sensors

Article
The Physical Layer Security Experiments of
Cooperative Communication System with
Different Relay Behaviors
Yishan Su 1,2 , Guangyao Han 2 , Xiaomei Fu 1,2, *, Naishen Xu 2 and Zhigang Jin 2
1 School of Marine Science and Technology, Tianjin University, 300072 Tianjin, China; yishan.su@tju.edu.cn
2 School of Electrical Automation and Information Engineering, Tianjin University, 300072 Tianjin, China;
hanguangyao@tju.edu.cn (G.H.); xunaishen@126.com (N.X.); zgjin@tju.edu.cn (Z.J.)
* Correspondence: fuxiaomei@tju.edu.cn

Academic Editor: Yuh-Shyan Chen

Received: 10 January 2017; Accepted: 31 March 2017; Published: 6 April 2017

Abstract: Physical layer security is an attractive security mechanism, which exploits the randomness
characteristics of wireless transmission channel to achieve security. However, it is hampered by
the limitation of the channel condition that the main channel must be better than the eavesdropper
channel. To alleviate the limitation, cooperative communication is introduced. Few studies have
investigated the physical layer security of the relay transmission model. In this paper, we performed
some experiments to evaluate the physical layer security of a cooperative communication system, with
a relay operating in decode-and-forward (DF) cooperative mode, selfish and malicious behavior in real
non-ideal transmission environment. Security performance is evaluated in terms of the probability
of non-zero secrecy capacity. Experiments showed some different results compared to theoretical
simulation: (1) to achieve the maximum secrecy capacity, the optimal relay power according to the
experiments result is larger than that of ideal theoretical results under both cooperative and selfish
behavior relay; (2) the relay in malicious behavior who forwards noise to deteriorate the main channel
may deteriorate the eavesdropper channel more seriously than the main channel; (3) the optimal relay
positions under cooperative and selfish behavior relay cases are both located near the destination
because of non-ideal transmission.

Keywords: physical layer security; different relay behavior; experiment

1. Introduction
When it comes to the secure transmission of information, cryptographic algorithms are traditionally
employed as the security technology to achieve information confidentiality [1]. Information encryption
usually involves the use of a key to encrypt and decrypt the messages. Unlike conventional wired
communication that uses wired media (optical cable, copper, etc.) for data transmission, the broadcast
nature of wireless channels allows an illegal receiver to wiretap legal messages easily during the
transmission from the source to the destination. Moreover, with the development of cryptography, new
algorithms and methods to decipher traditionally encrypted messages have been developed, which
means old encryption methods are faced with higher risk of being cracked. Cryptographic methods
face numerous risks due to the rapid development of the computing capabilities of eavesdroppers [2].
Besides, complex cryptographic algorithms require devices capable of dealing with heavy computing
loads, leading to high power and resource costs. The physical layer security adopts the physical
characteristics of wireless channels to guarantee its security through coding and signal processing,
in which confidential message can be decoded only by their legitimate receiver and avoid complex

Sensors 2017, 17, 781; doi:10.3390/s17040781 www.mdpi.com/journal/sensors

Sensors 2017, 17, 781 2 of 12

computing. In this way, physical layer security has been proposed to safeguard wireless transmission,
for instance, 5G wireless communication networks and heterogeneous cellular networks [3,4].
Physical layer security was first introduced by Wyner [5], who adapted Shannon’s concepts
of entropy and equivocation [6] to investigate the physical layer security. The achieved maximum
transmission rate of confidential message communication is defined by Wyner as the secrecy capacity.
However, the feasibility of physical layer security is hampered by the limitation when the channel
between the source and the destination (the main channel) is worse than that between the source and
the eavesdropper (the wiretap channel). An effective way to improve the main channel condition is
cooperative communication with a relay’s help [7].
Cooperative relays can enhance the secrecy capacity of wireless communications by exploiting
decode and forward (DF), amplify and forward (AF) [8]. In the DF and AF cooperative strategies,
the relay utilizes its overheard information to reinforce the transmitting signal of the source [9–11].
In [12–14], the physical layer security is improved by the advantages of large scale MIMO arrays of
relays with AF and DF strategies. In [15,16], the optimal transmission with maximized secrecy rate
under energy harvesting requirements is investigated, where the secrecy capacity is sensitive to the
position among the nodes. It was found that the distance from the source to the relay has an important
influence on secrecy capacity. Therefore, to find the optimal relay location is an important requirement
to achieve maximum secrecy capacity [17]. In this paper, we investigated a cooperative communication
system with one source, one relay, one destination and one eavesdropper with experiments for
further discussion.
In practical applications, relays may not always select a cooperative strategy because of its own
benefits. Some research has investigated the secrecy capacity with untrusted relays. In [18,19] the
authors discussed the case of untrusted relays with both a cooperator and an eavesdropper. The secrecy
capacity in a selfish behavior relay scheme in which the relay wants to save its own resources [20,21]
was also discussed.
Theoretical analysis usually guides the real implementation with simplified or ideal physical
factors, while applying them in practice may lead to different results. The authors in [22] firstly
analyzed the physical layer security with the probability of a non-zero secrecy capacity (P(CS > 0))
of a wiretap channel by experimental measurements. It was found that the shadowing deviation can
influence the security, which was seldom considered by theoretical analysis. The results from [23]
can only be applied in a point-to-point network scenario. To investigate the physical layer security of
multiple point relay transmission models, in this paper, we establish a cooperative communication
testbed based on a software-defined radio (SDR) platform [24]. Based on this platform, we investigate
the physical layer security with different relay behavior (cooperation, selfishness, maliciousness) with
DF scheme in real non-ideal wireless transmission environment. The P(CS > 0) was calculated to
evaluate physical layer security in an actual outdoor environment. We obtain some different results
which is ignored by theoretical or simulation analysis.

Table 1. Symbols Description.

hSD the channel fading coefficients from S to D (source to destination)

hSR the channel fading coefficients from S to R (source to relay)
h RD the channel fading coefficients from R to D (relay to destination)
Po f f the probability in the Off-State
P(CS > 0) the probability of Non-zero secrecy capacity
γM the average SNR of the main channel
γW the average SNR of the wiretap channel
γSD the average SNR of the S to D channel (source to destination)
γSRD the average SNR of the main channel which consists of channel from S to R and R to D
γSE the average SNR of the S to E path
γSRE the average SNR of the wiretap channel which consists of channel from S to R, and R to E
PSR the BER at S-R hop
PRD the BER at R-D hop
PSRD the BER of the two-hop S-R-D channel
Sensors 2017, 17, 781 3 of 12
Sensors 2017, 17, 781 3 of 12

The
The paper
paper is
is structured
structured as
as follows:
follows: Section
Section 22 introduces
introduces the
the system
systemmodel
modeland
anddifferent
differentrelays’
relays’
behavior;
behavior; Section 3 presents experimental results of bit error rate (BER) in a system withwith
Section 3 presents experimental results of bit error rate (BER) in a system different
different relay
relay behavior; Section 4 discusses the security performance under different relay behavior; finally,
behavior; Section 4 discusses the security performance under different relay behavior; finally, Section 5
Section
presents5 our
presents our conclusions.
conclusions. The symbolsTheused
symbols used
in this in this
paper arepaper are described
described in Table 1.in Table 1.

2.
2. System
System Description
Description

System Model
2.1. System Model
A relay transmission system model consists of a source (S), a destination (D), a relay (R) and an
eavesdropper (E), as depicted
depicted in
in Figure
Figure 1.
1. SS wants
wantstotocommunicate
communicatewith
withDDininthe
thepresence
presenceofofE.E.The
TheS
Sbroadcasts
broadcastssignals
signalswith
withaverage
average power
powerPS . R. R
receives signals
receives from
signals S with
from DFDF
S with strategy.
strategy.

Figure 1. Relay transmission model with with one

one eavesdropper,
eavesdropper,where wherehSDℎ , , hℎSR ,, ℎh RD denote
denote the
the channel
channel
hSDℎ ∼~CN (0, 2 ),,


fading coefficients from
from SS to
toD,
D,R,R,and
andfrom
fromR RtotoD.D. They
They areare modeled
modeled as follows:
as follows: 0, σSD
2
(0, 0, σ
ℎhSR~ ∼ CN ),SR
2ℎ , h ~RD (0,
∼ CN ),0,with
2 = σSD2 |ℎ = |E {|
, hSD |= }, σ|ℎ |
SR = E,{| hSR =
2 |2 }, σ|ℎ
2 | E. {|h RD |2 }.



σRD , with RD =

2.2.
2.2. Different
Different Behaviors
Behaviors of
of Relay
Relay
In
In this
this paper, three types
paper, three types of
of relay
relay behaviors
behaviors are
are discussed:
discussed:
1. Cooperative Behavior: The relay decodes the received signal and forwards it to the destination.
1. Cooperative Behavior: The relay decodes the received signal and forwards it to the destination.
2. Selfish Behavior: The relay refuses to cooperate with probability [14], while selecting to
2. Selfish Behavior: The relay refuses to cooperate with probability Po f f [14], while selecting to
cooperate with probability ( = 1 − ).
3. cooperate Behavior:
Malicious with probability
The relay Pon = 1 −random
Pon forwards Po f f . noise instead of the received signals of the
3. source.
Malicious Behavior: The relay forwards random noise instead of the received signals of the source.

The Probability
2.3. The Probability of Non-Zero Secrecy Capacity with Different Behavior
To investigate the physical layer security in experiment, we adopted the theories and equations
in [8]
[8]with
withthe wiretap
the communication
wiretap communication model modelwithout relay, as
without shown
relay, as in Figurein2. Figure
shown Then, we2. investigate
Then, we
the scenariothe
investigate ofscenario
a relay cooperative case. Suppose
of a relay cooperative a k-length
case. Suppose a k-length mk sent fromsent
messagemessage source
from is source
coded
to coded
is n-length codeword
to n-length codeword (2)[, .(1),
x n ( x n = [ x (1(), x= . . , x ((2), , ( )]),from
n)]),…chosen codebook
chosen X n ( x n ∈ X n (), the
from codebook ∈ main),
channel
the mainand the wiretap
channel and thechannel
wiretapoutput
channel is output
described as:
is described as:

y M ((i )) =
= hℎ ((i ))x ((i )) +
SD
+ n ,,
SD
(1)
(1)
(2)
yW ((i )) =
= hℎ ((i ))x ((i )) +
SE
+ n ,,
SE (2)
where ℎ , ℎ denote the channel fading coefficients of the main channel and wiretap channel
where hSD , hSE denote the channel fading coefficients of the main channel and wiretap channel
respectively with complex Gaussian distribution. Notations , denote the mean circularly
respectively with complex Gaussian distribution. Notations nSD , nSE denote the mean circularly
symmetric complex Gaussian noise in the main channel and wiretap channel respectively, and
symmetric complex Gaussian noise in the main channel and wiretap channel respectively, and nSD ∼
~ (0, ), ~ (0, ).
CN (0, NM ), nSE ∼ CN (0, NW ).
Sensors 2017, 17, 781 4 of 12
Sensors 2017, 17, 781 4 of 12

Figure 2. Wiretap communication model without relay.

Figure 2. Wiretap communication model without relay.
The channel power is limited by average transmit signal power P:
The channel power is limited by average transmit signal power P:
1
[| ( )| ] ≤ . (3)
1 n
n i∑
E[| x (i )|2 ] ≤ P. (3)
The average SNR of the main channel is
=1 given by:
= × [|ℎ | ]⁄ . (4)
The average SNR of the main channel is given by:
Similarly, the average SNR of the wiretap channel is described as:
γ M = P × E[| h |2 ]/N .
[|ℎSD| ]⁄ M
(4)
= × . (5)
The secrecy
Similarly, capacitySNR
the average in this
of model is defined
the wiretap as
channel = described
is − : as:
|ℎ | |ℎ |
− E[| h 1 |+
1 +γ = P × 2 , >
= W SE ] /NW . (6) (5)
, ≤
0
The secrecy capacity in this model is defined as CS = C M − CW :
We denote the instantaneous SNR of the main channel and wiretap channel by γ ′ and γ ,
[25]:     
|hSD |2 P |hSE |2 P
 log 1 + − log 1 + , i f γ M > γW
CS = NM′ = × [|ℎ ( )| ]N
⁄W , (7) (6)
, i f γm ≤ γW
0

′= × [|ℎ ( )| ]⁄ , (8)
from which the
We denote can see that γ SNR
we instantaneous (regardless of thechannel
of the main footnoteand
in different
wiretap channels) is γ
channel by 0
exponentially0
M and γW , [25]:
distributed as ∝ |ℎ| . Then:
γ M 0 = P × E1[|hSD (i )|2 ]/NM , (7)
( )= − , (9)

W by γ 0
hSD (i )|2 ]/NW ,
= P (× E>[|0)) (8)
and the probability of C > 0(denoted is calculated in [8], specifically:
from which we can see that γ (regardless of the footnote in different channels) is exponentially
distributed as γ ∝ |h|2 . Then:( > 0) = ( ) ( )
 
=
+
. (10)
1 γ
p(γ) = exp − , (9)
Then, we derive ( > 0) in relay transmission
γ γ for different relay behaviors:
model
and the
(1) probability of CS >
( > 0) under 0 (denoted
cooperative behavior: > 0relay
by P(CSThe )) is calculated in [8],
cooperatively specifically:
decodes and forwards the
received signal to D (or E).ZTheZdestination combines the received signals from S and R by
∞ γM γ M path and S-D path,
maximum ratioP(Ccombining
S > 0) =
(MRC). Thep(main
γ M ) pchannel
(γW )dγconsists
W dγ M =
of the S-R-D . (10)
0
and the wiretap channel includes 0 the S-E and S-R-E path. = M+
γ + γW is the SNR of the
main channel, and = + is the SNR of the eavesdropper channel. Therefore,
Then, we derive P(CS > 0() in>relay
according to (10), the
transmission model for different relay behaviors:
0) is given by:
(1) P(CS > 0) under cooperative behavior: The relay + cooperatively decodes and forwards the
( > 0) = , (11)
received signal to D (or E). The destination+combines + +
the received signals from S and R by
maximum
where ratio combining
is the (MRC).
average SNR of S The
to D main channelisconsists
channel, of the
the average SNRS-R-D
of thepath
mainand S-D path,
channel
andwhich
the wiretap
consists channel
of S to R, includes
and R to Dthe S-EThe
path. and S-R-ESNR
average path.
of the = Eγpath
γ MS to SRD +
is denoted
γ SD is the
as SNR
, of
the main channel, and γW = γSRE + γSE is the SNR of the eavesdropper channel. Therefore,
and is the average SNR of the wiretap channel which consists of S to R, and R to E path.
according to (10), the P(CS > 0) is given by:

γSRD + γSD
PC (CS > 0) = , (11)
γSRD + γSD + γSRE + γSE
Sensors 2017, 17, 781 5 of 12

where γSD is the average SNR of S to D channel, γSRD is the average SNR of the main channel
which consists of S to R, and R to D path. The average SNR of the S to E path is denoted as γSE ,
and γSRE is the average SNR of the wiretap channel which consists of S to R, and R to E path.
(2) P(CS > 0) under selfish behavior: If the relay is selfish with the probability Po f f , then, γ M = γSD
and γW = γSE , while the relay uses cooperative behavior with the probability (1 − Po f f ). Then,
P(CS > 0) is given by:
Sensors 2017, 17, 781 5 of 12
  γ + γSD γSD
(2) ( > 0) S > 0selfish
PS (Cunder ) = 1behavior: ×the relay isSRD
− Po f f If selfish with the
+ Po f f × , then,
probability
.
γ =
(12)
γSRD + γSD + γSRE + γSE γSD + γSE
and = γ , while the relay uses cooperative behavior with the probability (1 − ). Then,
(3) ( > 0) is given by:
P(CS > 0) under malicious behavior: The received signal from the malicious relay at D and E is
considered as a noise. γ M = γSD /(1 + γRD ) and+γW = γSE (1 + γRE ) , P(Cs > 0) is given by:
( > 0) = 1 − × + × . (12)
+ + + +
SD γ (1 + γ
RE )
(3) ( > 0) under maliciousPM (Cbehavior:
S > 0) =The received signal from the malicious. relay at D and E is (13)
considered as a noise. = /(1 +
γ (
) and
SD 1 + γ =
RE ) + γ
(1 +
SE (1 + γRD ) > 0) is given by:
) , P(Cs
(1 + )
2.4. Equivalent Signal-To-Noise Ratio( from
> 0) =
Source to
(1the
+
Destination
)+ (1 + )
. (13)

In theoretical analyses, most of the existing work considers the forward process as completely
ideal 2.4.
in the DF strategy
Equivalent [10,11,24],
Signal-To-Noise infrom
Ratio which thetorelay
Source correctly decodes and forwards the received
the Destination
messages.InHowever,
theoretical analyses, most of the existing work considersscenario.
A few researchers consider this a non-ideal the forwardDue to possible
process errors at the
as completely
relay,ideal
the S-R-D
in the channel is clearly
DF strategy non-linear
[10,11,24], in which and
thenon-Gaussian. However,
relay correctly decodes andone can think
forwards the of the BER in
received
S-R-Dmessages.
channel However,
as the error probability
A few at the
researchers receiver
consider thisof an equivalent
a non-ideal one-hop
scenario. Due toAWGN
possiblelink. The
errors at BER
of thethe relay, the
two-hop S-R-D
S-R-D channelisisgiven
channel clearlybynon-linear
[26]: and non-Gaussian. However, one can think of the
BER in S-R-D channel as the error probability at the receiver of an equivalent one-hop AWGN link.
The BER of the two-hop S-R-D = (1 −isPgiven
PSRDchannel SR ) ×byPRD + (1 − PRD ) × PSR
[26]: (14)
= (1 − )× + (1 − )× (14)
where PSR and PRD are the BER at source to relay (S-R) hop and relay to destination (R-D) hops,
where The
respectively. are the
andequivalent BERfrom
SNR at source to relay (S-R)
the measured BERhop
is and relayto
applied todescribe
destination (R-D)
this hops, DF
non-ideal
respectively. The equivalent SNR from the measured BER is applied to describe this non-ideal DF
transmission scenario. Then the equivalent SNR is obtained.
transmission scenario. Then the equivalent SNR is obtained.
3. Experimental Setup
3. Experimental Setup
Software defined
Software radio
defined (SDR)
radio (SDR)isisa arapidly
rapidly emerging conceptinin
emerging concept wireless
wireless communication
communication system.
system.
The main SDRSDR
The main involves using
involves GNU
using GNURadio Radioandand universal softwareradio
universal software radio peripheral
peripheral (USRP)
(USRP) in anin an
Ubuntu system
Ubuntu [18].[18].
system TheThe
experimental
experimental setup
setupisisshown
showninin Figure 3, where
Figure 3, whereeach
eachnode
node
is is equipped
equipped with a
with
laptop running
a laptop GNUGNU
running Radio connected
Radio connected to atoUSRP
a USRP inin
the
theexperiment.
experiment.

Figure3.3.Experimental
Figure Experimental setup.
setup.

Metal shields are placed between nodes to help attenuate any line-of-sight path. The distance
between every two points of S, R and D is 2 m. The noise power is measured about −60 dBm. The
source power is fixed on −39 dBm and relay power changes from −51 dBm to −39 dBm (all the power
presented in dBm is measured from the experiment). Figure 4 illustrates the different steps of signal
processing at the transmitter and relay designed by the authors.
Sensors 2017, 17, 781 6 of 12

Metal shields are placed between nodes to help attenuate any line-of-sight path. The distance
between every two points of S, R and D is 2 m. The noise power is measured about −60 dBm.
The source power is fixed on −39 dBm and relay power changes from −51 dBm to −39 dBm (all the
power presented in dBm is measured from the experiment). Figure 4 illustrates the different steps of
Sensors 2017, 17, 781 6 of 12
signal processing at the transmitter and relay designed by the authors.

Figure 4.
Figure Signal processing
4. Signal processing at
at each
each node.
node.

3.1. Source Operations

3.1. Source Operations
We
We use
useaaUSRP
USRPsink
sinkininthe
theGNU
GNU Radio platform
Radio platformto transmit
to transmita binary bit stream.
a binary Messages
bit stream. are first
Messages are
encoded in the laptop to baseband signals. Then the USRP sink broadcasts the signal at
first encoded in the laptop to baseband signals. Then the USRP sink broadcasts the signal at 2.45 GHz 2.45 GHz with
GMSK. A Python
with GMSK. program
A Python is used
program is to process
used the signals.
to process We transmit
the signals. 12,144 12,144
We transmit bits each
bitstime
eachand repeat
time and
every situation 25 times to exclude an accident case. The signal’s power is defined in
repeat every situation 25 times to exclude an accident case. The signal’s power is defined in the the software part
by limiting
software theby
part signal amplitude
limiting to aamplitude
the signal certain voltage.
to a certain voltage.

3.2. Relay Operations

3.2. Relay Operations
In this paper,
In this paper, we
we experiment
experiment onon different
different relay
relay strategy,
strategy, forward
forward power,
power, and
and locations
locations to discuss
to discuss
the
the influence on secrecy
influence on secrecy capacity.
capacity. In
In general,
general, the
the main
main task
task of
of aa relay
relay is
is to
to receive
receive the
the signal
signal from
from the
the
source, decode and forward them to the destination. After received desired signal,
source, decode and forward them to the destination. After received desired signal, it savesit saves the complex
the
signal
complexforsignal
further
forprocessing and finally
further processing andtransmit
finally processed signal in asignal
transmit processed certain
in power.
a certain power.
3.3. Data Receiving and Processing
3.3. Data Receiving and Processing
The USRP both in destination and eavesdropper node operates in the same way. They monitor at
The USRP both in destination and eavesdropper node operates in the same way. They monitor
the frequency of 2.45 GHz and save received complex signals to the laptop. Thanks to the extensive
at the frequency of 2.45 GHz and save received complex signals to the laptop. Thanks to the extensive
GNU Radio platform and the stable USRP hardware, we have little worries about how to generate
GNU Radio platform and the stable USRP hardware, we have little worries about how to generate a
a baseband bit stream and shifting it to a high frequency is much more easy and there is no carrier
baseband bit stream and shifting it to a high frequency is much more easy and there is no carrier
frequency offset problem compared to a traditional transmission device. The difficult part in the
frequency offset problem compared to a traditional transmission device. The difficult part in the
experiment however is processing the received data. The data are all saved in a complex signal form
experiment however is processing the received data. The data are all saved in a complex signal form
which can be read from MATLAB. We have to determine which part of the data is the desired signal
which can be read from MATLAB. We have to determine which part of the data is the desired signal
and leave out the white noise. Besides, complex signals received at the destination and eavesdropper
and leave out the white noise. Besides, complex signals received at the destination and eavesdropper
consists of the two parts received from the source and relay, respectively. Synchronization is required
consists of the two parts received from the source and relay, respectively. Synchronization is required
to combine these two parts and after synchronization, two different parts of the complex signal are
to combine these two parts and after synchronization, two different parts of the complex signal are
combined using the maximum ratio combining (MRC) method.
combined using the maximum ratio combining (MRC) method.
4. The Physical Layer Security Performance of the Relay Transmission Model
4. The Physical Layer Security Performance of the Relay Transmission Model
Considering the model depicted in Figure 5, S, D and E are located at fixed two-dimensional
Considering the model depicted in Figure 5, S, D and E are located at fixed two-dimensional
normalized coordinates (0; 0), (1; 0) and (0; 1). R is located along a horizontal line with S and D.
normalized coordinates (0; 0), (1; 0) and (0; 1). R is located along a horizontal line with S and D. We
execute the experiments and simulate the theoretical formulas according to Equations (11)–(13) for
comparison.
Sensors 2017, 17, 781 7 of 12
Sensors 2017, 17, 781 7 of 12

We execute the experiments and simulate the theoretical formulas according to Equations (11)–(13)
Sensors 2017, 17, 781 7 of 12
for comparison.

Figure 5. The system model.

Figure 5. The system model.

4.1. The Influence of the Relay Power forFigure
P(Cs >5.0)The system model.

4.1.
4.1. The
4.1.1. Influence of
Simulation
The Influence of the Relay
Results
the Relay Power
with for
for P(Cs
Various
Power >
Relay
P(Cs 0)
Power
> 0)
4.1.1.Figure 5 is aResults
Simulation typicalwith
cooperative relay Power
Various Relay communication model adopted from [7], with source,
4.1.1. Simulation Results with Various Relay Power
destination and eavesdropper nodes fixed and relay node’s location changing between the source
Figure 5 is a typical cooperative relay communication model adopted from [7], with source,
and Figure 5 is aThe
destination. typical cooperative
difference in thisrelay communication
paper is that the relay model hasadopted
differentfrom
relay[7],behaviors.
with source,
We
destination and eavesdropper nodes fixed and relay node’s location changing between the source and
simulate theand(eavesdropper
destination > 0) versesnodes / fixed
under and relay node’s
different relay location
behaviors, changing
where between and the aresource
the
destination. The difference in this paper is that the relay has different relay behaviors. We simulate the
and
power destination.
in relay and Thesource,
difference in this paper
respectively, and theis SNR
that the relay
of the sourcehas todifferent relay(S-D)
destination behaviors.
channelWeis
P(CS > 0) verses Pr /Ps under different relay behaviors, where Pr and Ps are the power in relay and
simulate the loss
5dB. The path ( in > 0)
the verses
simulation / is under
1/ , where
different isrelay behaviors,
the distance wheretwo nodes.
between and are the
source, respectively, and the SNR of the source to destination (S-D) channel is 5dB. The path loss in the
power Thein relay’s
relay and source,
normalized respectively,
location is and
fixed the
at SNR
(0.5; 0) of
(thethe source
middle of to
S destination
and D) and (S-D)
source channel
power is
simulation is 1/d2 , where d is the distance between two nodes.
5dB. Theatpath
is fixed −40 loss
dBm.inRelay power isis 1/
the simulation , where
changed from −70is the
dBmdistance
to −20between
dBm. Wetwo nodes.the ( >
simulate
The relay’s normalized location is fixed at (0.5; 0) (the middle of S and D) and source power PS is
0) under
The relay’s normalized
three different location
relay is fixed
behavior at (0.5; to
according 0) (the (12) ( of S=and
(11),middle 50%)D) and
and (13)
source as power
shown in
fixed at −40 dBm. Relay power Pr is changed from −70 dBm to −20 dBm. We simulate the P(CS > 0)
is fixed 6.
Figure atWe−40find
dBm. Relay
that power
there exists anisoptimal
changedrelay −70 dBm=to −20
frompower dBm. We
, where ( simulate
> 0) would ( >a
the reach
under three different relay behavior according to (11), (12) (Po f f = 50%) and (13) as shown in Figure 6.
0) under three different relay behavior according to
peak value in cooperative and selfish scenario. When the relay power (11), (12) ( = 50%) and
increases beyond in,
(13) as shown
We find that there exists an optimal relay power Pr = Ps , where P(CS > 0) would reach a peak value
the ( 6. >
Figure We 0)find that there
decreases. Weexists an optimal
also notice relay power
that increasing = , where
the malicious ( > 0)
relay power would
always reach a
decreases
in cooperative and selfish scenario. When the relay power Pr increases beyond Ps , the P(CS > 0)
the (value
peak > 0).
in cooperative and selfish scenario. When the relay power increases beyond ,
decreases. We also notice that increasing the malicious relay power always decreases the P(CS > 0).
the ( > 0) decreases. We also notice that increasing the malicious relay power always decreases
the ( > 0).

Figure 6. Simulation results: P(CS > 0) versus Pr /PS under different behavior.
Figure 6. Simulation results: ( > 0) versus / under different behavior.

4.1.2.
4.1.2. Experimental
Experimental Results
Results with
with Various
Various Relay Power
Figure 6. Simulation results: Relay
( > Power
0) versus / under different behavior.
In
In the experiment, S, R, D and E are locatedatatfixed
the experiment, S, R, D and E are located two-dimensional
fixed two-dimensionalcoordinates
coordinates(0; (0;
0), 0),
(1.8(1.8
m; m;
0),
4.1.2.
(3.6 Experimental
m; m;
0) and (0; (0; Results
3.63.6
m),m), with Various
respectively. InIn Relay
both DDPower
and
0), (3.6 0) and respectively. both andE Enode,
node,BER
BERofofboth
bothmain
mainchannel
channeland and wiretap
wiretap
In the
channel experiment,
are obtained byS, R,combining
D and E are thelocated at fixed
complex two-dimensional
signal from S and Rcoordinates (0; 0), (1.8 m;
using maximum-ratio-
0), (3.6 m; 0)(MRC),
combining and (0; 3.6 m), respectively.
which In both D and of
a weighted superposition E node, BER ofand
the source both main
relay channel
signals and wiretap
arriving at the
channel are obtained by combining the complex signal from S and R using maximum-ratio-
combining (MRC), which a weighted superposition of the source and relay signals arriving at the
Sensors 2017, 17, 781 8 of 12

channel are obtained by combining the complex signal from S and R using maximum-ratio-combining
Sensors 2017,
(MRC), which17, 781
a weighted superposition of the source and relay signals arriving at the destination 8 of 12
are
collected with the maximum available diversity order. We calculate P(CS > 0) by SNR according to
destination
(11), (12) andare collected
(13). The relay with the is
power maximum
changed available
about 3 dBmdiversity order.
every step andWe thecalculate
source power( >is 0) by
fixed
SNR according to (11), (12) and (13). The relay power is changed about
at −39 dBm. In malicious behavior case, pseudo random noise is forwarded by the relay, in which3dBm every step and the
source
there power
are is fixed
5972 error bitsatin-39dBm. In bits.
all 12,144 malicious behavior case, pseudo random noise is forwarded by
the relay, in which there are 5972
The experimental data of cooperative error bits in all 12,144
relay bits.
and malicious relay are shown in Tables 2 and 3
The experimental data of cooperative relay and
respectively, where PD is the BER at the destination and PE is the BER malicious relayatare
the shown in Tables
eavesdropper. In 2Table
and 2,3
γrespectively, where is the BER at the destination and is the BER at the eavesdropper. In Table
M = γSRD + γSD is the SNR of the main channel, and γW = γSRE + γSE is the SNR of the
2, = + is the SNR of the main channel,
eavesdropper channel. The last row of Table 2 is the calculated probability and = + of Cis > the0 (denoted
SNR of the by
S
eavesdropper channel. The last row of Table 2 is the calculated probability
P(CS > 0). In Table 3, γ M = γSD /(1 + γRD ) is the SNR of the main channel, and γW = γSE (1 + γRE of C > 0(denoted by)
( > 0). = /(1 + ) SNR of the main channel, and
is the SNR of the eavesdropper channel. The graphic experimental results are shown in Figure 7. is
In Table 3, is the = (1 + )
the SNR of the eavesdropper channel. The graphic experimental results are shown in Figure 7.
Table 2. Experimental results with different transmit power under cooperative relay.
Table 2. Experimental results with different transmit power under cooperative relay.
Transmit
Transmit −51 −49 −46.5 −44 −41.5 −39 −36 −34 −31 −28.5
Power (dBm) −51 −49 −46.5 −44 −41.5 −39 −36 −34 −31 −28.5
Power (dBm)
PD 0.3383
0.3383 0.2472
0.2472 0.1873
0.1873 0.113
0.113 0.061
0.061 0.0196
0.0196 0.0026
0.0026 0.0002
0.0002 0.0001
0.0001 0.0000
0.0000
γM 0.0974
0.0974 0.2614
0.2614 0.4415
0.4415 0.825
0.825 1.3386
1.3386 2.3814
2.3814 4.3727
4.3727 7.1942
7.1942 8.1577
8.1577 ∞ ∞
PE 0.3472 0.2639 0.2096 0.152 0.1118 0.0692 0.0226 0.0054 0.0005 0.0000
0.3472 0.2639 0.2096 0.152 0.1118 0.0692 0.0226 0.0054 0.0005 0.0000
γW 0.0864 0.2231 0.3652 0.593 0.8293 1.2289 2.2451 3.6327 6.1224 ∞
0.0864 0.2231 0.3652 0.593 0.8293 1.2289 2.2451 3.6327 6.1224 ∞
P(CS > 0) 0.5299 0.5394 0.5473 0.582 0.6175 0.6596 0.6608 0.6645 0.5713 0.5
( > 0) 0.5299 0.5394 0.5473 0.582 0.6175 0.6596 0.6608 0.6645 0.5713 0.5

Table 3. Experimental results with different transmit power under malicious relay.
Table 3. Experimental results with different transmit power under malicious relay.

Transmit
Transmit
−51−51 −49
−49 −46.5
−46.5 −44
−44 −
−41.5
41.5 −−39
39 −36
−36 −34
−34 −31−31 −28.5−28.5
Power
Power (dBm)
(dBm)
PD 0.3670
0.3670 0.3810
0.3810 0.4017
0.4017 0.4198
0.4198 0.4358
0.4358 0.4459
0.4459 0.4635
0.4635 0.4811
0.4811 0.4878 0.4918
0.4878 0.4918
γM 0.0647
0.0647 0.0513
0.0513 0.0343
0.0343 0.0229
0.0229 0.0146 0.0104
0.0104 0.0047
0.0047 0.0013
0.0013 0.0005 0.0002
0.0005 0.0002
PE 0.3609
0.3609 0.3725
0.3725 0.3902
0.3902 0.4104
0.4104 0.4251 0.4359
0.4359 0.4530
0.4530 0.4685
0.4685 0.4847 0.4918
0.4847 0.4918
γW 0.0710
0.0710 0.0592
0.0592 0.0416
0.0416 0.0287
0.0287 0.0200 0.0146
0.0146 0.0078
0.0078 0.0035
0.0035 0.0010
0.0010 0.0002
0.0002
( P(>C0)
S > 0) 0.4767
0.4767 0.4624
0.4624 0.4519
0.4519 0.4438
0.4438 0.4231 0.4155
0.4155 0.3761
0.3761 0.2652
0.2652 0.3333
0.3333 0.5 0.5

In our
In our experimental
experimental setup,
setup, the
the results
results from
from Figure
Figure 77 indicate
indicate that
that the
the optimal
optimal relay
relay power
power value
value
is larger
is larger than
than the
the sources
sourcespower
power(P / s=
( r /P = 5 dB),
5 dB), which
which means
means the relay need
the relay need more
more power
power toto achieve
achieve
the optimal security performance in real non ideal DF environment. Therefore, generally
the optimal security performance in real non ideal DF environment. Therefore, generally speaking, speaking,
in practice
in practice we
we can
can assume,
assume, the
the source
source toto destination(S-D)
destination(S-D) channel
channel isis not
not ideal,
ideal, the
the more
more relay
relay power
power
will be
will be needed
needed to to achieve
achievethe
thesame
sameresults
resultsasasthat
thatin
inthe
theideal
idealS-D
S-Dchannel.
channel.

Figure 7. Experimental
Figure 7. results: P((CS >
Experimental results: > 0) / s under
versus Pr /P
0) versus under different
different behavior.
behavior.

Moreover, in malicious relay behavior case, ( > 0) increases when / is beyond 5 dB,
which is different to theoretical simulation result. It indicates that when the malicious relay forwards
noise power high enough, the eavesdropper channel will be deteriorated more heavily than the main
Sensors 2017, 17, 781 9 of 12

Sensors 2017, 17, 781in malicious relay behavior case, P (CS > 0) increases when Pr /Ps is beyond 59 of
Moreover, dB,
12
which is different to theoretical simulation result. It indicates that when the malicious relay forwards
channel.
noise power As high
for the influence
enough, of relay power,
the eavesdropper it can will
channel be found that there more
be deteriorated existsheavily
a maximum
than thevalue
mainof
( > 0) for different relay powers.
channel. As for the influence of relay power, it can be found that there exists a maximum value of
P(CS > 0) for different relay powers.
4.2. Influence of Relay Location for P(CS > 0)
4.2. Influence of Relay Location for P(CS > 0)
4.2.1. Simulation Results with Various Relay Location
4.2.1. Simulation Results with Various Relay Location
We assume equals , and ( = / ) is the normalized distance from S to R. Let
We from
change assume 0 toPr1.equals
The simulation = dSRof/dSD
Ps , and x ( xresult ( ) is>the
0) normalized
is shown indistance
Figure 8.from S toexists
There R. Letanx optimal
change
relay location where ( > 0) would reach
from 0 to 1. The simulation result of P ( CS > 0 ) is shown in Figure 8. There exists an optimal
a peak value in both cooperative and selfish behavior relay
location where P(CS > 0) would reach a peak value in both cooperative and selfish behavior of a relay.
of a relay.

Figure8.
Figure 8. Simulation results: P((CS >
Simulation results: > 00) versusrelay
) versus relaylocation
locationunder
undervarious
variousbehaviors.
behaviors.

In the simulation = 5 is the SNR of S to D channel, the optimal location is in the middle of
In the simulation γSD = 5 is the SNR of S to D channel, the optimal location is in the middle of
source and destination ( = / = 0.5). ( > 0) will decreases quickly when R moves towards
source and destination (x = dSR /dSD = 0.5). P(CS > 0) will decreases quickly when R moves towards
D with a malicious behavior relay.
D with a malicious behavior relay.

4.2.2. Experimental
4.2.2. ExperimentalResults
Resultswith
withVarious
VariousRelay
RelayLocation
Location
Inthis
In thisexperiment,
experiment,
thethe transmit
transmit power
power of Sof S and
and R bothR both are at
are fixed fixed
−39at −39 S,
dBm. dBm. S, EDare
D and and E are
located
at fixed two-dimensional coordinates (0, 0), (3.6 m, 0) and (0, 3.6 m), respectively. The coordinate ofThe
located at fixed two-dimensional coordinates (0, 0), (3.6 m, 0) and (0, 3.6 m), respectively. R
coordinate of R changes from (1.2 m, 0), (1.8 m, 0) to (2.4 m, 0). The experimental results
changes from (1.2 m, 0), (1.8 m, 0) to (2.4 m, 0). The experimental results table are shown in Tables 4 table are
shown
and in Tables 4 The
5 respectively. and graphic
5 respectively. The graphic
experimental result experimental result9.is shown in Figure 9.
is shown in Figure

Table4.4.Experimental
Table Experimentalresults
resultswith
withvarious
variousrelay
relaylocations
locationsof
ofcooperative
cooperativebehavior.
behavior.

Relay Location Left (1.2 m, 0) Middle (1.8 m, 0) Right (2.4 m, 0)

Relay Location Left (1.2 m, 0) Middle (1.8 m, 0) Right (2.4 m, 0)
327.08 237.84 383.27
PD 327.08 237.84 383.27
2.081 2.3814 1.9336
γM 2.081 2.3814 1.9336
PE 625.71625.71 840.78
840.78 1174.15
1174.15
γW 1.488 1.488 1.2289
1.2289 0.9472
0.9472
(C0)
( P> S > 0) 0.58310.5831 0.6596
0.6596 0.6712
0.6712

ItItcan
canbe
beseen
seenthat
that the experimental P((CS >
the experimental > 00) valuein
) value inFigure
Figure99isisdifferent
differentfrom
fromthe
thetheoretical
theoretical
curvesininFigure
curves Figure8.8.The
Theoptimal
optimal location
location of of
thethe relay
relay is aisnearer
a nearer the destination
the destination in experiment
in experiment with with
non
non ideal DF case, whereas in theoretical ideal DF analysis, the location is in the middle of the Source
to destination link. That phenomenon is also appeared in selfish relay case. For malicious behavior,
the near destination malicious relay deteriorates the main channel seriously, but it is not as serious
as that in theoretical simulation.
Sensors 2017, 17, 781 10 of 12

ideal DF case, whereas in theoretical ideal DF analysis, the location is in the middle of the Source to
destination link. That phenomenon is also appeared in selfish relay case. For malicious behavior, the
near destination malicious relay deteriorates the main channel seriously, but it is not as serious as that
in theoretical
Sensors 2017, 17, simulation.
781 10 of 12

5. Experimental results
Table 5.
Table results with
with different
different relay
relay location
location of
ofmalicious
maliciousbehavior.
behavior.

Relay Location Left (1.2 m, 0) Middle (1.8 m, 0) Right (2.4 m, 0)

Relay Location Left (1.2 m, 0) Middle (1.8 m, 0) Right (2.4 m, 0)
0.441 0.4459 0.4535
PD 0.441 0.4459 0.4535
0.0123 0.0104 0.0076
γM 0.0123 0.0104 0.0076
0.4382 0.4359 0.4302
PE 0.4382 0.4359 0.4302
γW 0.0135
0.0135 0.0146
0.0146 0.0173
0.0173
( P>
(C0) 0.4763 0.4155 0.3063
S > 0) 0.4763 0.4155 0.3063

Figure 9.
Figure 9. Experimental results: P((CS >
Experimentalresults: > 0)
0) versus
versusrelay
relaylocation
locationunder
undervarious
variousbehavior.
behavior.

5.
5. Conclusions
In
In this
this paper,
paper, we
we implement
implement several
several experiments
experiments to to investigate
investigate the
the physical
physical layer
layer security
security of
of aa
relay transmission model with different relay behaviors. Experimental
relay transmission model with different relay behaviors. Experimental results are compared
compared with aa
with
theoretical
theoretical analysis. We find aa gap
gap between
between the
the ideal
ideal theoretical
theoretical simulations
simulations and
and the
the non-ideal
non-ideal real
real
experiments:
experiments:
(1) The relays
(1) relaysneed
needmoremore
power power to achieve
to achieve the maximum-security
the maximum-security performance performance in a real
in a real environment.
environment. In thethe
In the experiment, experiment,
optimal relaythe optimal
power isrelaylargerpower
thanisthelarger
sourcethan the source
power in both power in both
cooperative
cooperative and selfish
and selfish behavior relaybehavior
scenariosrelay scenarios
because of thebecause
non-idealof source
the non-ideal source (S-D)
to destination to destination
channel;
(2) (S-D)
In thechannel;
malicious behavior relay case, the experimental value of P(CS > 0) is different from
(2) In
thethe malicious
ideal behavior
theoretical relay case,
simulation resultthe experimental
when the relay value ( > than
powerofis larger 0) is the
different from
critical the
power
ideal theoretical simulation result when the relay power is larger than the
(Pr /PS = 5 dB), which indicates that when s malicious relay forwards high enough noise power, critical power ( / =
5 dB), which indicates that when s malicious relay forwards
the eavesdropper channel is deteriorated more seriously than the main channel; high enough noise power, the
eavesdropper channel in
(3) Optimal relay location is deteriorated
both cooperative moreandseriously
selfish than the main
behavior of relaychannel;
is in the middle of source
(3) Optimal relay location in both cooperative and selfish behavior
and destination (dSR /dSD = 0.5) in the ideal theoretical simulation. However, of relay is in theinmiddle of source
the experiment,
and destination
the optimal /
relay( location =
is 0.5) in the
nearer ideal
to the theoretical
destination simulation.
because However,source
of the non-ideal in the experiment,
to relay and
the optimal relay location
to destination (S-R-D) channel. is nearer to the destination because of the non-ideal source to relay
and to destination (S-R-D) channel.
Acknowledgments: This work is supported by the National Natural Science Foundation of China (61571323)
(61571318), and the Guangxi
Acknowledgments: Science
This work and Technology
is supported by the Project
NationalAC16380094.
Natural Science Foundation of China (61571323)
(61571318), and the Guangxi Science and Technology Project AC16380094.

Author Contributions: X.F. conceived and designed the study. Y.S., G.H. and N..X performed the experiments.
Z.J. gave some very important suggestions in the study. Y.S. and N.X. wrote the paper. X.F. and G.H. reviewed
and edited the manuscript. All authors read and approved the manuscript.

Conflicts of Interest: The authors declare no conflict of interest.

Sensors 2017, 17, 781 11 of 12

Author Contributions: X.F. conceived and designed the study. Y.S., G.H. and N..X performed the experiments.
Z.J. gave some very important suggestions in the study. Y.S. and N.X. wrote the paper. X.F. and G.H. reviewed
and edited the manuscript. All authors read and approved the manuscript.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Schneier, B. Description of a new variable-length key, 64-bit block cipher (Blowfish). In Proceedings of the
Fast Software Encryption: 7th International Workshop, FSE 2000, New York, NY, USA, 10–12 April 2000;
pp. 191–204.
2. Bruce, S. Applied Cryptography: Protocols, Algorithms, and Source Code in C; John Wiley & Sons, Inc.: New York,
NY, USA, 1996.
3. Yang, N.; Wang, L.; Geraci, G.; Elkashlan, M.; Yuan, J.; Di Renzo, M. Safeguarding 5G wireless communication
networks using physical layer security. IEEE Commun. Mag. 2015, 53, 20–27. [CrossRef]
4. Wang, H.M.; Zheng, T.X. Physical Layer Security in Heterogeneous Cellular Network. In Physical Layer
Security in Random Cellular Networks; Springer: Singapore, 2016; pp. 61–84.
5. Wyner, A.D. The Wire-Tap Channel. Bell Labs Techn. J. 1975, 54, 1355–1387. [CrossRef]
6. Shannon, C.E. Communication theory of secrecy systems. Bell Labs Techn. J. 1949, 28, 656–715. [CrossRef]
7. Lai, L.; El Gamal, H. The relay–eavesdropper channel: Cooperation for secrecy. IEEE Trans. Inf. Theory 2008,
54, 4005–4019. [CrossRef]
8. Dong, L.; Han, Z.; Petropulu, A.P.; Poor, H.V. Improving wireless physical layer security via cooperating
relays. IEEE Trans. Signal Process. 2010, 58, 1875–1888. [CrossRef]
9. Bassily, R.; Ulukus, S. Deaf cooperation and relay selection strategies for secure communication in multiple
relay networks. IEEE Trans. Signal Process. 2013, 61, 1544–1554. [CrossRef]
10. Bassily, R.; Ulukus, S. Deaf cooperation for secrecy with multiple antennas at the helper. IEEE Trans. Inf.
Forensics Secur. 2012, 7, 1855–1864. [CrossRef]
11. Bassily, R.; Ulukus, S. Secure communication in multiple relay networks through decode-and-forward
strategies. J. Commun. Netw. 2012, 14, 352–363. [CrossRef]
12. Chen, X.; Lei, L.; Zhang, H.; Yuen, C. Large-scale MIMO relaying techniques for physical layer security: AF
or DF? IEEE Trans. Wirel. Commun. 2015, 14, 5135–5146. [CrossRef]
13. Chen, X.; Zhong, C.; Yuen, C.; Chen, H.H. Multi-antenna relay aided wireless physical layer security.
IEEE Commun. Mag. 2015, 53, 40–46. [CrossRef]
14. Li, Q.; Yang, Y.; Ma, W.K.; Lin, M.; Ge, J.; Lin, J. Robust cooperative beamforming and artificial noise design
for physical-layer secrecy in AF multi-antenna multi-relay networks. IEEE Trans. Signal Process. 2015, 63,
206–220. [CrossRef]
15. Zhang, J.; Yuen, C.; Wen, C.K.; Jin, S.; Wong, K.K.; Zhu, H. Large system secrecy rate analysis for SWIPT
MIMO wiretap channels. IEEE Trans. Inf. Forensics Secur. 2016, 11, 74–85. [CrossRef]
16. Salem, A.; Hamdi, K.A.; Rabie, K.M. Physical layer security with RF energy harvesting in AF multi-antenna
relaying networks. IEEE Trans. Commun. 2016, 64, 3025–3038. [CrossRef]
17. Liu, Y.; Chen, H.H.; Wang, L. Physical layer security for next generation wireless networks: Theories,
Technologies, and Challenges. IEEE Commun. Surv. Tutor. 2016, 19, 347–376. [CrossRef]
18. Xiang, H.; Yener, A. Secrecy and reliable Byzantine detection in a Gaussian untrusted two-hop link.
In Proceedings of the 2010 IEEE Information Theory Workshop on Information Theory (ITW 2010), Cairo,
Egypt, 6–8 January 2010; pp. 1–5.
19. He, X.; Yener, A. Cooperation with an untrusted relay: A secrecy perspective. IEEE Trans. Inf. Theory 2010,
56, 3807–3827. [CrossRef]
20. Hou, L.; Fu, X. Physical layer security with dynamic behaviour cooperator based on coalitional game.
IET Commun. 2014, 8, 1258–1264. [CrossRef]
21. Fu, X.; Li, L.; Zong, Q. Bayesian Coalitional Game in Physical Layer Security. Wirel. Pers. Commun. 2015, 85,
1237–1250. [CrossRef]
22. Chrysikos, T.; Dagiuklas, T.; Kotsopoulos, S. Wireless information-theoretic security in an outdoor topology
with obstacles: Theoretical analysis and experimental measurements. EURASIP J. Wirel. Commun. Netw.
2011, 2011. [CrossRef]
Sensors 2017, 17, 781 12 of 12

23. Mitola, J. The software radio architecture. IEEE Commun. Mag. 1995, 33, 26–38. [CrossRef]
24. Zhang, R.; Comaniciu, C.; Poor, H.V. Outage capacity and partial secrecy for energy efficient physical layer
security in Gaussian fading channels. In Proceedings of the 2013 16th International Symposium on Wireless
Personal Multimedia Communications (WPMC), Atlantic City, NJ, USA, 24–27 June 2013; pp. 1–5.
25. Barros, J.; Rodrigues, M.R. Secrecy capacity of wireless channels. In Proceedings of the 2006 IEEE
International Symposium on Information Theory, Honolulu, HI, USA, 9–14 July 2006; pp. 356–360.
26. Wang, T.; Cano, A.; Giannakis, G.B.; Laneman, J.N. High-performance cooperative demodulation with
decode-and-forward relays. IEEE Trans. Commun. 2007, 55, 1427–1438. [CrossRef]

© 2017 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).