Distributed Spatial Modulation based Cooperative Diversity Scheme

International Journal of Wireless & Mobile Networks (IJWMN) Vol. 10, No. 2, April 2018
Distributed Spatial Modulation based Cooperative
Diversity Scheme
Quang-Trung Hoang
School of Information and Communication Technology
Thai Nguyen University, Vietnam
hqtrung@ictu.edu.vn
ABSTRACT: In this paper, a distributed spatial modulation based cooperative diversity scheme for relay
wireless networks is proposed. Where, the space-time block code is exploited to integrate with distributed
spatial modulation. Therefore, the interested transmission scheme achieves high diversity gain. By using
Monte-Carlo simulation based on computer, we showed that our proposed transmission scheme outperforms
state-of-the-art cooperative relaying schemes in terms bit error rate (BER) performance.
KEYWORDS: Space-Time Block Codes, Distributed STBC, Distributed Spatial Modulation, Cooperative
Communications, Cooperative Diversity Schemes.
1 Introduction
In the future, wireless networks are faced too many user demands for high rate and low delay
transmission. Especially, modern mobile devices are rapidly increasing over each year. This
requires next mobile communication systems as 5G to overcome challenges of technologies.
Among other solutions, cooperative communication technique is an expected solution, and
is considered more by the research community, recently.
With cooperative wireless networks, mobile users are not only able to make direct con-
nectivities, as device-to-device (D2D) communications in [1], but they also can use mediate
relays for collaboration to forwarding data from source to destination. Further, since the
cooperative communication technique is combined with some techniques as network coding
(NC) or space-time block code (STBC) and spatial modulation (SM), the performance of
wireless networks is significantly improved.
Distributed space-time block code (D-STBC) is an STBC technique applied for distributed
wireless networks to support diversity gain similar to the multi-input multi-output (MIMO)
systems. In reference [2], Bansal et al. showed the efficiency of Alamouti typed D-STBC
when it is used for cooperative relaying schemes. In addition, limitations on transmission
delay are also overcome by physical layer network coding [3]. This enables cooperative re-
laying networks to increase the high throughput in case there exists bidirectional traffic that
is necessary for widespread information exchange in future wireless services.
The last years, the spatial modulation, also known as a new technique, is interested in
more-and-more researchers. This method uses two constellations, one for signal modulation
and another one for selecting antennas to transmit modulated signals. In MIMO systems,
the technique SM is proved that it can highly increase the energy efficiency of a system
[4]. By exploiting capabilities of SM to apply for distributed wireless networks, the system
performance will be enhanced, and SM is then called DSM (Distributed SM).
DOI: 10.5121/ijwmn.2018.10203 31

In references [5]–[6], authors investigated cooperative transmission schemes based on DSM,
and showed that DSM enables to achieve high diversity gain being closed to the case of ones
without using DSM. Further, this technique can also reduce transmission energy. However,
researchers in [5]–[6] still omit some problems that are necessary to continue researching
more: The first, authors in [5] just consider in relaying transmission without cooperation
among relays or between a source and any relay to forward data, simultaneously. In this
case, the transmission scheme can achieve the gain of DSM but not spatial diversity gain.
The second, although the DSM-STBC protocol in [7] achieves both advantage of DSM and
diversity gain of distributed Alamouti-STBC, the performance of this protocol is signifi-
cantly affected by imperfect demodulation of two relays chosen.
To deal with the above problems, in this paper, we propose a DSM based cooperative
diversity scheme, so-called the DSM-CD scheme. Where, we consider in keeping both spa-
tial diversity and modulation gain for enhancing the system performance in terms BER.
By letting the source node cooperate with the relay node in data forwarding phase1, the
proposed scheme will obtain better spatial diversity gain than the protocol in [5]. Further,
thank using collaboration between the source node and one of the relay nodes instead of
transmission cooperation the among only relays, our DSM-CD scheme reduces the amount
of errors occurred by relays. The reason is that the source plays a role as one relay node in
the data forwarding phase (the second phase), so there is just one relay which must decode
the source’data, and a number of decoding errors occurred at relays are decreased. Then,
the destination node can decode data better than the case of protocol in [7] over considered
SNR ranges.
Our contributions include as follows:
– The DSM-CD scheme, which achieves two advantages: (i) the Alamouti distributed
STBC is integrated with DSM to enhance diversity gain; (ii) the cooperative mechanism
between the source node and the activated relay, which significantly decreases errors
occurred by the distributed spatial modulation, and the BER performance of the system
is improved.
– The Monte-Carlo simulations are used to show that our proposal outperforms state-of-
the-art cooperative diversity schemes.
The remainder of this paper is organized as follows. In Section 2, the considered system
model is introduced. Next, the proposed DSM-CD scheme is studied in Section 3. In Section
4, simulation results and evaluations are commented. Finally, Section 5 shows the conclusion
of this paper.
2 System model
For simply, the considered system model of cooperative wireless networks is shown as in
Figure 1. Where, it consists of one source (S), one destination (D) and two relays (Rk, with
k = 1, 2). Assume that each network node is equipped only one single antenna and works
in the half-duplex mode. The process of cooperative transmission is implemented in two
phases: (i) in the first phase, the source broadcasts its data to both the destination and
1
Cooperative transmission is implemented in two phases: the first phase, the source node broadcasts its
data to relays. In the second phase, relay nodes forward source’data to the destination node.
32

S D
R1
R2
S D
R1
R2
S D
R1
R2
dS =0
1RID 0
2RID 1
1RID 0
2RID 1
S D
R1
R2
1RID 0
2RID 1
Sd =1
1RID 0
2RID 1
R1 is activated. R2 is activated.
Fig. 1. The DSM-CD scheme with a source, a destination and two relays.
relays nodes; (ii) in the second phase, one of two relays collaborates with the source to for-
ward data block of Alamouti STBC to the destination node. With the interested model, the
distributed spatial modulation [5] is used at the relays, so the cooperative diversity scheme
can alternatively switch on between two relays. At each instant time, the activated relay
will participate cooperative transmission process while the rest one (being not activated)
keeps silent.
The channel between node X and Y is affected by Rayleigh fading, with a gain hXY . The
noise occurred at receiver of each network node is a complex Additive White Gaussian ran-
dom variable with zero mean and variance σ2
n = N0/2 per dimension.
Assume that, the source transmits Phase Shift Keying (PSK) or Quadrature Amplitude
Modulation (QAM) symbols belonging to constellation A with size M while the relays use
PSK/QAM modulation with constellation size N. Because DSM is used at the relays, value
M and N can be different. Where, the number of potential relays used for cooperative
transmission is Nr ≥ M. In case Nr = 2, as the considered model, the constellation size
is M = N = 2. In addition, each relay is assigned a unique identify IDRk
. With two
cooperative relays, k = 1, 2. For example, IDR1 = 0 and IDR2 = 1.
3 Proposed DSM-CD Scheme
In this section, we detail the operation of DSM-CD scheme. Where, we also use the dis-
tributed spatial modulation (DSM) proposed in [5]. However, the different point between
our scheme and [5] is that we let the source cooperate with the relay to transmit data
simultaneously in the relaying phase while the source using protocol proposed in [5] keeps
silent. By the way, our DSM-CD scheme does not only achieve high diversity but it also
decreases the computing complex of the demodulator at the destination node.
3.1 Broadcasting phase
Assuming the source needs to send three data symbols d1, d2 and d3 to the destination.
With Nr = 2, we use PSK modulation for both the source and relays, so M = N = 2. Then,
modulated symbols are s1, s2 and s3, respectively. In the broadcasting phase, the source
sequentially transmits these symbols to both the destination and relays. Therefore, signals
received at the relays are determined as follows:
ySRk
t = ES
1 hSRk
st + nRk
t , (1)
33

where, ySRk
t is signal received at the kth relay (k = 1, 2), at time-slots t = 1, 2, 3; ES
1 is the
source’s transmit energy during the first phase; hSRk
is the channel coefficient between the
source and kth relay, which is assumed to be constant over the transmission of s1, s2 and
s3; nRk
t is the Adaptive White Gauss Noise (AWGN) element.
Similarly, the signals received at the destination node in the broadcasting phase (the first
phase) are written by:
ySD1
t = ES
1 hSDst + nD1
t . (2)
The signal ySD1
t is used to provide information for demodulation process at the destination
node. In order to demodulate the source’s signals, the Maximal-Likelihood (ML) criterion
is used. Then, the demodulated signals are written as follows:
ˆsSRk
t = arg min˜s
SRk
t ∈A
ySRk
t − ES
1 hSRk
˜sSRk
t
2
(3)
Thus, the estimated data symbols
ˆ
dSRk
t are formulated by ˆdSRk
t = Φm ˆsSRk
t . Where, Φm (·)
is the PSK modulation - to - bit mapping function.
3.2 Relaying phase
After estimating the source’s data symbols, each kth relay will be activated to collaborate
with the source if its IDRk
and dSRk
1 is coincident. Then, a pair of this relay and the source
transmit Alamouti STBC formed by
ˆsSRa
2 s3
−ˆsSRa
3
∗
(s2)∗ , (4)
where, ˆsSRa
t , (t = 2, 3), is the activated relay’s modulation symbols estimated from the
source’s signals; s2 and s3 are the modulated signals of d2 and d3, respectively. Thus,
although the source plays a role as one relay but it just transmits the original data symbols
without the estimated ones. By this way, the signals received at the destination, during two
consecutive time slots of the relaying phase, are formulated as follows:
yD
1 = ERhRaDˆsSRa
2 + ES
2 hSDs3 + nD
1 (5)
yD
2 = ERhRaD −ˆsSRa
3
∗
+ ES
2 hSD (s2)∗
+ nD
2 (6)
Where, ER is the activated relay’s transmit energy and ES
2 is the source’s transmit energy
during the relaying phase.
3.3 Demodulator at The Destination
In this paper, we adopt the error-aware ML demodulator in [7] for detecting the source’s
data from the distributed Alamouti STBC received at the destination. However, because
one of two cooperating relays forwarding the source’s data is just the source, so the number
of errors occurred by the distributed spatial modulation (DSM) depends on only one rest
34

relay. The data symbols demodulated at the destination node are formulated by (7) and
(8). Where, ˆsSRa
t , t = {2, 3}, are the modulation symbols of the activated relay.
ˆsD
= arg max˜sD
t ∈A,t=1,2,3


˜sD
1 ∈A ˜sD
2 ∈A ˜sD
3 ∈A ˆs
SR1
1 ∈A ˆs
SR1
2 ∈A ˆs
SR1
3 ∈A ˆs
SR2
1 ∈A ˆs
SR2
2 ∈A ˆs
SR2
3 ∈A
Pr yD
1 , yD
2 | ˜sD
,ˆsSR
Pr ˜sD
,ˆsSR



(7)
Pr yD
1 , yD
2 | ˜sD
,ˆsSR
= exp




yD
1 −
√
ERhRaDˆsSRa
2 + ES
2 hSDs3
N0




× exp




yD
2 −
√
ERhRaD −ˆsSRa
3
∗
+ ES
2 hSD (s2)∗
N0



 (8)
Where, ˆsD
= sD
1 , sD
2 , sD
3 presents the estimated data vector of the source at the destina-
tion; ˜sD
= ˜sD
1 , ˜sD
2 , ˜sD
3 is the 1x3 trial data vector at the destination for the hypothesis
detection problem; ˆsSR
= [sSR1
1 , sSR1
2 , sSR1
3 , sSR2
1 , sSR2
2 , sSR2
3 ] denotes the 1x6 demodulated
data vector at the relays; Pr ˜sD
,ˆsSR
is the priori joint probability of the demodulation
errors at the relays. At high SNR values, the ML demodulator in (7) is rewritten as follows:
ˆsD
= arg min˜sD
t ∈A,t=1,2,3 Λhigh−snr
yD
1 , yD
2 ,˜sD
,ˆsSR
(9)
where, Λhigh−snr yD
1 , yD
2 ,˜sD
,ˆsSR
is determined as in the formulation (10).
Λhigh−snr
yD
1 , yD
2 ,˜sD
,ˆsSR
= Q + yD
1 − ERhRaDˆsSRa
2 + ES
2 hSDs3
+ yD
2 − ERhRaD −ˆsSRa
3
∗
+ ES
2 hSD (s2)∗
+
2
k=1
3
t=1
ES
1 hSRk
2
˜sD
t − ˆsSRk
t
2
. (10)
Q = ySD
1 − ES
1 hSDs1
2
+ ySD
2 − ES
1 hSDs2
2
+ ySD
3 − ES
1 hSDs3
2
(11)
Observe the formulations (7), (9) and (10), clearly by using our proposed DSM-CD scheme
the demodulator at destination can reduce a number of cooperating relays while the dis-
tributed Alamouti STBC is still transmitted.
4 Simulation Results
In order to evaluate the performance of proposed scheme, we use Monte-Carlo simulation
based on computer for the network model as in Figure 1. Where, the links between two
35

SNR (dB)
0 5 10 15 20 25
BER
10-6
10-5
10-4
10-3
10-2
10-1
100
DSM-STBC
DSM-Relay
DSM-Source
DSM-CD (Proposed)
Fig. 2. The BER performance of the proposed DSM-CD scheme.
any nodes are the Rayleigh fading channels, with variance σ2 = σ2
SD = σ2
SR = σ2
RD = 1
and zero mean. The noise at receivers is AWGN with zero and variance σ2
n = N0/2. We as-
sume the average transmit energy per symbol is E. Thus, in the second transmission phase,
the source transmits symbol with the energy ES
2 and the activated relay transmits symbol
with the energy ER, ES
2 = E − ER. Further, the PSK modulation is used. The simulation
is implemented to investigate the BER performance of the transmission schemes over the
common SNR ranges (SNRSD = SNRSR = SNRRD = SNR). Then, the BER performance of
the DSM-CD scheme is compared with schemes in [5] and [7]. Simulation results are shown
as in Figure 2. Clearly, the DSM-CD scheme obtains the better BER performance than the
others. can be explained as follows. With the DSM scheme proposed in [5], the distribution
Alamouti STBC is not used, so in the second transmission phase the diversity gain is not
achieved. More especially, if let relays transmit their symbols estimated from the source
(case of DSM-Relay), errors occurred at the destination will be increased. Thus, the BER
performance is low. However, if let the relays transmit the source’s original data symbols
(case of DSM-Source), the diversity gain will be highly increased. In fact, the demodulation
of relays is imperfect. The DSM-CD scheme allows the error-aware ML demodulator of des-
tination combine information of both the source and relays, so the high diversity is achieved.
Compared to the DSM-STBC scheme in [7], our proposed DSM-CD scheme reduces a
number of relays demodulating the source’s data symbols. Thus, the errors occurred by
the relays used DSM is decreased. In contrast, the demodulation errors at relays make the
36

DSM-STBC scheme in [7] decrease the diversity gain. At the considered SNR ranges, our
DSM-CD scheme also outperforms the DSM-STBC scheme in terms the BER performance.
In the cooperative networks, the network nodes always get opportunities to collaborate with
each other for increasing the transmission rate. Thus, even if there exist good links, each
node still needs to cooperate with any node to obtain the best network performance.
5 Conclusion
In this paper, we proposed the DSM-CD scheme for the cooperative wireless networks. Our
proposal can deal disadvantages of the state-of-the-art schemes on the diversity gains as
well as the system complex. Especially, the proposed cooperative diversity scheme in this
paper can be well applied for wireless ad-hoc and sensor networks, which are considered
more in the future wireless network designs. In the next research, we will consider in the
cross-layer design consisted of MAC layer.
Acknowledgment
This paper was supported by the School of Information and Communication Technology -
Thai Nguyen University in Vietnam under Project: T2017-07-17.
References
1. E. Datsika, A. Antonopoulos, N. Zorba, and C. Verikoukis, “Green cooperative device - to - device
communication: a social - aware perspective,” IEEE Access, vol. 4, pp. 3697–3707, 2016.
2. A. Bansal, M. R. Bhatnagar, and A. Hjorungnes, “Decoding and performance bound of demodulate-
and-forward based distributed alamouti stbc,” IEEE Transactions on Wireless Communications, vol. 12,
pp. 702–713, February 2013.
3. S. Zhang, S. C. Liew, and P. P. Lam, “Hot topic: Physical-layer network coding,” in Proceedings of the
12th Annual International Conference on Mobile Computing and Networking, MobiCom ’06, (New York,
NY, USA), pp. 358–365, ACM, 2006.
4. M. D. Renzo, H. Haas, A. Ghrayeb, S. Sugiura, and L. Hanzo, “Spatial modulation for generalized mimo:
Challenges, opportunities, and implementation,” Proceedings of the IEEE, vol. 102, pp. 56–103, Jan 2014.
5. S. Narayanan, M. D. Renzo, F. Graziosi, and H. Haas, “Distributed spatial modulation for relay networks,”
in Vehicular Technology Conference (VTC Fall), 2013 IEEE 78th, pp. 1–6, Sept 2013.
6. S. Narayanan, M. D. Renzo, F. Graziosi, and H. Haas, “Distributed spatial modulation: A coopera-
tive diversity protocol for half-duplex relay-aided wireless networks,” IEEE Transactions on Vehicular
Technology, vol. 65, pp. 2947–2964, May 2016.
7. S. Narayanan, A. Stavridis, M. D. Renzo, F. Graziosi, and H. Haas, “Distributed spatially-modulated
space-time-block-codes,” in 2013 IEEE 18th International Workshop on Computer Aided Modeling and
Design of Communication Links and Networks (CAMAD), pp. 159–163, Sept 2013.
37

Distributed Spatial Modulation based Cooperative Diversity Scheme

More Related Content

Distributed Spatial Modulation based Cooperative Diversity Scheme