Physical Communication 57 (2023) 101977

Contents lists available at ScienceDirect

Physical Communication
journal homepage: www.elsevier.com/locate/phycom

Full length article

Distributed spatial media-based modulation for relay networks with

RF mirrors
∗
Taissir Y. Elganimi a , , Khaled M. Rabie b , Ammar M. Abu-Hudrouss c
a
Department of Electrical and Electronic Engineering, University of Tripoli, Libya
b
Department of Engineering, Manchester Metropolitan University, United Kingdom
c
Electrical Engineering and Smart Systems Department, Islamic University of Gaza, Palestine

article info a b s t r a c t

Article history: A novel distributed spatial media-based modulation scheme is proposed in this paper by cleverly
Received 29 October 2021 utilizing distributed spatial modulation (DSM) and media-based modulation (MBM) principles. This
Received in revised form 16 December 2022 proposed scheme is referred to as distributed channel modulation (DCM) for relay networks. In
Accepted 17 December 2022
this scheme, decode-and-forward relaying protocol is adopted, and the channel states are exploited
Available online 27 December 2022
for transmitting extra information bits by using a number of radio frequency (RF) mirrors that are
Keywords: placed near each relay. To provide a fair comparison with the conventional state-of-the-art schemes,
Distributed channel modulation (DCM) the symbol error rate (SER) performance of DSM scheme is evaluated. Besides, a low complexity
Media-based modulation (MBM) detection technique known as iterative maximum ratio combining (i-MRC) is used in order to reduce
RF mirrors the receiver complexity of the proposed scheme. Simulation results demonstrate that the proposed
Index modulation
DCM scheme significantly outperforms DSM scheme for the same average rate. It is also shown
Channel estimation errors
that there is a negligible degradation in the SER performance of the proposed DCM scheme when
i-MRC detection is used as compared to the performance with maximum likelihood (ML) detection.
Furthermore, a significant reduction in the receiver complexity is achieved by using i-MRC detection
technique in contrast to the results with ML detector. It has been also revealed that the proposed DCM
scheme shows a performance drop of about 3 dB when the availability of an imperfect channel state
information (CSI) is assumed with the presence of channel estimation errors (CEEs). Finally, simulation
results have confirmed the analytical findings.
© 2022 Elsevier B.V. All rights reserved.

1. Introduction MIMO schemes such as spatial multiplexing (SMX) scheme that

requires many transmit antennas to be equipped and activated
Cooperative communication techniques and relaying protocols simultaneously [10–12]. Even with utilizing the orthogonality
have been recently emerged as an alternative solution and poten- of the in-phase and quadrature-phase components as in [13],
tial candidate for future wireless systems and applications [1,2]. the emerging SM systems are impractical with large number of
On the other hand, distributed multi-relay cooperative diversity antennas at the transmitter and useful only in downlink multiuser
systems such as distributed spatial modulation (DSM) scheme are scenarios with equipping a large number of base station antennas.
regarded effective in overcoming the effect of both the multi-path
To circumvent the aforementioned issue, DSM concept is pro-
fading and the heavy shadowing, reduce the transmit power con-
posed in [3], and studied more extensively by the same authors
sumption, and improve the coverage of the fifth generation (5G)
in [4]. The concept of SM schemes is utilized in this cooperative
wireless communications [3–5]. However, many researchers have
diversity protocol in order to allow the active relay to re-convey
focused on the benefits of combining multiple input multiple
output (MIMO) schemes with relay-aided cooperation systems, the frames of the source to the destination. More recently, DSM
e.g., [6–8]. scheme is generalized in [14] with activating multiple relays at
Spatial modulation (SM) with single radio frequency (RF) chain each time instant based on the concept of generalized spatial
is considered as the most popular index modulation (IM) tech- modulation (GSM) [15]. This multi-relay cooperative diversity
nique that has been regarded as an innovative technique in wire- protocol is referred to as distributed generalized spatial modula-
less communications [9]. In terms of the computational com- tion (DGSM) scheme [14]. Additionally, the research studies [16–
plexity, however, it is capable of outperforming the conventional 18] have extensively studied SM-based relay-aided protocols and
took advantages of applying SM principle in cooperative relaying
∗ Corresponding author. communications. All relaying protocols in these papers treat the
E-mail address: [email protected] (T.Y. Elganimi). SM scheme as a conventional modulation technique, and aim to

https://doi.org/10.1016/j.phycom.2022.101977
1874-4907/© 2022 Elsevier B.V. All rights reserved.
T.Y. Elganimi, K.M. Rabie and A.M. Abu-Hudrouss Physical Communication 57 (2023) 101977

to the spatial domain. For instance, the indices of transmit an-

Nomenclature tennas in SM systems are exploited in the spatial domain [10],
5G Fifth generation while in [19] the subcarrier indices of multicarrier systems are
bpcu Bits per channel use used which is known as subcarrier index modulation (SIM). The
enhanced schemes of SIM are also considered in [20–22] where
CEE Channel estimation error
enhanced orthogonal frequency division multiplexing with index
CSI Channel state information
modulation (OFDM-IM) schemes aimed at achieving higher spec-
DCM Distributed channel modulation tral efficiency and diversity gain are proposed. In [23,24], the
DF Decode-and-forward indices of time slots in the time domain are utilized to modulate
DGSM Distributed generalized spatial modula- additional bits, while in [25,26] the indices of channel states
tion activated by the RF mirrors are used, and termed as media-
DSM Distributed spatial modulation based modulation (MBM). Moreover, several multidimensional IM
DSMBM Double spatial media-based modulation schemes and their significant performance enhancements have
FD Full-duplex caught the attention of plenty of researchers in recent years, and
FDR Full-duplex relaying have been considered as innovative techniques for 5G and beyond
GQCM Generalized quadrature channel modu- wireless systems [[27], and references therein].
lation Recently, the emerging MBM transmission technique has at-
tracted the researchers’ attention, where the transmission is car-
GQSM generalized quadrature spatial modula-
ried out in the channel domain. The innovative concept of MBM
tion
technique is based on varying the RF properties of the propa-
GSM Generalized spatial modulation
gation environment around transmit antennas, such as perme-
i-MRC iterative maximum ratio combining
ability, permittivity and resistivity. In this technique, digitally
IM Index modulation controlled parasitic elements can be added to each transmit an-
MAPs Mirror activation patterns tenna and act as an RF mirror [25,26]. Each mRF mirror is a passive
MBM Media-based modulation antenna element that contains a positive-intrinsic-negative (PIN)
MBM-TU MBM transmit unit diode that can be switched ON and OFF in response to the in-
MBSTBC-SM Media-based space–time block coded coming information bits which can be detected at the receiver
spatial modulation as each state has a different channel realization. However, inde-
MIMO Multiple-input multiple-output pendent channel fade realizations can be generated by different
ML Maximum likelihood mirror activation patterns (MAPs), and the available mRF mirrors
MRC Maximum ratio combining generate 2mRF MAPs.
In general, SM and space shift keying (SSK) schemes use mul-
OFDM-IM Orthogonal frequency division multi-
tiple antennas at the transmitter in order to create the channel
plexing with index modulation
modulation alphabet. The number of these required antennas
PDF Probability density function
grows exponentially with the spectral efficiency. Whereas MBM
PEP Pairwise error probability transmission technique requires a number of RF mirrors that
PIN Positive-intrinsic-negative increases linearly with the spectral efficiency. Therefore, MBM
PSK Phase shift keying transmission scheme with only one transmit antenna and mRF
QAM Quadrature amplitude modulation mirrors can be regarded as an equivalent SSK or SM scheme with
QCM Quadrature channel modulation 2mRF antennas at the transmitter.
QSM Quadrature spatial modulation
RF Radio frequency 1.1. Related work
SER Symbol error rate
SIM Subcarrier index modulation Integrating several MIMO schemes with MBM transmission
technique is considered in several recent studies and showed
SIMO Single-input multiple-output
significant performance improvements that are achieved over the
SM Spatial modulation
conventional modulation techniques due to the good distance
SMX Spatial multiplexing properties of MBM signal vectors [28]. For example, in [25,26],
SNR Signal-to-noise ratio single-input multiple-output MBM (SIMO-MBM) scheme is in-
SSK Space shift keying troduced where the index bits are transmitted by the index
STBC Space–time block codes of a selected wireless channel state in SIMO systems. In this
STBC-SM Space–time block coded spatial modu- scheme, the single-antenna source is equipped with a number of
lation RF mirrors to convey extra bits. In [29], GSM scheme is combined
STCM Space–time channel modulation with MBM concept, and showed that the GSM-MBM scheme is
USTLD Uncoded space–time labeling diversity capable of outperforming its conventional MBM-based counter-
parts for the same spectral efficiency. At the end of the same
year, quadrature spatial modulation (QSM) scheme [30] has been
integrated with MBM technique by other authors in [31], and
improve either the network coverage or the link reliability of SM known as quadrature channel modulation (QCM). In addition, the
technique. authors in [32] have combined the conventional space–time block
codes (STBC) with MBM approach and termed as space–time
Multidimensional IM schemes have been intensively investi-
channel modulation (STCM), where the index bits are conveyed
gated in recent years as promising techniques for future wireless
by the index of selected channel state vectors in Alamouti STBC
communications. In these schemes, each dimension can be used schemes. Moreover, a further enhancement was achieved in [33]
to modulate information bits by providing more exploitation of by combining the uncoded space–time labeling diversity (USTLD)
the multiple transmission entities including but not restricted scheme with MBM technique, and referred to as USTLD-STCM
2
T.Y. Elganimi, K.M. Rabie and A.M. Abu-Hudrouss Physical Communication 57 (2023) 101977

Table 1
A comparison between the related work and the proposed DCM scheme.
Reference Year Scheme Number of Detection technique Error Achievable rate CSI knowledge
antennas in the performance
source
[25] 2013 SIMO-MBM 1 Sequence detection in ✗ ✓ Perfect CSI
time
[29] 2017 GSM-MBM 4 ML ✓ ✗ Perfect CSI
[31] 2017 QCM 4 ML ✓ ✗ Perfect and
imperfect CSI
[32] 2017 STCM 2 ML ✓ ✗ Perfect CSI
[33] 2018 USTLD-STCM 2 ML ✓ ✗ Perfect and
imperfect CSI
[34] 2019 MBSTBC-SM 3, 4, 5 and 8 ML and near-ML ✓ ✗ Perfect CSI
[36] 2020 MBSTBC-SM-LD 4 ML and near-ML ✓ ✗ Perfect CSI
[37] 2021 GQCM 2 and 3 ML ✓ ✗ Perfect and
imperfect CSI
[38] 2021 DSMBM 2 and 4 Soft-output ML ✓ ✗ Perfect CSI
[42] 2019 FDR-MBM 4 ML ✓ ✓ Perfect CSI
[43] 2020 SSK-MBM 2 and 4 ML ✓ ✗ Perfect CSI
Proposed 2022 DCM 1 ML and i-MRC ✓ ✗ Perfect and
scheme imperfect CSI

scheme. The reported results in [32,33] highlighted the advan- 1.2. Contributions and paper organization
tages of MBM-based schemes over their classical counterparts. In
another work [34], the application of MBM in space–time block Against this backcloth, and motivated by the performance
coded spatial modulation (STBC-SM) schemes [35] is proposed, improvement that can be achieved through the use of extra index
and named as media-based STBC-SM (MBSTBC-SM). Later, the bits in IM schemes, the main contributions of the present paper
error performance of MBSTBC-SM is further improved in [36] by can be summarized as follows:
applying the labeling diversity, and termed as MBSTBC-SM-LD.
The results in [34,36] showed that the error performance is signif- 1. The concept of DCM scheme is introduced by combining
icantly improved over the conventional schemes. Another study DSM scheme for MIMO relay networks and MBM transmis-
has been recently reported in [37] where QCM scheme is gen- sion principle. This was proposed to further enhance the
eralized to another framework that is referred to as generalized data rate of DSM schemes by further exploitation of the
quadrature channel modulation (GQCM) by combining MBM and space domain by sending additional information bits via
generalized quadrature spatial modulation (GQSM) techniques. It the channel states domain without increasing the number
is shown in [37] that the GQCM scheme further improves the of relays. In this paper, the symbol error rate (SER) per-
error performance as compared to its conventional counterparts. formance with maximum likelihood (ML) detection is as-
More recently, a new technique is proposed in [38] where double sessed in comparison with the conventional DSM scheme,
spatial modulation scheme [39] is extended to employ MBM tech- and the mathematical analysis is shown and confirmed the
nique. This new modulation technique is termed as double spatial computer simulation results.
media-based modulation (DSMBM), and showed that DSMBM 2. A low complexity detection scheme based on iterative
scheme outperforms the conventional schemes in terms of the maximum ratio combining (i-MRC) is used in the pro-
error performance at the same spectral efficiency. posed scheme in order to reduce the high computational
Further, the emerging MBM technique has been widely consid- complexity of ML detection technique.
ered in a range of scenarios such as differential MBM [40], mul- 3. Finally, the effect of the presence of channel estimation
tiuser environment [41], full-duplex (FD) decode-and-forward errors (CEEs) on the SER performance of the proposed
(DF) relaying cooperative communications [42] which is referred scheme with imperfect channel state information (CSI) is
to as FD relaying scheme with MBM (FDR-MBM), and hybrid evaluated with ML and i-MRC detection techniques.
cooperative space shift keying media-based modulation (SSK-
MBM) schemes [43]. In [43], two novel schemes were proposed It has been shown with simulation results that there is a neg-
with considering a multi-antenna source that communicates with ligible performance drop in the SER performance of DCM scheme
multiple cooperative relays and the destination by applying (SSK- with i-MRC detection over the performance with ML detectors.
MBM) at the source. In these two schemes, even the relays Furthermore, a significant reduction in the receiver complexity is
are equipped with MBM using RF mirrors. All these foregoing obtained by using i-MRC detector as compared to the results with
MBM-based systems have showed significant performance im- ML detection.
provements over their non-IM-aided counterparts. The rest of the present paper is organized as follows. Section 2
For better clarity, Table 1 summarizes and compares the pre- presents the system model of the proposed DCM scheme. In
vious related work and the proposed distributed channel mod- Section 3, the detection techniques used in the proposed scheme
ulation (DCM) scheme which is also referred to as distributed are described. The performance analysis of the proposed DCM
spatial media-based modulation scheme. Then, it is compared scheme is formulated and presented in Section 4. The computa-
to the conventional DSM scheme that is equipped with single- tional complexity is presented in Section 5, and the SER perfor-
antenna source and at the same average rate for the purpose of mance is shown in Section 6. Finally, the conclusions and possible
ensuring a fair comparison. future research directions are summarized in Section 7.
3
T.Y. Elganimi, K.M. Rabie and A.M. Abu-Hudrouss Physical Communication 57 (2023) 101977

Fig. 1. Block diagram of the proposed DCM scheme with single-antenna source, R relays, and both the source and the cooperative relays are equipped with RF
mirrors, and the destination is equipped with Nr receive antennas.

1.3. Notations technique with constellation size MR , and one active relay is
selected to convey the complex symbol at each particular time
In this paper, the upper-case boldface letters denote the ma- instant. The number of bits that can be conveyed in DCM scheme
trices, while the lower-case boldface letters denote the vectors. depends on the constellation diagram that used, and the number
AH , |A| and ∥A∥F denote the Hermitian (conjugate transpose), of relays that equipped, where log2 M + mRF s information bits
the absolute value
( )and the Frobenius norm operation of matrix can be conveyed at each time instant. Moreover, each relay is
a assigned a unique digital identifier IDr for r = 1, 2, . . . , R, where
A, respectively. is the binomial coefficient which represents
b the number of cooperative relays must be a power of two, and
the combinations that result from choosing b elements from M2mRF s = RMR 2mRF r . In this corresponding, the constellation size
a elements without repetition, while CM ×N denotes a complex MR at the relays which must be a power of two can be found
matrix having a size of M × N. according to the following expression
M
2. Distributed channel modulation MR = . (1)
R2(mRF r −mRF s )
In this section, the concept of the proposed DCM scheme for The key idea of designing DCM scheme is based on cleverly
MIMO relay networks is introduced. The system model of the combining both DSM and MBM techniques, where M = MR R is
proposed DCM scheme is depicted in Fig. 1 with equipping a considered in this paper.
single-antenna source that uses the ordinary M-ary phase shift In the proposed DCM scheme, the transmission of data occurs
keying (PSK) or M-ary quadrature amplitude modulation (QAM) over two phases and lasts for two time slots as follows
technique, where M is the constellation size. Indexing in DCM
scheme is performed in the space and channel domains with 2.1. Broadcasting phase of DCM scheme
the aid of MBM transmission principle, where both the source
and the cooperative relays are equipped with RF mirrors. In this In the broadcasting phase of the proposed DCM scheme, let pS
scheme, however, the RF mirrors change the state of wireless be the log2 M +mRF s digital bits emitted by the source to a number
channels between the source and the relays, and between each of relays. In this scheme, s = ψS (pS ) is the transmitted complex
relay and the destination based on the incoming information symbol that is conveyed from the source with s ∈ {s1 , s2 , . . . , sM },
bits. The RF mirrors are placed near each relay in the proposed where ψS (.) stands for the bit-to-symbol modulation mapping
DCM scheme in order to allow the incident RF signal to pass function at the single-antenna source. The received signal vector,
through transparently when it is ON [29], and to facilitate a secure yr ∈ CNRr ×1 , at the rth relay in DCM scheme is expressed as [3]
transmission [44]. Placing RF mirrors near a relay, however, is √
equivalent to placing scatterers in the wireless propagation en- yr = ES hb s + nr , (2)
vironment close to the relay. In the transmitter structure of this
where b ∈ (1, 2, . . . , 2 mRF s
) corresponding to the active channel,
scheme, the single-antenna source is equipped with mRF s mirrors
ES denotes the source’s average transmit energy per symbol, hb is
to convey log2 M + mRF s bits to R relays that share their transmit mRF
s
the bth column of the Rayleigh fading channel Hb ∈ CNRr ×2 for
antennas and form a virtual antenna array, where each relay’s
the rth source-to-relay link having a zero-mean and unit variance,
transmit antenna is equipped with mRF r mirrors, which change
NRr is the number of receive antennas in the broadcasting phase,
the channel state between each relay and the destination. These
and nr ∈ CNRr × 1 is the complex additive Gaussian noise
mirrors create disturbances by changing the RF characteristics
vector at the rth relay with zero-mean and noise variance σn2 per
around the relay and the status of the mirrors, and hence convey-
dimension.
ing additional information through the different channel states. In
With the aid of ML criterion, the source’s estimated symbol at
this proposed scheme, DF relaying protocol is adopted. At each
the rth relay is obtained by demodulating the received signal, yr ,
time instant, similar to the conventional SM scheme, only one
of the proposed DCM scheme as follows
selected relay is allowed to be active in the distributed system to { }
forward the estimated bits to the destination which is equipped √
{ŝ(r) , b̂} = ⏐⏐yr − ES hb sm ⏐⏐2 ,
⏐⏐ ⏐⏐
with multiple receive antennas. Therefore, the overall transmit arg min (3)
m
b∈{1,2,...,2 RF s },sm ∈{s1 ,s2 ,...,sM }
power at the relays is saved.
Consider the proposed multi-relaying network of DCM system where {s1 , s2 , . . . , sM } is the complex data symbol that transmit-
(r)
with single-antenna source and R relays that employ PSK or QAM ted from the single-antenna source. Thus, the estimate of pS at
4
T.Y. Elganimi, K.M. Rabie and A.M. Abu-Hudrouss Physical Communication 57 (2023) 101977

the rth relay is found as After demodulating the signal received from the single-ante-
(r) nna source by using (3), each cooperative relay applies the SM
p̂S = ψS (ŝ ),
−1 (r)
(4) principle [10] in order to simultaneously forward the source’s
where ψS (.) is the inverse bit-to-symbol modulation mapping
−1 data. Therefore, the modulated symbol of the rth relay is chosen
function. as
(r)
ψR (xR ) IDr = p̂S ,
{
2.2. Relaying phase of DCM scheme xr (r) = (r) (7)
0 IDr ̸ = p̂S ,

In the relaying phase of the proposed DCM scheme, each rth where ψR (.) is the bit-to-symbol modulation mapping at each
relay divides the demodulated information bits into three groups. relay, and xR is the log2 R + log2 MR + mRF r bits emitted by the
The first group of the incoming bit sequence with a length of relay if activated.
log2 R information bits is used to determine the index of the The received signal vector, yd ∈ CNr ×1 , where d = 1, 2, . . . , Nr
active relay. The subsequent group with a length of log2 MR digital and Nr is the number of receive antennas at the destination, can
bits is used to determine the constellation size of PSK or QAM be written as
√
complex symbol that conveyed by the active relay. The active yd = Er hv xr (r) + nd , (8)
relay compares the first log2 R bits with its own digital identifier,
and then modulates the remaining bits of the second group using where v ∈ (1, 2, . . . , R2 mRF r
) which is an index to the active
a modulation scheme if the first log2 R bits match with its own channel, Er denotes the rth relay’s average transmit energy per
digital identifier. This active relay is responsible for forwarding symbol, and hv is the v th column of the Rayleigh fading channel
mRF
r
the modulated symbol to the destination. Finally, the last mRF r Hv ∈ CNr ×R2 for the link between the rth MBM-TU and
bits select the active wireless channel state which corresponds to the dth receive antenna of the destination with zero-mean and
the ON or OFF status of the available RF mirrors that associated unit variance. In addition, xr is the transmitted symbol at the
with the active relay. rth MBM-TU, and nd ∈ CNr ×1 is the complex additive Gaussian
The proposed DCM scheme extends the ordinary DSM scheme noise vector at the destination with zero-mean and noise variance
for relay networks into a third dimension, which is the channel σn2 per dimension. Since only one relay is active at each time
states dimension, in order to convey additional bits. Regarding instant in the relaying phase, it is assumed in this paper that
the operation of DCM scheme, consider the following example the total transmission energy is equally divided between the
with M = 64, mRF s = mRF r = 2, and R = 4 relays. In this single-antenna source and the active relay.
scheme, MR can be found as 16. If the input bits are grouped
[ ] 3. Detection schemes
as 10 1011
 01 , the first group of log2 R = 2 bits (10)
   
log2 R log2 MR mRF r After receiving the signal symbols at the destination of the
is used to determine the active relay’s antenna index, ra = 3, proposed DCM scheme, a signal detector is used in order to jointly
over which the data symbol will be transmitted, while the next estimate the index of the active MBM-TU, the channel state,
log2 MR = 4 bits (1011) are modulated to obtain the 16-QAM and the symbol that is transmitted by the active MBM-TU, x̂, as
symbol x = 3 − j. The last mRF r = 2 bits (01) select the second follows.
channel state that corresponds to the ON and OFF status of the
available RF mirrors of the active relay, where the first RF mirror 3.1. Maximum Likelihood (ML)
is OFF, and the second RF mirror is ON.
In DCM scheme, each relay has single transmit antenna with Assuming that a perfect CSI is available at the destination, the
mRF r mirrors which is known as MBM transmit unit (MBM-TU). ML optimum detection rule is given by
One of the available R MBM-TUs is chosen and one of the mRF r is { √ }
r̂a , x̂, m̂RF r = ⏐⏐yd − Er hv qn ⏐⏐2 ,
[ ] ⏐⏐ ⏐⏐
activated to transmit the modulated signal symbol at each time arg min
instant based on log2 (R2mRF r ) information bits. The number of mRF
v∈{1,2,...,R2 r },qn ∈{q1 ,q2 ,...,qM }
R
bits per channel use (bpcu) that are transmitted from the active
relay in DCM scheme can be expressed as (9)
where {q1 , q2 , . . . , qM } is the complex data symbol that trans-
ηR = log2 M + mRF s = log2 R + log2 MR + mRF r , (5) R
mitted from the active MBM-TU.
where MR depends on the choice of M, R, mRF s and mRF r . Since
there are two phases in the proposed DCM scheme, and the data 3.2. Iterative Maximum Ratio Combining (i-MRC)
transmission occurs in two time slots, the average rate (in bpcu)
can be expressed as In this paper, an i-MRC detection technique has been used
1[ 1[ ] ] at the destination of DCM scheme in order to reduce the re-
η= log2 M + mRF s = log2 R + log2 MR + mRF r . (6) ceiver complexity. The i-MRC detection algorithm is basically
2 2 based on the fact that only one transmit antenna is active at a
It is worth mentioning that increasing the number of RF mir- time. Therefore, the active antenna number may change at each
rors improves the average rate, but a penalty is paid in increasing subsequent transmission instant, and only one transmit antenna
the constellation size M at the source of DCM scheme if mRF s is is transmitting the information at each particular time instant. In
kept constant. It is also clear that increasing the number of RF addition, this detection technique considers the channel vectors
mirrors at the relays requires lower constellation size, MR , for the between each transmit antenna and the number of multiple
same average rate. receive antennas separately at the receiver. Thus, the receiver
The main advantage of integrating MBM into DSM systems is iteratively computes the maximum ratio combining (MRC) results
that the number of information bits that conveyed by indexing between the channel paths from each transmit antenna to the
the RF mirrors increases linearly with the number of mirrors, corresponding receive antennas. Finally, the receiver chooses the
while the number of index bits in the conventional DSM schemes active antenna number which gives the highest correlation when
increases logarithmically with the number of cooperative relays. full knowledge of the channel at the receiver side is assumed [10].
5
T.Y. Elganimi, K.M. Rabie and A.M. Abu-Hudrouss Physical Communication 57 (2023) 101977

In i-MRC detection scheme, the received signal is iteratively 4.2. Relaying phase
multiplied by the channel path gains. These gains are assumed to
be known at the destination to estimate the index of the active Following the same procedure in the previous subsection, the
MBM-TU, the channel state, and the data symbol that conveyed unconditional average error probability in the relaying phase of
by the active MBM-TU as [10] the proposed DCM scheme can be derived as
ĉa = arg max(zc ), for c = 1, 2, . . . , R2mRF r , (10) Nr −1 (
∑ Nr − 1 + k)
Pe2 a → â = γ
¯ Nr
[1 − γ ]k .
( )
c (19)
k
x̂ = D(zĉa ), (11) k=0
⏐
⏐ H ⏐
⏐ The average symbol error probability at any destination at this
⏐hv yd ⏐
where zc = , and D is the constellation demodulator func- stage is given by
∥hv ∥2F
k k
tion. As channel states are independent, the demapped ĉa can be 2 2
2 ∑∑
P¯e2 ai → âj .
( )
used for joint detection for the active relay and the active channel PsD = (20)
state, while x̂ is used to detect the transmitted data symbol. 2k
i=1 j=i

4. Performance analysis 4.3. Combined error performance

In this section, the performance analysis of the proposed DCM The error performance for the complete cycle of transmission
scheme is introduced. In particular, the error performance of from the transmitter to the receiver of the proposed DCM scheme,
the broadcasting phase between the transmitter and the relays, a successful transmission is needed in both stages. If an error
and the error performance of the relaying phase between the occurs in the first stage, it may propagate to the second stage.
relays and the receiver are studied separately. Consequently, the Therefore, the SER in the first stage occurs if any relay decode
combined error performance in the two phases is introduced. the vector wrongly and the relay is active in the second stage, or

4.1. Broadcasting phase P̄sR = RP̄sRi /R = P̄sRi . (21)

In other words, the total SER at the relay is the summation of the
Let a and â be two distinct vectors presuming that â is an
probability at each relay which is represented by RP̄sRi . However,
estimated vector at the relay when the vector a is transmitted.
due to random distribution of bits, the probability that one of the
Then, the pairwise error probability (PEP) is given by
channels of the relay will be active in the second stage is equal
to 1/R. Therefore, the total probability of SER is equal to the SER
⎛√   ⎞
a − â2 (√ )
ζ ,
[ ]
Pr a → â|H = Q ⎝ ⎠=Q (12) of a single relay. Consequently, the combined SER performance is
2σn2
given by

where Q (.) is the Q -function, and SER = P̄sR (1 − P̄sD ) + P̄sD (1 − P̄sR ) + P̄sR P̄sD . (22)
1 ( )H (
ζ = a − â .
)
a − â (13) 5. Computational complexity
2σn
2

With NRr receive antennas per relay, the PEP is given by In this section, the computational complexity of the proposed
⎛ ⎞ DCM scheme with ML detection technique is compared to the
 NRr complexity of the same scheme with i-MRC detector. The receiver
∑ (√ )
ζi ⎠ = Q Z ,
[ ]
Pr a → â|H = Q ⎝√ (14) complexity is computed as the number of the real-valued multi-
⎜ ⎟
i=1 plicative operations (×, ÷) that required by each algorithm. The
computational complexity of ML receiver in (9) is found as
which results in Z being a chi-squared random variable with a
probability density function (PDF) given by CDCM-ML = 8Nr R2mRF r MR , (23)
⏐⏐ √ ⏐⏐
as evaluating ⏐⏐yd − Er hv qn ⏐⏐2 requires 2 complex multiplica-
( )
1 Z
PZ (Z ) = Z NRr −1 exp − , (15) tions, and each complex-valued multiplication requires 4 real-
Γ (NRr )ζ̄ NRr ζ̄
valued multiplications.
where Γ (.) denotes the Gamma function, and ζ̄ is the expected In contrast, i-MRC detector requires Nr R2mRF r complex mul-
value of ζ , given by tiplications, therefore; the computational complexity of DCM-
{ ⏐ i-MRC receiver in terms of the number of real multiplicative
⏐s − ŝ⏐2 hSr = ĥŜr ,
⏐
ES operations can be written as
ζ̄ = ⏐ ⏐2 (16)
2σn2 |s|2 + ⏐ŝ⏐ hSr ̸ = ĥŜr .
CDCM-i-MRC = 4Nr R2mRF r . (24)
The average unconditional PEP is given by [45,46]
It can be clearly seen that the receiver complexity in (23)
NRr −1
∑ (NR − 1 + k) and (24) can be specialized to the conventional DSM schemes by
P̄1Ru = P¯e a → â = γ NRr [1 − γ ]k ,
( ) r
(17) substituting mRF r = 0.
k The percentage of reduction in the receiver complexity of
k=0
( √ ) DCM scheme with i-MRC detector as compared to the receiver
ζ̄ /2
where γ = 1
2
1− 1+ζ̄ /2
. complexity using ML detection can be expressed as
The average symbol error probability at any relay is given by (
CDCM-i-MRC
) (
1
)
2 2 k k Cred % = 1− × 100 = 1 − × 100. (25)
2 ∑∑ CDCM-ML 2MR
P¯e1 ai → âj ,
( )
PsRi = (18)
2k It is worth noting that the percentage of reduction in the
i=1 j=i
receiver complexity of the proposed DCM scheme increases with
where k is double the spectral efficiency given in bps/Hz. increasing the constellation size, MR . It is also clear from (25)
6
T.Y. Elganimi, K.M. Rabie and A.M. Abu-Hudrouss Physical Communication 57 (2023) 101977

Fig. 2. SER performance of DSM-ML and the proposed DCM-ML schemes with R = 4 relays and NRr = 4 receive antennas are equipped at each relay: (a) For 8 bpcu;
(b) For 10 bpcu.

that DCM scheme with i-MRC detection has much lower com- This is mainly because a number of RF mirrors are employed
plexity than DCM scheme with ML detection technique. For ex- in DCM schemes which causes a performance improvement as
ample, a reduction of almost 75%, 87.5%, 93.75% and 96.875% is compared to the conventional DSM schemes. The main reason for
achieved with MR = 2, 4, 8 and 16, respectively. This shows this SER performance improvement can be also explained by the
that the computational complexity is significantly reduced using fact that the conventional DSM schemes need to employ higher
i-MRC detection technique without any noticeable loss in the SER order constellations to reach the same average rate as that of the
performance as will be shown in Section 6. proposed DCM schemes. Similarly, the proposed DCM schemes
can achieve a higher spectral efficiency than the conventional
6. Simulation results and discussion DSM schemes for the same modulation order since the DCM
scheme employs the MBM transmission technique. This clearly
In this section, the SER performance of the proposed DCM shows that the proposed DCM scheme is better than DSM scheme
scheme is evaluated through extensive computer simulations and in terms of the SER performance, especially with employing RF
comparisons. Throughout the simulation, the channels between mirrors at both the source and the cooperative relays.
the source and the cooperative relays and between these relays The SER performance of DCM schemes with increasing the
and the destination are modeled as Rayleigh fading channels. A number of RF mirrors is shown in Fig. 3(a) and (b) using ML
perfect channel knowledge at the receiver is also assumed for and i-MRC detection techniques, respectively. For the sake of
perfect CSI unless otherwise stated. In addition, it is assumed fair comparison, the SER performance of DSM schemes is also
that each relay is equipped with one transmit antenna that is evaluated in these figures. In these comparisons, all schemes
surrounded by a number of RF mirrors in MBM-aided schemes. are equipped with 4 relays, each consists of NRr = 2 receive
In Fig. 2(a) and (b), the SER performance of the proposed DCM antennas, and 256-QAM technique is employed at the source. In
scheme versus the signal-to-noise ratio (SNR) is compared to that addition, MR is kept constant at 4 for 4-QAM technique in all
of the conventional DSM scheme for 8 and 10 bpcu at each time DCM schemes, while 64-QAM is employed in DSM schemes. It
slot, respectively. The optimum ML detection is used in these is obvious from these two figures that increasing the number
comparisons, R = 4 relays is also considered with NRr = 4 receive of RF mirrors improves the average rate of the proposed DCM
antennas are equipped at each relay, and mRF s = 2 and mRF r = 4 scheme. In other words, an additional bpcu is achieved by adding
and 6 are considered in DCM schemes. The average rate of DCM each two RF mirrors at the source and two RF mirrors at each
scheme, η, is found as expressed in (6). In this comparison, DCM relay, while the SER performance degrades by about 0.5 − 1 dB
schemes are denoted by DCM-ML (Nt , Nr , NRr , M-QAM, MR -QAM, when each two RF mirrors are increased at the source and each
mRF s , mRF r ). Without loss of generality, the source is assumed in relay. The reason for the average rate improvements is that the
this paper to be equipped with one transmit antenna, i.e., Nt = RF mirrors in the proposed DCM scheme create disturbances by
1. Moreover, DSM scheme is also taken into consideration as a changing the RF characteristics around the relays and the status of
special case of DCM scheme by substituting mRF r = 0 and mRF s = the mirrors, and hence conveying extra information bits through
0, and another special case of DCM scheme is also considered different channel states.
by substituting mRF s = 0 at the source. It is noticeable from Fig. 4 shows the SER performance of the proposed DCM
Fig. 2 that the proposed DCM schemes outperform the conven- schemes using ML and i-MRC detection techniques for different
tional DSM schemes by about 5.5 dB and 6 dB for 8 and 10 bits average rates, where all schemes are equipped with 4 relays, each
transmission cases, respectively. This also shows that increasing consists of NRr = 2 receive antennas, and MR is kept constant at
the number of transmitted bits by increasing the modulation 4 for 4-QAM technique in all DCM schemes. It is clear from this
orders leads to increasing the performance gain that achieved figure that the proposed DCM schemes with the two detection
in DCM schemes, regardless the performance degradation that techniques match each other, and DCM schemes with i-MRC
occurs due to increasing the number of bpcu. In addition, it is detection technique show a slight degradation at low SNR values
shown that a negligible performance drop occurs in DCM schemes as compared to their DCM-ML counterparts.
with mRF s = 0 over their DSM counterparts for the same average The SER performance with ML detection is depicted in Fig. 5
rate. In these cases, DSM schemes employ higher modulation in order to investigate the performance of the proposed DCM
order in the relays than the modulation order of DCM schemes. scheme in the presence of CEEs in both broadcasting and relaying
7
T.Y. Elganimi, K.M. Rabie and A.M. Abu-Hudrouss Physical Communication 57 (2023) 101977

Fig. 3. SER performance of the proposed DCM scheme with increasing the number of RF mirrors. It is assumed that all DSM and DCM schemes are equipped with
4 relays and NRr = 2 receive antennas are equipped at each relay: (a) with ML detection, (b) with i-MRC detection.

Fig. 4. SER performance of the proposed DCM schemes with ML detection (solid lines) and i-MRC detection (dashed lines). It is assumed that R = 4 relays and
NRr = 2 receive antennas are equipped at each relay: (a) For DCM schemes with η = 5 bpcu; (b) For DCM schemes with η = 6 bpcu.

phases. Similarly, the performance of DCM scheme is shown in 7. Conclusions

Fig. 6 with the low-complexity i-MRC detection technique. In
this paper, the estimated channel matrices are obtained as H̃b = In this paper, a new multidimensional scheme for MIMO re-
Hb − E and H̃v = Hv − E, which are taken into consideration in laying protocols is proposed, and referred to as distributed spatial
(3) with the presence of channel imperfection in the broadcasting media-based modulation scheme in which antennas and RF mir-
phase and (9)–(11) in the presence of channel imperfection in rors are indexed simultaneously. This scheme is also termed as
the relaying phase, respectively. In addition, E represents the distributed channel modulation for relay networks. It is shown
CEE matrix with entries having zero-mean and error variance σe2 , from the simulation results that DCM-i-MRC schemes benefit
which is assumed in this paper to be equal to the noise variance from the substantially lower receiver complexity, and have al-
σn2 [37]. From these comparisons, it is noticed that a performance most similar error performance as DCM-ML schemes. Addition-
degradation of about 3 dB is found when an imperfect CSI is ally, it is revealed that the presence of CEEs degrades the SER
assumed with the presence of CEEs in all DCM schemes. performance of the proposed schemes by about 3 dB when chan-
nel imperfection is assumed in both relaying and broadcasting
Finally, the analytical results of the SER performance of the
phases. It is also demonstrated that the analytical results of
proposed DCM-ML scheme with R = 4 relays, MR = 4 and NRr =
DCM scheme is shown to agree well with simulation results.
4 receive antennas are equipped at each relay are shown in Fig. 7
Furthermore, the improved average rate makes the proposed
and compared to the computer simulation results. In addition,
DCM scheme a promising candidate for 5G and beyond wire-
the analytical and simulation results are compared without and less systems. For further works, it is of great interest to focus
with CEEs for 4 bits transmission with M = 64, mRF s = 2 and on different channel estimation methods for the proposed DCM
mRF r = 4, and for 5 bits transmission with M = 256, mRF s = 2 and scheme. In addition, the capacity analysis of DCM scheme may be
mRF r = 6. As it can be seen from Fig. 7, the derived analytical SER also considered in a future research work. Finally, considering the
performance matches closely with computer simulation results of activation of more than one relay, in addition to other practical
DCM-ML scheme, especially at high SNR region. impairments, will be also considered in a future work.
8
T.Y. Elganimi, K.M. Rabie and A.M. Abu-Hudrouss Physical Communication 57 (2023) 101977

Fig. 5. SER performance of the proposed DCM scheme using ML detection technique in the presence of CEEs. All schemes are equipped with 4 relays and NRr = 2
receive antennas at each relay: (a) DCM scheme with M = 256, MR = 4, mRF s = 2 and mRF r = 6 for η = 5 bpcu; (b) DCM scheme with M = 256, MR = 4, mRF s = 4
and mRF r = 8 for η = 6 bpcu.

Fig. 6. SER performance of the proposed DCM scheme using i-MRC detection technique in the presence of CEEs. All schemes are equipped with 4 relays and NRr = 2
receive antennas at each relay: (a) DCM scheme with M = 256, MR = 4, mRF s = 2 and mRF r = 6 for η = 5 bpcu; (b) DCM scheme with M = 256, MR = 4, mRF s = 4
and mRF r = 8 for η = 6 bpcu.

Fig. 7. Comparison of analytical and simulation results of SER performance of the proposed DCM-ML scheme in the presence of CEEs with R = 4 relays and NRr = 4
receive antennas are equipped at each relay, and MR = 4: (a) For 8 bpcu with M = 64, mRF s = 2 and mRF r = 4; (b) For 10 bpcu with M = 256, mRF s = 2 and
mRF r = 6.

9
T.Y. Elganimi, K.M. Rabie and A.M. Abu-Hudrouss Physical Communication 57 (2023) 101977

Declaration of competing interest [20] M. Wen, B. Ye, E. Basar, Q. Li, F. Ji, Enhanced orthogonal frequency division
multiplexing with index modulation, IEEE Trans. Wireless Commun. 16 (7)
(2017) 4786–4801, http://dx.doi.org/10.1109/TWC.2017.2702618.
The authors declare that they have no known competing finan-
[21] J. Li, S. Dang, M. Wen, X.-Q. Jiang, Y. Peng, H. Hai, Layered orthogonal
cial interests or personal relationships that could have appeared frequency division multiplexing with index modulation, IEEE Syst. J. 13
to influence the work reported in this paper. (4) (2019) 3793–3802, http://dx.doi.org/10.1109/JSYST.2019.2918068.
[22] M. Wen, Q. Li, E. Basar, W. Zhang, Generalized multiple-mode OFDM
Data availability with index modulation, IEEE Trans. Wireless Commun. 17 (10) (2018)
6531–6543, http://dx.doi.org/10.1109/TWC.2018.2860954.
[23] S. Jacob, T.L. Narasimhan, A. Chockalingam, Space-time index modula-
The data that has been used is confidential. tion, in: 2017 IEEE Wireless Communications and Networking Conference
(WCNC), 2017, pp. 1–6, http://dx.doi.org/10.1109/WCNC.2017.7925922.
References [24] T.Y. Elganimi, K.M. Rabie, Time-indexed parallel spatial modulation for
large-scale MIMO systems with antenna grouping, in: 2021 IEEE 93rd
[1] A. Nosratinia, T. Hunter, A. Hedayat, Cooperative communication in wire- Vehicular Technology Conference (VTC2021-Spring), 2021, pp. 1–5, http:
less networks, IEEE Commun. Mag. 42 (10) (2004) 74–80, http://dx.doi.org/ //dx.doi.org/10.1109/VTC2021-Spring51267.2021.9449015.
10.1109/MCOM.2004.1341264. [25] A.K. Khandani, Media-based modulation: A new approach to wireless
[2] J. Laneman, D. Tse, G. Wornell, Cooperative diversity in wireless networks: transmission, in: 2013 IEEE International Symposium on Information
Efficient protocols and outage behavior, IEEE Trans. Inf. Theory 50 (12) Theory, 2013, pp. 3050–3054, http://dx.doi.org/10.1109/ISIT.2013.6620786.
(2004) 3062–3080, http://dx.doi.org/10.1109/TIT.2004.838089. [26] A.K. Khandani, Media-based modulation: Converting static Rayleigh fading
[3] S. Narayanan, M. Di Renzo, F. Graziosi, H. Haas, Distributed spatial to AWGN, in: 2014 IEEE International Symposium on Information Theory,
modulation for relay networks, in: 2013 IEEE 78th Vehicular Technology 2014, pp. 1549–1553, http://dx.doi.org/10.1109/ISIT.2014.6875093.
Conference (VTC Fall), 2013, pp. 1–6, http://dx.doi.org/10.1109/VTCFall. [27] E. Basar, M. Wen, R. Mesleh, M. Di Renzo, Y. Xiao, H. Haas, Index
2013.6692167. modulation techniques for next-generation wireless networks, IEEE Access
[4] S. Narayanan, M. Di Renzo, F. Graziosi, H. Haas, Distributed spatial 5 (2017) 16693–16746, http://dx.doi.org/10.1109/ACCESS.2017.2737528.
modulation: A cooperative diversity protocol for half-duplex relay-aided [28] B. Shamasundar, S. Jacob, S. Bhat, A. Chockalingam, Multidimensional index
wireless networks, IEEE Trans. Veh. Technol. 65 (5) (2016) 2947–2964, modulation in wireless communications, in: 2017 Information Theory and
http://dx.doi.org/10.1109/TVT.2015.2442754. Applications Workshop (ITA), 2017, pp. 1–10, http://dx.doi.org/10.1109/ITA.
[5] K. Sankhe, S. Chaudhari, G.R. Murthy, Distributed spatial modulation with 2017.8023446.
dynamic frequency allocation, Phys. Commun. 23 (2017) 65–75, http: [29] Y. Naresh, A. Chockalingam, On media-based modulation using RF mirrors,
//dx.doi.org/10.1016/j.phycom.2017.02.005. IEEE Trans. Veh. Technol. 66 (6) (2017) 4967–4983, http://dx.doi.org/10.
[6] Y. Rong, X. Tang, Y. Hua, A unified framework for optimizing linear 1109/TVT.2016.2620989.
nonregenerative multicarrier MIMO relay communication systems, IEEE [30] R. Mesleh, S.S. Ikki, H.M. Aggoune, Quadrature spatial modulation, IEEE
Trans. Signal Process. 57 (12) (2009) 4837–4851, http://dx.doi.org/10.1109/ Trans. Veh. Technol. 64 (6) (2015) 2738–2742, http://dx.doi.org/10.1109/
TSP.2009.2027779. TVT.2014.2344036.
[7] W. Xu, X. Dong, Y. Huang, Asymptotic achievable rate analysis for selection [31] I. Yildirim, E. Basar, I. Altunbas, Quadrature channel modulation, IEEE
strategies in amplify-and-forward MIMO two-hop networks with feedback, Wirel. Commun. Lett. 6 (6) (2017) 790–793, http://dx.doi.org/10.1109/LWC.
IEEE Trans. Veh. Technol. 59 (7) (2010) 3662–3668, http://dx.doi.org/10. 2017.2742508.
1109/TVT.2010.2054844. [32] E. Basar, I. Altunbas, Space-time channel modulation, IEEE Trans. Veh.
[8] W. Xu, X. Dong, W.-S. Lu, Joint precoding optimization for multiuser multi- Technol. 66 (8) (2017) 7609–7614, http://dx.doi.org/10.1109/TVT.2017.
antenna relaying downlinks using quadratic programming, IEEE Trans. 2674689.
Commun. 59 (5) (2011) 1228–1235, http://dx.doi.org/10.1109/TCOMM. [33] N. Pillay, H. Xu, Uncoded space-time labeling diversity—Application of
2011.041111.090713. media-based modulation with RF mirrors, IEEE Commun. Lett. 22 (2)
[9] M. Wen, B. Zheng, K.J. Kim, M. Di Renzo, T.A. Tsiftsis, K.-C. Chen, N. Al- (2018) pp. 272–275, http://dx.doi.org/10.1109/LCOMM.2017.2770145.
Dhahir, A survey on spatial modulation in emerging wireless systems: [34] B.S. Adejumobi, N. Pillay, RF mirror media-based space-time block coded
Research progresses and applications, IEEE J. Sel. Areas Commun. 37 (9) spatial modulation techniques for two time-slots, IET Commun. 13 (15)
(2019) 1949–1972, http://dx.doi.org/10.1109/JSAC.2019.2929453. (2019) 2313–2321.
[10] R. Mesleh, H. Haas, C.W. Ahn, S. Yun, Spatial modulation - A new [35] E. Başar, U. Aygölü, E. Panayirci, H.V. Poor, Space-time block coded spatial
low complexity spectral efficiency enhancing technique, in: 2006 First modulation, IEEE Trans. Commun. 59 (3) (2011) pp. 823–832, http://dx.
International Conference on Communications and Networking in China, doi.org/10.1109/TCOMM.2011.121410.100149.
2006, pp. 1–5, http://dx.doi.org/10.1109/CHINACOM.2006.344658. [36] B.S. Adejumobi, T. Shongwe, Labeling diversity for media-based space-time
[11] R.Y. Mesleh, H. Haas, S. Sinanovic, C.W. Ahn, S. Yun, Spatial modulation, block coded spatial modulation, IEEE Access 8 (2020) pp. 99870–99879,
IEEE Trans. Veh. Technol. 57 (4) (2008) 2228–2241, http://dx.doi.org/10. http://dx.doi.org/10.1109/ACCESS.2020.2995945.
1109/TVT.2007.912136. [37] T.Y. Elganimi, K.M. Rabie, Multidimensional generalized quadrature index
[12] M. Di Renzo, H. Haas, Bit error probability of SM-MIMO over generalized modulation for 5G wireless communications, in: 2021 IEEE 93rd Vehicular
fading channels, IEEE Trans. Veh. Technol. 61 (3) (2012) 1124–1144, http: Technology Conference (VTC2021-Spring), 2021, pp. pp. 1–6, http://dx.doi.
//dx.doi.org/10.1109/TVT.2012.2186158. org/10.1109/VTC2021-Spring51267.2021.9448997.
[13] J. Li, M. Wen, X. Cheng, Y. Yan, S. Song, M.H. Lee, Generalized precoding- [38] R. Tsvaki, N. Pillay, H. Xu, Double spatial media-based modulation with
aided quadrature spatial modulation, IEEE Trans. Veh. Technol. 66 (2) RF mirrors, in: Annals of Telecommunications, Vol. 20, Springer, 2021,
(2017) 1881–1886, http://dx.doi.org/10.1109/TVT.2016.2565618. http://dx.doi.org/10.1007/s12243-021-00841-3.
[14] T.Y. Elganimi, F.I. Alwerfly, A.A. Marseet, Distributed generalized spatial [39] Z. Yigit, E. Basar, Double spatial modulation: A high-rate index modulation
modulation for relay networks, in: 2020 28th Signal Processing and scheme for MIMO systems, in: 2016 International Symposium on Wireless
Communications Applications Conference (SIU), Sep. 2020, pp. 1–4, http: Communication Systems (ISWCS), 2016, pp. 347–351, http://dx.doi.org/10.
//dx.doi.org/10.1109/SIU49456.2020.9302199. 1109/ISWCS.2016.7600927.
[15] A. Younis, N. Serafimovski, R. Mesleh, H. Haas, Generalised spatial modula- [40] Y. Naresh, A. Chockalingam, A low-complexity maximum-likelihood detec-
tion, in: 2010 Conference Record of the Forty Fourth Asilomar Conference tor for differential media-based modulation, IEEE Commun. Lett. 21 (10)
on Signals, Systems and Computers, 2010, pp. 1498–1502, http://dx.doi. (2017) 2158–2161, http://dx.doi.org/10.1109/LCOMM.2017.2687921.
org/10.1109/ACSSC.2010.5757786. [41] B. Shamasundar, S. Jacob, L.N. Theagarajan, A. Chockalingam, Media-based
[16] N. Serafimovski, S. Sinanovic, M. Di Renzo, H. Haas, Dual-hop spatial modulation for the uplink in massive MIMO systems, IEEE Trans. Veh.
modulation (Dh-SM), in: 2011 IEEE 73rd Vehicular Technology Conference Technol. 67 (9) (2018) 8169–8183, http://dx.doi.org/10.1109/TVT.2018.
(VTC Spring), Yokohama, Japan, May 2011, pp. 1–5, http://dx.doi.org/10. 2839706.
1109/VETECS.2011.5956586. [42] Y. Naresh, A. Chockalingam, Performance analysis of full-duplex decode-
[17] R. Mesleh, S. Ikki, H. Aggoune, A. Mansour, Performance analysis of space and-forward relaying with media-based modulation, IEEE Trans. Veh.
shift keying (SSK) modulation with multiple cooperative relays, EURASIP Technol. 68 (2) (2019) 1510–1524, http://dx.doi.org/10.1109/TVT.2018.
J. Adv. Signal Process. 201 (2012) (2012). 2887062.
[18] P. Som, A. Chockalingam, Performance analysis of space-shift keying in [43] F. Yarkin, I. Altunbas, E. Basar, Cooperative space shift keying media-based
decode-and-forward multihop MIMO networks, IEEE Trans. Veh. Technol. modulation with hybrid relaying, IEEE Syst. J. 14 (1) (2020) 500–509,
64 (1) (2015) 132–146, http://dx.doi.org/10.1109/TVT.2014.2318437. http://dx.doi.org/10.1109/JSYST.2019.2899876.
[19] E. Başar, U. Aygölü, E. Panayırcı, H.V. Poor, Orthogonal frequency division [44] J. Luo, H. Wang, F. Wang, S. Wang, Secure spatial modulation via radio
multiplexing with index modulation, IEEE Trans. Signal Process. 61 (22) frequency mirrors, IEEE Trans. Veh. Technol. 69 (8) (2020) 9168–9173,
(2013) 5536–5549, http://dx.doi.org/10.1109/TSP.2013.2279771. http://dx.doi.org/10.1109/TVT.2020.2999917.

10
T.Y. Elganimi, K.M. Rabie and A.M. Abu-Hudrouss Physical Communication 57 (2023) 101977

[45] J.G. Proakis, Digital communications McGraw-Hill series in electrical and His current research interests focus on designing and developing next-generation
computer engineering, in: Computer Engineering, McGraw-Hill, 2001. wireless communication systems.
[46] M.K. Simon, M.-S. Alouini, Digital Communication over Fading Channels, Khaled serves regularly on the technical program committee (TPC) for several
Vol. 95, John Wiley & Sons, 2005. major IEEE conferences, such as GLOBECOM, ICC, and VTC. He has received many
awards over the past few years in recognition of his research contributions
including the Best Paper Awards at the 2021 IEEE CITS and the 2015 IEEE
Taissir Y. Elganimi (Senior Member, IEEE) was born ISPLC, and the IEEE ACCESS Editor of the month award for August 2019. He
in Tripoli, Libya, in 1988. He received his B.Sc. degree is currently serving as an Editor of IEEE COMMUNICATIONS LETTERS, an Editor
in Electronics and Communication Engineering from of IEEE Internet of Things Magazine, an Associate Editor of IEEE ACCESS and an
Department of Electrical and Electronic Engineering at Executive Editor of the TRANSACTIONS ON EMERGING TELECOMMUNICATIONS
University of Tripoli, Libya, in 2010, and his M.Sc. TECHNOLOGIES (Wiley). Khaled is also a Fellow of the U.K. Higher Education
degree in Wireless Communications (with Distinction) Academy (FHEA).
from University of Southampton, UK, in 2015. He is
currently working as a lecturer in the Department of
Electrical and Electronic Engineering at University of
Tripoli, Libya. Ammar M. Abu-Hudrouss (Senior Member, IEEE) was
Taissir serves as a technical reviewer for several born in Khan-Younis, Palestine, in 1977. He received
IEEE transaction journals, and has been a member of Technical Program Com- the B.Sc. degree from Islamic University Gaza (IUG),
mittees (TPC) for several IEEE conferences such as ICC, WCNC, GLOBECOM, etc. Palestine, in 2000. He received the M.Sc. degree in
He is a Publicity Co-chair of the 8th EAI International Conference on IoT as a Telecommunication Engineering and the Ph.D. de-
Service (EAI IoTaaS 2022). He has received many awards over the past few years gree in Communication Engineering from Birmingham
including the Best Paper Award at the IEEE International Conference on Com- University, Birmingham, U.K., in 2003 and 2007, re-
puter, Information and Telecommunication Systems (CITS 2021). His research spectively. He was a visiting researcher at University
interests mainly include multi-functional MIMO, space modulation techniques, of York from 9/2012 to 9/2013 as a holder of Distin-
multidimensional index modulation, optical communications, millimeter-wave guished Scholar Award from the Arab Fund for Social
massive MIMO communications, and reconfigurable intelligent surface-assisted and Economic Development. He is currently a Professor
MIMO systems for 6G communications. of Communications at IUG, Palestine. He is also a member of CTRC, PICTRA.
During his work at IUG, Prof. Abu-Hudrouss has been granted to presume
several short scientific visits to international universities in European and
Khaled M. Rabie (Senior Member, IEEE) received the
Canada. Prof. Abu-Hudrouss is currently a member of many local and national
Ph.D. degree in Electrical and Electronic Engineering
committees and councils including Palestinian Engineering Syndicate, and the
from the University of Manchester in 2015. Prior to
Palestine Academy for Science and Technology. His current research interests
that, he received the M.Sc. degree (with Distinction) in
are localization, digital signal processing for wireless communications, software
Communication Engineering and the B.Sc. degree (with
defined radio, index modulation, and coding.
Hons.) in Electrical and Electronic Engineering from
the University of Manchester and University of Tripoli,
respectively, securing first rank in both degrees. He is
currently a Reader with the Department of Engineering
at Manchester Metropolitan University (MMU), UK. He
has worked as a part of several largescale industrial
projects and has published 200+ journal and conference articles (mostly IEEE).

11