAPC for
power allocation, which satisfies the linearity constraint of the power amplifier and makes the
system realizable. Simulation studies of the proposed system are carried out for various p-values
in the range 1<p<∞ for both perfect and imperfect CS condition. An optimum solution is obtained
when the value of p approaches unity (p=1.76). Hence the proposed mathematical model for
MU-MIMO transceiver system meets the practical requirement of a vehicular network design. It
can also be implemented for other RF-MIMO & visual MIMO communication networks which
required low ABER and high security.
ACKNOWLEDGMENTS
Finally, I would like to thank god for giving me an opportunity, encouragement, and strength to
finish this work.
REFERENCES
[1] Karagiannis, Georgios, Onur Altintas, Eylem Ekici, Geert Heijenk, Boangoat Jarupan, Kenneth Lin,
& Timothy Weil (2011) “Vehicular networking: A survey and tutorial on requirements, architectures,
challenges, standards and solutions”, IEEE communications surveys & tutorials, Vol. 13, No. 4,
pp584-616.
[2] Saliyah Kahar, Riza Sulaiman, Anton Satria Prabuwono, Nahdatul Akma Ahmad, & Mohammad
Ashri Abu Hassan(2012) “ Review of Wireless Technology Usage for Mobile Robot Controller,”
2012 International Conference on System Engineering and Modeling, Vol. 34, pp7-12.
[3] Jetmir Haxhibeqiri, Elnaz Alizadeh Jarchlo, Ingrid Moerman, & Jeroen Hoebeke, (2018) “Flexible
Wi-Fi Communication among Mobile Robots in Indoor Industrial Environments,” Hindawi Mobile
Information Systems, Vol. 2018, pp1-19.
[4] Ningxiao Sun, Yuejin Zhao, Lin Sun, & Qiongzhi Wu, (2017) “Distributed and Dynamic Resource
Management for Wireless Service Delivery to High-Speed Trains,” IEEE Access, Vol. 5, pp622-632.
[5] Jankiraman, Mohinder. (2004). Space-time codes and MIMO systems. Artech House
[6] Wong, Kai-Kit. (2004).“Performance analysis of single and multiuser MIMO diversity channels using
Nakagami-m distribution,” IEEE Transactions on Wireless Communications, Vol.3, No. 4, pp1043-
1047.
[7] Spencer, Quentin H., A. Lee Swindlehurst, & Martin Haardt. (2004). “Zero-forcing methods for
downlink spatial multiplexing in multiuser MIMO channels,” IEEE transactions on signal processing,
Vol.52, No. 2, pp461-471.
[8] Serbetli, Semih, & Aylin Yener. (2004),“Transceiver optimization for multiuser MIMO systems.”
IEEE Transactions on Signal Processing, Vol.52, No. 1, pp241-226.

31
[9] Schubert, Martin, Shuying Shi, Eduard A. Jorswieck, & Holger Boche. (2005). “Downlink sum-MSE
transceiver optimization for linear multi-user MIMO systems,” Conference Record of the Thirty-
Ninth Asilomar Conference on Signals, Systems and Computers, pp1424-1428.
[10] Zhang, Jinfan, Yongle Wu, Shidong Zhou, & Jing Wang. (2005),“ Joint linear transmitter and receiver
design for the downlink of multiuser MIMO systems,” IEEE communications letters, Vol.9, No. 11,
pp991-993.
[11] Shi, Shuying, & Martin Schubert. (2005), “MMSE transmit optimization for multi-user multi-antenna
systems.” IEEE International Conference on Acoustics, Speech, and Signal Processing, pp409-412.
[12] Songnan Xi, & Michael D. Zoltowski. (2006), “Transmit Beamforming and Detection Design for
Uplink Multiuser MIMO Systems, “Fortieth Asilomar Conference on Signals, Systems and
Computers, pp. 1593-1600.
[13] Khachan, Ali M., Adam J. Tenenbaum, & Raviraj S. Adve. (2006). “Linear processing for the
downlink in multiuser MIMO systems with multiple data streams,” IEEE International Conference on
Communications,pp4113-4118.
[14] Xiao, Pei, & Mathini Sellathurai. (2010). “Improved linear transmit processing for single-user and
multi-user MIMO communications systems.” IEEE Transactions on Signal Processing, Vol.58, No. 3,
pp1768-1779.
[15] Raja, M., Muthuchidambaranathan.P, & Ha H. Nguyen (2013). “Transceiver Design for MIMO
Systems with Improper Modulations,” Wireless Personal Communications, Vol. 68, pp265–280.
[16] Raja, M. & Muthuchidambaranathan.P, (2012). “Joint Precoding and Decoding in MU-MIMO
Downlink Systems with Perfect Channel State Information (CSI).” Procedia Technology, Vol.6,
pp708-715.
[17] Raja, M. & Muthuchidambaranathan.P, (2014). “Multiuser MIMO Transceiver Design for Uplink and
Downlink with Imperfect CSI,” Wireless Personal Communications, Vol. 75, pp1215–1234.
[18] Raja, M. & Muthuchidambaranathan.P, (2013). “SVD-Assisted Joint Precoder and Decoder Design
for the Uplink of MU-MIMO Systems With Improper Modulation.” Wireless Personal
Communications, Vol. 73, No. 3, pp1129–1142.
[19] C. Manikandan, P. Neelamegam, A. Srivishnu & B. Sabari Ganesh, (2015). “Transceiver Design for
SU-MIMO System with Improper Modulations Using per Antenna Power Constraint and Perfect
CSI,” International Journal of Mathematical Analysis, Vol. 9, No. 31, pp1519 - 1528.
[20] Andrew Rukhin, Juan Soto, James Nechvatal, Miles Smid, Elaine Barker, Stefan Leigh, Mark
Levenson, Mark Vangel, David Banks, Alan Heckert,&James Dray (2010). “A Statistical Test Suite
for Random and Pseudorandom Number Generators for Cryptographic Applications” National
Institute of Standards and Technology Special Publication 800-22 revision 1a, pp 1 – 131.
AUTHORS
C.Manikandan received his B.E. Degree in Electronics and Communication Engineering
from Anna University and M.Tech. Degree in Digital Communication and Networking
from SRM University, Currently, he is working as an Assistant Professor-III in School of
Electrical and Electronics Engineering, SASTRA, Deemed University. He has a teaching
experience of 9 years, and he has published 18 research articles in refereed International
Journals He is currently working towards his Ph.D. Degree in SASTRA, Deemed
University. His research area is improving security and reliability for Wireless & Visual
MIMO Communication System. He is a Life Member of ISSE.
P.Neelamegam received his M.Sc. Degree from Madras University, Chennai and Ph.D.
Degree from Bharathidasan University, Trichy. Currently, he is working as a Professor in
the School of Electrical and Electronics Engineering, SASTRA, Deemed University,
Thanjavur. He has teaching experience of 45 years, and he has published 80+ research
articles in refereed International journals His research is related to Signal Processing,
Smart Instrumentation, Wireless embedded system, and Networking. He has guided 8
Ph.D. students. He is a Life member of ISSE.

32
S.Rakesh Kumar received his B.E and M.E in the field of instrumentation engineering.
He completed his Ph.D. in the area of sensor fusion techniques to improve autonomous
navigation of wheeled mobile robots. He has nine years of teaching experience and
currently working as an assistant professor in the Department of instrumentation
engineering, School of electrical and electronics, SASTRA Deemed University, India.
He has around 35 research articles published in peer-reviewed international journals.
His area of interest includes sensor fusion, robotics and automation, and machine
learning techniques.
G.Venkat Babu received his B.E. Degree in Electronics and Communication Engineering
from Bharathidasan University, Trichy, M.E. Degree in Microwave communication and
RADAR from Dr.B.R.Ambedkar University, Agra and Ph.D. Degree from Bhagwant
University, Ajmer. Currently, he is working as an Assistant Professor in School of
Electrical and Electronics Engineering, SASTRA, Deemed University, Thanjavur. He has
a teaching experience of 16 years, and he has published seven research articles in
National and International journals. He Co-authored a book in the field of
Electromagnetic Fields. His research area includes Smart Antennas, and RF-MEMS. He
is a member of IEEE, Life member of IETE and ISTE.
Sai Siva Satwik Kommi received his Degree in Electronics and Communication
Engineering from SASTRA Deemed University, Thanjavur India. He is pursuing masters
in Computer and Information Systems Security from Illinois Institute of Technology,
Chicago. He has around four research articles published in peer-reviewed international
journals and Conferences His research area includes Information Security, Image
Steganography and Wireless Communication.

DESIGN OF SECURE AND RELIABLE MU-MIMO TRANSCEIVER SYSTEM FOR VEHICULAR NETWORKS

More Related Content

What's hot (20)

Similar to DESIGN OF SECURE AND RELIABLE MU-MIMO TRANSCEIVER SYSTEM FOR VEHICULAR NETWORKS (20)

More from IJCNCJournal (20)

Recently uploaded (20)

DESIGN OF SECURE AND RELIABLE MU-MIMO TRANSCEIVER SYSTEM FOR VEHICULAR NETWORKS