SM Mimo Mod

1124
IEEE TRANSACTIONS ON VEHICULAR TECHNOLOGY, VOL. 61, NO. 3, MARCH 2012
Bit Error Probability of SM-MIMO Over

Generalized Fading Channels
Marco Di Renzo, Member, IEEE, and Harald Haas, Member, IEEE
AbstractIn this paper, we study the performance of spatial

modulation (SM) multiple-inputmultiple-output (MIMO) wireless systems over generic fading channels. More precisely, a comprehensive analytical framework to compute the average bit error
probability (ABEP) is introduced, which can be used for any
MIMO setup, for arbitrary correlated fading channels, and for
generic modulation schemes. It is shown that, when compared with
state-of-the-art literature, our framework 1) has more general
applicability over generalized fading channels, 2) is, in general,
more accurate as it exploits an improved union-bound method,
and, 3) more importantly, clearly highlights interesting fundamental trends about the performance of SM, which are difficult
to capture with available frameworks. For example, by focusing
on the canonical reference scenario with independent identically
distributed Rayleigh fading, we introduce very simple formulas
that yield insightful design information on the optimal modulation
scheme to be used for the signal constellation diagram, as well
as highlight the different roles played by the bit mapping on
the signal and spatial constellation diagrams. Numerical results
show that, for many MIMO setups, SM with phase-shift-keying
(PSK) modulation outperforms SM with quadrature-amplitude
modulation (QAM), which is a result never reported in the literature. In addition, by exploiting asymptotic analysis, closed-form
formulas of the performance gain of SM over other single-antenna
transmission technologies are provided. Numerical results show
that SM can outperform many single-antenna systems and that,
for any transmission rate, there is an optimal allocation of the
information bits onto spatial and signal constellation diagrams.
Furthermore, by focusing on the Nakagami-m fading scenario
with generically correlated fading, we show that fading severity
plays a very important role in determining the diversity gain
of SM. In particular, the performance gain over single-antenna
systems increases for fading channels less severe than Rayleigh
fading, whereas it gets smaller for more severe fading channels.
In addition, it is shown that the impact of fading correlation at
the transmitter is reduced for less severe fading. Finally, analytical
frameworks and claims are substantiated through extensive Monte
Carlo simulations.
Manuscript received August 25, 2011; revised December 15, 2011; accepted
January 17, 2011. Date of publication January 31, 2012; date of current
version March 21, 2012. This paper was presented in part at the IEEE/ICST
International Conferrence on Communications and Networking in China, Beijing, China, August 2010. This work was supported in part by the European
Union under PITN-GA-2010-264759, GREENET project. The work of M.
Di Renzo was supported by the Laboratory of Signals and Systems under
research project Jeunes Chercheurs. The work of H. Haas was supported in
part by the Engineering and Physical Sciences Research Council under Grant
EP/G011788/1. The review of this paper was coordinated by Dr. E. K. S. Au.
M. Di Renzo is with the Laboratoire des Signaux et Systmes, Unit
Mixte de Recherche 8506, Centre National de la Recherche Scientifique-cole
Suprieure dlectricit-Universit Paris-Sud XI, 91192 Gif-sur-Yvette Cedex,
France (e-mail: marco.direnzo@lss.supelec.fr).
H. Haas is with the Institute for Digital Communications (IDCOM), School
of Engineering, College of Science and Engineering, The University of
Edinburgh, EH9 3JL Edinburgh, UK (e-mail: h.haas@ed.ac.uk).
Color versions of one or more of the figures in this paper are available online
at http://ieeexplore.ieee.org.
Digital Object Identifier 10.1109/TVT.2012.2186158
Index TermsLarge-scale antenna systems, massive

multiple-inputmultiple-output (MIMO) systems, performance
analysis, single-RF MIMO design, spatial modulation (SM).
I. I NTRODUCTION
PATIAL MODULATION (SM) is a digital modulation

concept for multiple-inputmultiple-output (MIMO) wireless systems that has recently been introduced to increase
the data rate of single-antenna systems by keeping a lowcomplexity transceiver design and by requiring no bandwidth
expansion [1][4]. Unlike conventional spatial multiplexing
schemes [2], [5], in SM, the multiplexing gain is realized through mapping a block of information bits into two
information-carrying units: 1) the conventional signal constellation diagram [e.g., phase-shift-keying (PSK) or quadratureamplitude modulation (QAM)] and 2) the so-called spatial constellation diagram, which is the antenna array at the transmitter.
Like in conventional modulation schemes, the first subblock of
information bits determines the point of the signal constellation
diagram that is actually transmitted. Specific to SM is that the
second subblock identifies the single active transmit antenna.
As a result, the point of the signal constellation diagram is transmitted through a single active antenna belonging to the spatial
constellation diagram. This simple modulation concept brings
two main advantages: 1) For each channel use, the data rate
increases by a factor equal to the logarithm of the number of
antennas at the transmitter [2], [5], and 2) the receiver can detect
the whole block of bits through single-stream demodulation,
as the second subblock of bits is only implicitly transmitted
through activation of the transmit antenna [6]. With respect to
single-antenna systems, the net gain is a multiplexing gain for
the same decoding complexity, whereas the price to pay is the
need of more antennas at the transmitter. Recent results have
showcased the performance gain of SM with respect to other
state-of-the-art transmission technologies for single- and multiantenna systems [5][19]. Finally, it is worth mentioning that
SM seems to be an appealing transmission technology for highrate and low-complexity MIMO implementations that exploit
the recently proposed massive MIMO or large-scale antenna
systems paradigm [20], [21]. In fact, in these systems, it is
envisaged that improved performance and energy efficiency can
be achieved by using tens or hundreds of antenna elements at
the base station instead of exploiting base station cooperation.
In this perspective, SM can be regarded as a low-complexity
modulation scheme that exploits the massive MIMO idea but
with a single active RF chain. The design of MIMO schemes
that retain the benefits of multiple-antenna transmission while
having a single active RF element is another recent and major
trend in current and future MIMO research [22].
0018-9545/$31.00 2012 IEEE
DI RENZO AND HAAS: BIT ERROR PROBABILITY OF SM-MIMO OVER GENERALIZED FADING CHANNELS
Since its introduction, many researchers have been studying the performance of SM-MIMO over fading channels
either through time-consuming Monte Carlo simulations or
through analytical modeling. Despite being more challenging,
analytical modeling is, in general, preferred because of the
following:
1) It allows a deeper understanding of the system performance.
2) It enables a simpler comparison with other competing
transmission technologies.
3) It provides opportunities for system optimization.
A careful look at current state-of-the-art reveals the following contributions. The vast majority of analytical frameworks
are useful for a special case of SM, which is called spaceshift-keying (SSK) modulation [8]. SSK modulation is a lowcomplexity and low-data-rate version of SM where only the
spatial constellation diagram is exploited for data modulation.
This transmission technology is extensively studied in [23]
[29] for various MIMO setups and channel models. The analytical study conducted in [23][29] has highlighted fundamental
properties of the spatial constellation diagram with respect to
fading severity, channel correlation, power imbalance, transmit
diversity, as well as robustness to multiple-access interference
and channel estimation errors. However, these frameworks are
of limited use to understand the performance of SM as the
signal constellation diagram is neglected. On the other hand,
analytical modeling of SM is limited to a very few papers,
which have various limitations. In [5] and [30], the authors
study a suboptimal receiver design, and the symbol error
probability is computed by resorting to numerical integrations,
which are not easy to compute and, in some cases, are numerically unstable. In [6], the authors study the average bit
error probability (ABEP) of the maximum-likelihood (ML) optimum receiver over independent identically distributed (i.i.d.)
Rayleigh fading. The framework is based on the union-bound
method. Due to the absence of a scaling factor in the final
formula [31], this bound is rather weak. Furthermore, and
more importantly, the framework is valid for real-valued signal
constellation points, and thus, it cannot be used, e.g., for PSK
and QAM. In [9], the authors provide a first closed-form framework to compute the ABEP of SM-MIMO over generically
correlated Rician fading and for arbitrary modulation schemes.
In addition, the framework highlights some fundamental behaviors of SM, such as its incapability to achieve transmit diversity [17]. However, [9] has the following important
limitations.
1) The analysis is applicable to Rician fading only.
2) The framework is based on conventional union-bound
methods, whose accuracy degrades for high modulation
orders and small numbers of receive antennas.
3) Signal and spatial constellation diagrams are treated as a
single entity, which does not highlight the role played by
each of them for various fading channels and modulation
schemes.
In this depicted context, this paper is aimed at proposing
a comprehensive analytical framework to study the ABEP of
1125
SM-MIMO over generalized fading channels. More specifically, we are interested in studying the following:
1) the interplay of signal and spatial constellation diagrams
and whether an optimal allocation of the information bits
between them exists;
2) the effect of adding the spatial constellation diagram
on top of the signal constellation diagram, and whether
conventional signal modulation schemes (e.g., PSK and
QAM) are the best choice for SM, or whether new
optimal modulation schemes should be designed to fully
exploit the benefits of this hybrid modulation scheme;
3) the advantages and disadvantages of SM with respect to
conventional single-antenna PSK/QAM and SSK modulations as a function of the MIMO setup and fading
scenario.
To this end, we propose a new analytical framework that
foresees to write the ABEP as the summation of the following
three contributions:
1) a term that mainly depends on the signal constellation
diagram;
2) a term that mainly depends on the spatial constellation
diagram;
3) a joint term that depends on both constellation diagrams
and highlights their interactions.
This new analytical formulation allows us to introduce an
improved union-bound method that is more accurate than
conventional union-bound methods and enables a deeper understanding of the role played by both information carrying
units for various channel models and MIMO setups. Some of
the most important and general results of this paper are as
follows.
1) We show that SM outperforms single-antenna PSK/QAM
schemes only for data rates greater than 2 bits per channel
use (bpcu), and that SM with QAM-modulated points
in the signal constellation diagram is never superior to
single-antenna QAM for single-antenna receivers. On the
other hand, for multi-antenna receivers and higher data
rates, SM can significantly outperform single-antenna
PSK/QAM. Closed-form expressions of this performance
gain over i.i.d. Rayleigh fading are given.
2) Unlike single-antenna systems, where QAM always outperforms PSK, we show that SM with PSK-modulated
points in the signal constellation diagram can outperform
SM with QAM-modulated points. This is due to the
interactions of signal and spatial constellation diagrams,
and for i.i.d. Rayleigh fading, we show analytically that
the ABEP of SM does not depend only on the Euclidean
distance of the points in the signal constellation diagram.
This provides important information on how to conceive
new modulation schemes that are specifically optimized
for SM.
3) By considering, as a case study, Nakagami-m fading, we
show that the fading severity mNak plays an important
role on the performance of SM. More specifically, like
conventional modulation schemes, the ABEP gets worse
for wireless channel with fading more severe (0.5
mNak < 1) than Rayleigh. However, with respect to
1126
single-antenna PSK/QAM modulation, the performance

gain of SM increases thanks to the higher diversity gain
experienced by the information bits mapped onto the
spatial constellation diagram. On the contrary, the performance gain decreases for less severe (mNak > 1) fading
because the diversity gain of the bits mapped onto the
spatial constellation diagram is independent of the fading
parameter mNak . In addition, it is shown that channel
correlation at the transmitter has less impact when 0.5
mNak < 1.
The remainder of this paper is organized as follows. In
Section II, the system model is described. In Section III, the
improved union bound is introduced, and the specific contribution of spatial and signal constellation diagrams is shown. In
Section IV, closed-form expressions of the ABEP for various
fading channels and modulation schemes are provided. In Section V, the canonical i.i.d. Rayleigh fading channel is studied in
detail, and closed-form expressions of the gain with respect to
single-antenna PSK/QAM and SSK modulations are given. In
Section VI, numerical results are shown to substantiate claims
and analytical derivations. Finally, Section VII concludes this
paper.
II. S YSTEM M ODEL
We consider a generic Nt Nr MIMO system, where Nt
and Nr denote the antennas at the transmitter and receiver,
respectively. We assume that the transmitter can send digital
information via M complex symbols. In SM literature [4],
the set of Nt antennas (nt = 1, 2, . . . , Nt ) is called spatial
constellation diagram, whereas the set of M complex points
(l for l = 1, 2, . . . , M ) is called signal constellation diagram.
The basic idea of SM is to map blocks of information bits
into two information carrying units [5]: 1) a symbol, which is
chosen from the complex signal constellation diagram, and 2) a
single active transmit antenna, which is chosen from the spatial
constellation diagram.
More specifically, SM works as follows. At the transmitter,
the bitstream is divided into blocks containing log2 (Nt ) +
log2 (M ) bits each, with log2 (Nt ) and log2 (M ) being the
number of bits needed to identify a transmit antenna in the
spatial constellation diagram and a symbol in the signal constellation diagram, respectively. Each block is split into two
subblocks of log2 (Nt ) and log2 (M ) bits each. The bits in the
first subblock are used to select the transmit antenna that is
switched on for transmission, whereas all the other antennas are
kept silent. The bits in the second subblock are used to choose a
symbol in the signal constellation diagram. At the receiver, the
detector can recover the whole block of log2 (Nt ) + log2 (M )
information bits by solving an Nt M hypothesis detection
problem, which jointly estimates the transmit antenna that is
not idle and the signal waveform that is transmitted from it.
In this paper, the generic block of log2 (Nt ) + log2 (M ) bits
is called message, and it is denoted by (nt , l ), where nt =
1, 2, . . . , Nt and l = 1, 2, . . . , M univocally identify the active
transmit antenna nt and the complex symbol l transmitted
from it, respectively. The Nt M messages are assumed to be
equiprobable.
A. Notation
Throughout this paper, we use the notation as follows.
1) We adopt
a complex envelope signal representation.
2) j = 1 is the imaginary unit.
3) () is the complex-conjugate operator.
+
4) (x y)(t) = x()y(t )d is the convolution of
x() and y().
5) | | is the absolute value.
6) E{} is the expectation operator.
7) Re{} and Im{} are real and imaginary part operators.
+
8) (x) = 0 x1 exp()d is the Gamma function.
+
9) Q(x) = (1/ 2) x exp(t2 /2)dt is the Q-function.
10) Em is the average energy per transmission.
11) Tm is the transmission time slot of each message.
+
12) w() is the unit energy, i.e., |w(t)|2 dt = 1, elementary transmitted pulse waveform that is nonzero only in
[0, Tm ].
from
13) The signal related to (nt , l ) and transmitted
antenna nt is denoted by s(t|(nt , l )) = Em l w(t).

14) The generic point of the signal constellation diagram
I
exp(jl ), where
l is defined as l = R
l + jl = l
I
R
l = Re{l }, l = Im{l }, l =
I 2
2
(R
l ) + (l ) ,
arctan(Il /R
l ).
and l =
15) Pr{} denotes probability.
16) The noise nr at the input of the nr th (nr = 1, 2, . . . , Nr )
receive antenna is a complex additive white Gaussian
noise (AWGN) process with power spectral density N0
per dimension. Across the receive antennas, the noises
nr are statistically independent.
17) We introduce = Em /(4N0 ).
18) () is the Dirac delta function.
19) x is the function that rounds x to the closest integer.
20) is the
function.
floor
)
(a
21) Gm,n
.| p
is the Meijer-G function defined in [32,
p,q
(bq )
Ch. 8, p. 519].
22) MX (s) = E{exp(sX)} is the moment generating
function (MGF) of random variable (RV) X.
d
23) X = Y denotes that the RVs X and Y are equal in
distribution or law, i.e., they have the same probability
density function (pdf).
24) (x!) is the factorial of x.
25) ()1 is the inverse of a square matrix.
26) Iv () is the modified Bessel function of first kind and
v [33, Ch. 9].
order
27) is the binomial coefficient.
28) NH ((
nt , l ) (nt , l )) is the Hamming distance of
messages (
nt , l ) and (nt , l ), i.e., the number of
positions where the information bits are different, with
nt , l ) (nt , l )) log2 (Nt M ).
0 NH ((
B. Channel Model
We consider the frequency-flat slowly-varying fading channel model as follows:
1) hnt ,nr () = nt ,nr ( nt ,nr ) is the channel impulse
response of the wireless link from the nt th transmit
antenna to the nr th receive antenna. nt ,nr = nRt ,nr +

jnI t ,nr = nt ,nr exp(jnt ,nr ) is the complex channel
gain, and nt ,nr is the propagation time delay. No
specific distribution for the channel envelopes nt ,nr =

(nRt ,nr )2 + (nI t ,nr )2 , the channel phases nt ,nr =
arctan(nI t ,nr /nRt ,nr ), and nRt ,nr = Re{nt ,nr },
nI t ,nr = Im{nt ,nr } is assumed a priori.
2) The delays nt ,nr are assumed to be known at the receiver, i.e., perfect time synchronization is considered. In
addition, we assume 1,1
= 1,2
= ...
= Nt ,Nr , which
is a realistic assumption when the distance between transmitter and receiver is much larger than the spacing of
the antenna elements [24]. Due to these assumptions, the
delays nt ,nr are neglected in the next sections.
C. ML Optimum Detector
Let (
nt , l ) be the transmitted message.1 The signal rent , l ) is transmitted, is
ceived by the nr th receive antenna, if (

nt , l ) + nr (t)
znr (t) = sch,nr t| (
(1)
nt , l )) = (s(|(
nt , l )) hn t ,nr )(t) =
where
sch,nr (t|(
Em n t ,nr l w(t) for n

t = 1, 2, . . . , Nt , nr = 1, 2, . . . , Nr ,
and l = 1, 2, . . . , M .
Equation (1) is a general Nt M hypothesis detection
problem [34, Sec. 7.1], [35, Sec. 4.2, pp. 257] in AWGN
when conditioning upon fading channel statistics. Thus, the
ML optimum detector with full channel state information and
perfect time synchronization at the receiver is given by [6],
[34, Sec. 7.1]
(
nt , l ) =
{D(nt , l )}
arg max
for nt =1,2,...,Nt and l=1,2,...,M
arg max
for nt =1,2,...,Nt and l=1,2,...,M
1 We emphasize that symbols with

identify the actual message that is
transmitted, while symbols without denote the trial message that is tested by
the detector to solve the Nt M -hypothesis detection problem. Also, symbols
with denote the message estimated by the detector. This notation does not
apply to the antenna-index nr at the receiver since there is no hypothesis-testing
in this case.

ABEP = E
1127

Nr
znr (t)sch,n (t|(nt , l )) dt

r
nr =1 T
m
|sch,nr (t|(nt , l ))|2 dt . (2)
Tm
The outcome of (2) is the estimated message (

nt , l ). Thus,
the receiver is successful in decoding the whole block of bits if
nt , l ), i.e., n
t = n
t and l = l .
and only if (
nt , l ) = (
III. AVERAGE B IT E RROR P ROBABILITY OVER
G ENERALIZED FADING C HANNELS :
I MPROVED U NION B OUND
The exact ABEP of the detector in (2) can be computed
in closed-form for arbitrary fading channels and modulation
schemes as in (3), shown at the bottom of the page [36,
eqs. (4) and (5)], where 1) is a shorthand to denote the
set of Nt Nr complex channel gains, i.e., nt ,nr for nt =
1, 2, . . . , Nt , nr = 1, 2, . . . , Nr , and 2) E {} is the expectation computed over all the fading channels.
For arbitrary MIMO systems, the estimation of
nt , l )} is very complicated as it
Pr{(
nt , l ) = (nt , l )|(
requires, in general, the computation of multidimensional
integrals. Because of that, it is common practice to exploit
union-bound methods [34] to compute the ABEP in (3) as in
(4), shown at the bottom of the page, where APEP((
nt , l )
nt , l ) (nt , l )}} is the average
(nt , l )) = E(nt ,nt ) {Pr{(
pairwise error probability (APEP), i.e., the probability of
nt , l ) is transmitted,
detecting (nt , l ) when, instead, (
nt , l ) are the only
under the assumption that (nt , l ) and (
two messages possibly being transmitted, and E(nt ,nt ) {} is
the expectation computed over the fading channels from the
t th transmit antennas and the Nr receive antennas.
nt th and n
The APEP is equal to [26, eqs. (10)(11)]

APEP (
nt , l ) (nt , l )

nt , l ) < D(nt , l )
= E(nt ,nt ) Pr D(

Nr

= E(nt ,nt ) Q
|n t ,nr l nt ,nr l |2 .
(5)
nr=1
Nt
Nt
M
M

1
1
nt , l )
NH (
nt , l ) (nt , l ) Pr (
nt , l ) = (nt , l ) (
Nt M log2 (Nt M ) n =1
(3)
Nt
Nt

1
1
NH (
nt , l ) (nt , l ) APEP (
nt , l ) (nt , l )
Nt M log2 (Nt M ) n =1
(4)
t =1
l=1 n
l=1
ABEP
t =1
l=1 n
l=1
1128
The union bound in (4) has been used in [6],2 [9], and
[24][26]. However, as mentioned in Section I, it has some
limitations, as follows.
1) The roles played by spatial and signal constellation diagrams (and the related bit mapping) are hidden in the
fourfold summation.
2) It is not accurate enough for large M and small Nr [37],
as shown in Section VI.
3) Its computational complexity is the same, irrespective of
modulation scheme and fading channel, when, instead,
simpler formulas can be obtained in several cases.
A. Improved Upper Bound
To avoid the limitations of the conventional union bound
when used for performance analysis of SM, and, more importantly, to get more insights about the expected performance of
SM, we propose an improved upper bound. The new bound is
summarized in Proposition 1.
Proposition 1: The ABEP in (3) can be tightly upper
bounded as follows:
ABEP ABEPsignal + ABEPspatial + ABEPjoint
(6)
where ABEPsignal , ABEPspatial , and ABEPjoint are defined as
Nt
"
log2 (M )
ABEPsignal = N1t log

ABEPMOD (nt )
(N
M
)
t
nt =1
M
"
ABEP
= 1 log2 (Nt )
ABEP
(l)
spatial
M log2 (Nt M )
SSK
l=1
Nt "
Nt
M
M
"
"
"
ABEPjoint = Nt1M log (N
M
)
t
2
nt =1 l=1 n
t =nt =1
l=l=1

NH (
nt nt ) + NH (l l ) (nt , l , n
t , l )
(7)
M
M "
"
1
1
ABEPMOD (nt ) = M
log2 (M )
l=1l=1

NH (l l )E(nt ) Pr l = l |l
(8)
Nt "
Nt
"
1
1
ABEPSSK (l) = Nt log (Nt )
nt =1 n
t =1
[NH (
nt nt )l (nt , n
t )]
and
1) NH (
nt nt ), NH (l l ) are the Hamming distances of the bits transmitted through spatial and signal
constellation diagrams, respectively;
2) E(nt ) {} is the expectation computed over the fading
channels from the nt th transmit antenna to the Nr receive
antennas; "
r
2
3) (nt ,nt ) = N
,n nt ,nr | ;
nr =1 |n
"Nr t r
4) (nt ,l ,nt ,l) = nr =1 |n t ,nr l nt ,nr l |2 ;
/2
t ) = (1/) 0 M(nt ,n t ) (
2l /2 sin2 ())d;
5) l (nt , n
/2
t , l ) = (1/) 0 M(nt , l , n t , ) (
/2
6) (nt , l , n
sin2 ())d.
2 In [6], the scaling factor 1/ log (N M ) is not present, which yields a weak
t
2
upper-bound [31].
Proof: See Appendix A.

Let us analyze each term in (6) as follows:
1) ABEPsignal is the summation of Nt addends
ABEPMOD (). By direct inspection, we notice that
each addend is the ABEP of a conventional modulation
scheme whose points belong to the signal constellation
diagram of SM and are transmitted only through the
nt th transmit antenna. Therefore, ABEPMOD () depends
only on the Euclidean distance of the points in the signal
constellation diagram, and thus, ABEPsignal can be
regarded as the term that shows how the signal constellation diagram affects the performance of SM.
2) ABEPspatial is the summation of M addends
ABEPSSK (). From, e.g., [24, eq. (35)], we observe
that ABEPSSK () is the ABEP of an equivalent SSKMIMO scheme, where is replaced by 2l . Except for
this scaling factor, ABEPSSK () only depends on the
Euclidean distance of the points in the spatial constellation diagram, and thus, ABEPspatial can be regarded as
the term that shows how the spatial constellation diagram
affects the performance of SM.
3) ABEPjoint has a more complicated structure, and it depends on the Euclidean distance of points belonging
to signal and spatial constellation diagrams. Thus, it is
called joint because it shows how the interaction of
these two non-orthogonal diagrams affects the ABEP
of SM.
Finally, let us emphasize that 1) although Proposition 1
might seem like a simple and less compact rearrangement
of (3), in Sections IV and V, we show that (6)(8) allow
us to get very simple and often closed-form expressions for
specific modulation schemes and fading channels, and 2) unlike ABEPSSK () and ABEPjoint , which are obtained through
conventional union-bound methods, ABEPMOD () is the exact
error probability related to the signal constellation diagram. In
other words, no union bound is used to compute this term.
The exact computation of ABEPMOD () avoids the inaccuracies
of using the union-bound method for performance analysis of
conventional modulation schemes, especially for large M and
small Nr [34], [37]. For this reason, we call the framework
in (6)(8) improved union bound. The better accuracy of this
new bound is substantiated in Section VI through Monte Carlo
simulations. For the convenience of the reader, in Table I, we
report the exact expression of ABEPMOD () in (8) for PSK
and QAM modulations. The formulas in Table I are useful for
arbitrary fading channels and when Gray coding is used to map
the information bits onto the signal constellation diagram.
IV. S IMPLIFIED E XPRESSIONS OF THE

AVERAGE B IT E RROR P ROBABILITY
Proposition 1 provides a very general framework to compute the ABEP for arbitrary fading channels and modulation
schemes. By direct inspection, we notice that (6)(8) can be
computed in closed-form if the MGFs of the signal-to-noise
ratios (SNRs) (nt ), (nt ,nt ) , and (nt ,l ,nt ,l) are available
in closed-form. If so, the ABEP can be obtained through the
1129
TABLE I
ABEPMOD () OF PSK AND QAM M ODULATIONS W ITH MRC AT THE R ECEIVER AND G RAY C ODING . F OR QAM M ODULATION , W E C ONSIDER
A
G ENERIC R ECTANGULAR M ODULATION S CHEME W ITH M = IM JM . S QUARE -QAM M ODULATION I S O BTAINED BY S ETTING IM = JM = M .
"Nr
T HE MGF OF (nt ) =
|hnt ,nr |2 , M(nt ) (), I S AVAILABLE IN C LOSED -F ORM IN [34] FOR M ANY C ORRELATED FADING C HANNELS . N OTE
nr =1
T HAT FADING C ORRELATION AT THE T RANSMITTER D OES NOT A FFECT ABEPMOD (), B UT FADING C ORRELATION AT THE R ECEIVER D OES
computation of simple single integrals and summations. More

specifically, M(nt ) () is available in [34] for many correlated
fading channels, which allows us to compute ABEPMOD ()
and, eventually, ABEPsignal . On the other hand, the computation of M(nt ,n t ) () and M(nt ,l ,n t ,) () deserves further
l
attention as they are not available in the literature for arbitrary fading channels. Thus, the objective of this section is
threefold:
To our best knowledge, and according to Section I, such a

comprehensive study is not available in the literature.
1) to compute closed-form expressions of M(nt ,n t ) () and

M(nt ,l ,n t ,) () for the most common fading channel
l
models;
Let us consider the scenario with identically distributed fading at the transmitter. We study uncorrelated and correlated fading, where in the latter case, the term identically distributed
2) to provide simplified formulas of the ABEP in (7) and (8)

for specific modulation schemes and fading channels;
3) to analyze the obtained formulas to better understand SM.
A. Identically Distributed Fading at the Transmitter
1130
means that all pairs of wireless links are equicorrelated. In for(s), M(nt ,n t ) (s) =
mulas, this implies M(nt ) (s) = MMOD
(s),
and
M
(s)
=
M
MSSK
(nt ,l ,n
(l ,) (s) for nt =
t ,)
l
l
1, 2, . . . , Nt and n
t = 1, 2, . . . , Nt , which means that the
t ).
MGFs are the same for each nt or for each pair (nt , n
Accordingly, the ABEP in Proposition 1 can be simplified, as
shown in Corollary 1.
Corollary 1: For identically distributed fading, (7) in Proposition 1 simplifies to (9), shown at the bottom of the page,
where ABEPMOD is the error probability in Table I with
(s). If a constant-modulus modulation is
M(nt ) (s) = MMOD
considered, i.e., l = 0 for l = 1, 2, . . . , M , then ABEPspatial

in (9) reduces to
ABEPspatial
Nt log2 (Nt )
=
2 log2 (Nt M )

/2
2
0
1
MSSK
d .
2 sin2 ()
t because it is the same for each pair

tion with indexes
"ntt and"n
Nt
t ), and 2) N
nt nt ) + NH (l
(nt , n
nt =1
n
t =nt =1 [NH (
2
l )] = (Nt /2) log2 (Nt ) + Nt (Nt 1)NH (l l ) for any
bit mapping. Finally, some simplifications lead to (9). Equation (11) can be obtained from (9) and the following
considerations:
1) If the channel phases are uniformly distributed,
" r
2 d
then (nt ,l ,nt ,l) = N
t ,nr
nr =1 |n
l nt ,nr l | =
"Nr
2
t ,nr
nr =1 |n
l nt ,nr l | . In fact, since adding a
constant phase term to a uniformly distributed phase
still yields a uniformly distributed phase, i.e., (nt ,nr +
d
l ) = nt ,nr , then nt ,nr l = [nt ,nr exp(jnt ,nr )]

d
(10)
Likewise, if a constant-modulus modulation, i.e., l = 0 for

l = 1, 2, . . . , M , and independent and uniformly distributed
channel phases are considered, then ABEPjoint in (9) simplifies to
%
M (Nt 1) log2 (M )
ABEPjoint =
2
log2 (Nt M )
&
Nt (M 1) log2 (Nt )
+
2
log2 (Nt M )

/2
2
0
1
MSSK
d . (11)
2 sin2 ()
0
Proof: ABEPsignal in (9) follows immediately from (7) by

taking into account that for identically distributed fading, the M
addends of the summation are all the same. ABEPspatial in (9)
can be obtained by noticing the following: 1) for identically
t ) in (8) is the same for nt =
distributed fading, l (nt , n
t = 1, 2, . . . , Nt , and
be moved
1, 2, . . . , Nt and n
"itNcan
" thus,
t
t
out of the twofold summation; and 2) N
N
nt
nt =1
n
t =1 H (
2
nt ) = (Nt /2) log2 (Nt ) for any bit mapping. Finally, some
algebraic manipulations lead to (9). Equation (10) follows from
(9) with l = 0 for l = 1, 2, . . . , M . ABEPjoint in (9) can
be obtained as follows: 1) for identically distributed fading,
t , l ) in (7) can be moved out of the twofold summa(nt , l , n
[l exp(jl )] = nt ,nr l exp(j(nt ,nr + l )) = nt ,nr

l exp(jnt ,nr ) = nt ,nr l .
2) If l = 0 for l = 1, 2, . . . , M , then (nt ,l ,nt ,l) =
" r
2
2
20 N
t ,nr nt ,nr | = 0 (nt ,
nt ) , which for
nr =1 |n
identically distributed fading implies M(l ,) (s) =
l
MSSK
(20 s). Thus, the integral in (9) can be replaced by
the integral in (11).

3) For a constant-modulus modulation, the integral in (9)
can be moved out of the twofold "
summation,
"M which can
be simplified using the identity M
l=1
l
l=1 NH (
l ) = (M 2 /2) log2 (M ) for any bit mapping. Finally,
some algebraic manipulations lead to (11). This concludes the proof.

Corollary 1 leads to two important considerations about the
performance of SM: 1) ABEPsignal and (9) show that, for
identically distributed fading, the ABEP of SM is independent
of the bit mapping of the spatial constellation diagram. This
result is reasonable and agrees with intuition: If the channels
are statistically identical, on average, the Euclidean distance of
pairs of channel impulse responses is the same. In this case,
the bit mapping has no role in determining the ABEP. On
the other hand, the complex-valued points of the signal constellation diagram have different Euclidean distances, and this
bit mapping plays an important role. 2) Under some realistic
assumptions (i.e., constant-modulus modulation and uniform
channel phases), ABEPjoint in (9), which in the most general
case depends on both spatial and signal constellation diagrams,
depends only on the signal constellation diagram. Thus, since
there are no terms in Table I, (10), and (11) that depend on both
constellation diagrams, we conclude that they can be optimized
individually. In particular, the best bit mapping for the signal
constellation diagram turns out to be the conventional one based
only on the Euclidean distance.
log2 (M )
ABEPsignal = log
ABEPMOD
2 (Nt M )
'
*
(
)
/2
M
2

"
log
(N
)
Nt
ABEP
1
1
t
2
l
MSSK
d
spatial = M log (Nt M ) 2
2 sin2 ()
2
l=1
0

'
*
+
,
(
)
/2
M
M "
"
N
log
(N
)
1
1
1
t
t
2
+ NH (l l )(Nt 1)
M(l ,) 2 sin2 () d
2
ABEPjoint = M log2 (Nt M ) l=1
l
l=l=1
(9)
Finally, we notice that, e.g., (10) and (11) are very simple to
compute and avoid the computation of fold summations on Nt
and M . This is an important difference with respect to other
frameworks available in the literature, where fourfold summations are always required, regardless of modulation scheme
and channel model [6], [9]. In addition, Corollary 1 simplifies
the frameworks in [24] and [26] for SSK modulation, as the
twofold summation can be avoided for some fading channels
and modulation schemes. Furthermore, we mention that Corollary 1 provides closed-form results if the MGFs, which depend
on the specific fading channel model, are available in closedform as well as if the related finite integral can be computed explicitly. In Section IV-B and C, we show some fading scenarios
(Nakagami-m and Rice fading with arbitrary fading parameters
and correlation) where the MGFs can be obtained in close-form.
In addition, in Section V, we study the i.i.d. Rayleigh fading
scenario where fold summations can be avoided and integrals
can be computed in closed-form, thus leading to a very simple
analytical framework for system analysis and optimization.
1131
2) ABEPspatial : For Nakagami-m fading, M(nt ,n t ) () is

not well known in the literature and only recently it has been
analyzed in [24] for single-antenna receivers, i.e., Nr = 1.
Thus, we need to generalize [24] for our system model. Two
case studies are considered: 1) correlated fading at the transmitter and independent fading at the receiver and 2) correlated fading at both ends of the MIMO channel. In the first
case study, by exploiting the independence
0 r of the fading at
(s),
the receiver, we have M(nt ,n t ) (s) = N
t ;nr )
nr =1 M(nt ,n
where M(nt ,n t ;nr ) () is the MGF of (nt ,nt ;nr ) = |n t ,nr
nt ,nr |2 . This latter MGF is available in closed-form in
[24, Sec. III] for generic correlated fading at the transmitter.
The second case study is analytically more complicated as
M(nt ,n t ) () requires the computation of the expectation of
2Nr correlated RVs. Proposition 2 provides the final expression
of M(nt ,n t ) () for Nr = 2.
Proposition 2: Given 2Nr arbitrary distributed and correlated Nakagami-m RVs with fading envelopes (nt ,nr and
n t ,nr ) distributed according to the multivariate Nakagami-m
pdf in [41, eq. (2)] and channel phases uniformly and i.i.d. in
[0, 2), then M(nt ,n t ) () for Nr = 2 is given in (13), shown at
the bottom of the page, where
B. Nakagami-m Envelopes With Uniform Phases

In this section, the fading envelopes nt ,nr are Nakagami-m
(nt ,nr )
RVs with fading severity mNak
= [E{n2 t ,nr }]2/E{[n2 t ,nr
(n ,nr )
t
E{n2 t ,nr }]2 } and mean square value Nak
(n ,n )
= E{n2 t ,nr }. We
(n ,n )
t
r
t
r
adopt the notation nt ,nr N (mNak
, Nak
). The ampli(nt ,nr ,
nt ,
nr )
tude correlation coefficient Nak
is defined as follows:
(n ,nr ,
nt ,
nr )
t
Nak
E{[nt ,nr E{nt ,nr }] [n t ,nr E{n t ,nr }]}

=- .
/- .
/.
E [nt ,nr E{nt ,nr }]2
E [n t ,nr E{n t ,nr }]2
(12)
In addition, the channel phases nt ,nr are independent and

uniformly distributed RVs in [0, 2). We adopt the notation
nt ,nr U(0, 2). Finally, channel phases and fading envelopes are assumed to be independent.
Given this fading model, let us analyze and explicitly compute each term in (6).
1) ABEPsignal : M(nt ) () has been widely studied in the
literature, and closed-form expressions for non-identically distributed and arbitrary correlated Nakagami-m fading can be
found in [34, Sec. 9.6.4].
M(nt ,n t ) (s) =
1) is the 2Nr 2Nr correlation matrix of the Gaussian

RVs associated with nt ,nr and n t ,nr , which can be
computed from the amplitude correlation coefficients
(nt ,nr ,
nt ,
nr )
by using the method in [42];
Nak
2) trid is the tridiagonal approximation of , which can be
obtained as described in [41, Sec. IV] and Appendix B;
1
3) pab = 1
trid (a, b) are the entries of trid ;
(nt ,nr )
(
nt ,nr )
= mNak
is the fading parameter
4) mNak = mNak
common to all links;
(p ,p )
(p ,p )
5) Fk 11 33 () and Fk 22 44 () are defined as
(p ,p )
(m
+k +k )
Fk 11 33 (s) = (1/4)sp33 Nak 2 3

G1,2 s2 1mNak k2 k3 1mNak k1
2,2
sp11 sp33
0
0
(p22 ,p44 )
(mNak +k1 +k2 ) (mNak +k3 )
Fk
(s) = (1/4)sp22
sp44

1m
1mNak k1 k2
2

Nak k3
1,2
G2,2 sp ssp
22
44
0
0
(14)
where sp = s + (p/2).
Proof: See Appendix B. Formulas for Nr > 2 can be
obtained as described in Appendix B. For arbitrary Nr , the final
formula is given by the (2Nr 1)-fold series of the product of
(,)
Nr terms Fk ().
1 mNak

trid
2(4mNak 4) (mNak )
'
*

(p ,p )
(p ,p )
+
+ k1 +k2 +k3
|p12 |2k1 |p23 |2k2 |p34 |2k3 Fk 11 33 (s)Fk 22 44 (s)

1
4
(k1 !)(k2 !)(k3 !)(k1 + mNak )(k2 + mNak )(k3 + mNak )
k1 =0 k2 =0 k3 =0
(13)
1132
It is worth mentioning that (14) gives an exact result when

is tridiagonal, i.e., = trid . On the contrary, for arbitrary
correlation and by using the Green method [41, Sec. IV],
it provides a very tight approximation (see Section VI). Finally,
we mention that the series in (14) converge very quickly
thanks to the factorial and the Gamma functions in the
denominator.
ABEPjoint : To compute M(nt ,l ,n t ,) (), we need Prol
position 3.
Proposition 3: For Nakagami-m fading envelopes and uniform
phases, (nt ,l ,nt ,l) reduces to (nt ,l ,nt ,l) = (nt ,l ,nt ,l) =
"Nr
d (l)
(
l)
(l)
(l)
2
nr=1 |n
t ,nr nt ,nr | , where nt ,nr = nt ,nr exp(jnt ,nr ),
(l)
(l)
(n ,nr )
t
nt ,nr = l nt ,nr , and nt ,nr N (mNak
(nt ,nr ;l)
(n
,n
)
t
r
Nak
= 2l Nak
.
(l)
(n ,nr ;l)
t
, Nak
2)
3)
) with
(l)
Proof: The equality in law nt ,nr = nt ,nr exp(jnt ,nr )

can be obtained by using the same analytical development used
for (11) in Corollary 1, and more specifically, the identity in law
d

(nt ,nr + l ) = nt ,nr . This concludes the proof.
Proposition 3 points out that (nt ,nt ) and (nt ,l ,nt ,l)
are related by a scaling factor in the mean power of each
(nt ,nr ;l)
(nt ,nr )
/Nak
= 2l . Accordingly,
channel envelope, i.e., Nak
M(nt ,l ,n t , () can be computed by using the same framel
works used to compute M(nt ,n t ) (). In other words, for arbitrary correlation, M(nt ,l ,n t ,) () is still given by (13) but
l
with a different correlation matrix .
Diversity Analysis: The accurate analysis of ABEPsignal ,
ABEPspatial , and ABEPjoint through closed-form expressions
of the MGFs allows us to provide important considerations
about the diversity gain [43] of SM in Nakagami-m fading as
well as to understand the constellation diagram that dominates
the performance of SM for high SNR. The main result is given
in Proposition 4.
Proposition 4: Let us assume, for the sake of simplic(nt ,nr )
(
nt ,nr )
= mNak
for each wireless link.
ity, that mNak = mNak
The diversity gain DivSM of SM is equal to DivSM =
min{Nr , mNak Nr }.
Proof: From (6), we have DivSM = min{Divsignal ,
Divspatial , Divjoint }, where Divsignal , Divspatial , and Divjoint
are the diversity gains of ABEPsignal , ABEPspatial , and
ABEPjoint , respectively. In fact, for high SNR, the worst term
dominates the slope of the ABEP and thus the diversity gain
[43]. From Section IV-B1 and [34, Sec. 9.6.4], it follows that
Divsignal = mNak Nr . From Section IV-B2 and B3, [24], and
[44], it follows that Divspatial = Divjoint = Nr . In fact, as
(,)
analytically shown in [44], M(nt ,n t ;nr ) () and Fk () have
unit diversity gain regardless of the fading severity mNak , and
thus, the diversity is determined only by the number Nr of
antennas at the receiver. This concludes the proof.

Proposition 4 unveils important properties of SM and provides information about the best scenarios where SM should be
used. More specifically, we have the following:
1) In scenarios with less severe fading than Rayleigh, i.e.,
mNak > 1, we have DivSM = Divspatial = Divjoint =
Nr . We conclude that the ABEP is mainly determined
by the spatial constellation diagram (i.e., ABEPspatial
4)
5)
ABEPsignal and ABEPjoint ABEPsignal ), and that the

diversity gain is independent of fading severity.
In scenarios with more severe fading than Rayleigh,
i.e., 0.5 mNak < 1, we have DivSM = Divsignal =
mNak Nr . We conclude that the ABEP is mainly
determined by the signal constellation diagram
(i.e., ABEPsignal ABEPspatial and ABEPsignal
ABEPjoint ), and that the diversity gain strongly depends
on fading severity.
Due to the increasing diversity gain of ABEPsignal with
mNak [34], it is expected that ABEPsignal provides a
negligible contribution for increasing mNak , and that the
ABEP gets better with mNak . This behavior is similar
to conventional modulations [34] but different from SSK
[24], [26].
From [34], it is known that conventional single-antenna
systems have the same diversity gain as ABEPsignal ,
i.e., Divconventional = Divsignal = mNak Nr . Thus, with
respect to conventional modulations, the performance
gain of SM is expected to increase for 0.5 mNak < 1,
whereas it is expected to decrease for mNak > 1. This
conclusion agrees with intuition since SM encodes part of
the information bits onto the spatial constellation diagram
whose points are more closely spaced if mNak > 1.
From [24] and [44], it is known that the diversity
gain of SSK modulation is the same as ABEPspatial ,
i.e., DivSSK = Divspatial = Nr , which is independent of
mNak . Thus, unlike conventional modulation schemes
and SM, SSK modulation does not experience any diversity reduction when 0.5 mNak < 1, and it can be concluded that, thanks to the higher diversity gain, it turns out
to be, among SM and conventional modulations, the best
transmission scheme in scenarios with fading less severe
than Rayleigh. The price to be paid is the need of many
antennas at the transmitter to achieve the same rate. On
the contrary, in more severe fading channels, SSK modulation turns out to be worse than conventional modulation.
In conclusion, the performance of SM in Nakagami-m

fading strongly depends on mNak , and there is no clear transmission technology better than others for any mNak . This
important result suggests the adoption of a multimode adaptive
transmission scheme, which can switch among the best modulations according to the fading severity and the desired rate.
Finally, we close this section with the following corollary.
Corollary 2: For Nakagami-m fading envelopes, uniform
channel phases, and a constant-modulus modulation, i.e., l =
0 for l = 1, 2, . . . , M , ABEPspatial and ABEPjoint in (7) can
be simplified as in (15), shown at the bottom of the next page.
Proof: Equation (15) can be obtained through analytical
steps similar to (10) and (11) but without the assumption of
identically distributed fading.

Corollary 2 shows that, for a constant-modulus modulation, ABEPspatial and ABEPjoint can be computed only
(). This makes even more evthrough M(nt ,n t ) () = MSSK
ident the connection established between M(nt ,n t ) () and

M(nt ,l ,n t ,) () in Proposition 3. We note that in (15), neither
l
ABEPspatial nor ABEPjoint depend on the bit mapping used for
the signal constellation diagram. Thus, the optimality of usual

bit mappings adopted for ABEPsignal seems to be preserved.
C. Rician Fading
Let us consider a generic Rician fading [26], [45]. In this
case, nt ,nr are generically correlated complex Gaussian RVs,
and nRt ,nr and nI t ,nr are independent by definition. We adopt
R
I
I
the notation R
nt ,nr = E{nt ,nr }, nt ,nr = E{nt ,nr }, and
2
I
I
2
n2 t ,nr = E{(nRt ,nr R
nt ,nr ) } = E{(nt ,nr nt ,nr ) }.
R
R
2
In addition, we use the shorthands nt ,nr G(nt ,nr , nt ,nr )
and nI t ,nr G(Int ,nr , n2 t ,nr ).
The analysis of Rician fading is simpler than Nakagami-m
fading, and a unified framework can be used to compute (6).
The main enabling result is summarized in Proposition 5.
Proposition 5: Given a complex Gaussian RV nt ,nr ,
(l)
then nt ,nr = l nt ,nr is still a complex Gaussian RV such
(l)
(l)
2 2
that Re{nt ,nr } G(R
Im{nt ,nr }
nt ,nr ,l , l nt ,nr ),
R
G(Int ,nr ,l , 2l n2 t ,nr ), with R
nt ,nr ,l = nt ,nr l cos(l )
I
I
I
nt ,nr l sin(l ) and nt ,nr ,l = nt ,nr l cos(l ) + R
nt ,nr
(l)
(l)
l sin(l ). In addition, Re{nt ,nr } and Im{nt ,nr } are

independent RVs.
(l)
Proof: By definition, nt ,nr = l exp(jl )(nRt ,nr +
(l)
(l)
(l)
jnI t ,nr ) = Re{nt ,nr } + jIm{nt ,nr } with Re{nt ,nr } =
1133
m fading, in Rician fading, the ABEPs in (6) have the same

slope. However, [26, Sec. V] has pointed out that ABEPspatial
and ABEPsignal have opposite behavior with the Rician factor:
ABEPspatial increases and ABEPsignal decreases when the Rician factor increases, respectively. Therefore, the ABEPs in (6)
have different coding gains depending on the fading severity.
The Moschopoulos framework can be simplified for some
fading channels, as shown in Corollary 3.
Corollary 3: For Rician fading, a constant-modulus moduI
lation, and zero-mean fading, i.e., R
nt ,nr = nt ,nr = 0, then
ABEPspatial and ABEPjoint in (7) can be simplified as shown
in (15).
Proof: For zero-mean fading and a constant-modulus
d
modulation, (nt ,l ,nt ,l) = 20 (nt ,nt ) because R
nt ,nr ,l and
I
nt ,nr ,l are independent of l . Then, considerations similar to
Corollary 2 lead to (15).

We emphasize that, although Corollaries 2 and 3 provide
the same result, the assumptions are different. In Nakagamim fading, the channel phases need to be uniformly distributed.
On the other hand, in Rician fading, the complex channel gains
need to have zero mean (i.e., zero Rician factor). Furthermore, it
should be noted that although Rayleigh fading is a special case
of either Rician or Nakagami-m fading, for correlated channels,
exploiting the framework for Rician fading leads to a more
straightforward analytical derivation.
(l)
Im{nt ,nr }
l nRt ,nr cos(l ) l nI t ,nr sin(l ) and

= l
nI t ,nr cos(l ) + l nRt ,nr sin(l ). Then, by taking into
account that nRt ,nr and nI t ,nr are Gaussian distributed,
independent, and have the same variance, simple algebraic
manipulations conclude the proof.

From Proposition 5, we conclude that (nt ) , (nt ,nt ) , and
(nt ,l ,nt ,l) in Table I and (6)(8) are all given by the summation of Nr square envelopes of arbitrary distributed and
correlated Gaussian RVs. Thus, the related MGF can be computed by using the so-called Moschopoulos method [45]. More
specifically, we have the following.
1) M(nt ) () can be obtained from [45, eq. (25)].
2) M(nt ,n t ) () can be found in [26, eq. (15)].
3) M(nt ,l ,n t ,) () can still be obtained from [26, eq. (15)]
l
thanks to Proposition 5.
and
The
only
differences
between
M(nt ,n t ) ()
M(nt ,l ,n t ,) () are the parameters of each Gaussian
l
RV, which, however, can be related to one another, as shown in
Proposition 5. The same applies to the covariance matrices.
1) Diversity Analysis: Diversity can be studied by using the Moschopoulos method. In fact, [45, Sec. IV-B] and
[26, Sec. III-C] show that each ABEP term in (6) has diversity
gain equal to Nr , i.e., DivSM = Nr . Thus, unlike Nakagami-
ABEPspatial =
ABEPjoint =
1
Nt log2 (Nt M )
1
Nt log2 (Nt M )
Nt "
Nt
"
'
nt =1 n
t =1
Nt
"
Nt
"
nt =1 n
t =nt =1
/2
0
+
MSSK
M log2 (M )
2
V. AVERAGE B IT E RROR P ROBABILITY

OVER I NDEPENDENT AND I DENTICALLY
D ISTRIBUTED R AYLEIGH FADING
In this section, we study the canonical i.i.d. Rayleigh fading
scenario. Our contribution is threefold:
1) i.i.d. Rayleigh fading has already been studied in [6].
However, [6] is useful only for real-valued signal constellation points, whereas our framework is simple and
applicable to generic signal constellation diagrams. In
addition, we provide asymptotically tight bounds that
highlight fundamental properties of SM.
2) i.i.d. Rayleigh fading is a special case of either Rician or
Nakagami-m fading [34]. We show how the integrals in
(6) can be computed in closed-form.
3) Closed-form expressions of the SNR difference between
SM and other similar transmission technologies are provided. This allows us to understand the best transmission
technology to use for every MIMO setup and data rate.
To our best knowledge, these contributions make this
section novel and important to understand the achievable
performance of SM.
20
2 sin2 ()
)
d
,
+ NH (
nt nt )(M 1)
'
1
/2

0
MSSK
20
2 sin2 ()
)
d
(15)
1134
Let us consider the channel model in Section IV-C, which

I
for i.i.d. Rayleigh fading reduces to R
nt ,nr = nt ,nr = 0 and
2
2
nt ,nr = 0 . Corollary 4 summarizes the ABEP of SM over
i.i.d. Rayleigh fading.
Corollary 4: For i.i.d. Rayleigh fading, the ABEP in (6)
reduces to (16), shown at the bottom of the page, where
and R() are defined in Table I. Furthermore, for
ABEPRayleigh
MOD
a constant-modulus modulation, ABEPspatial and ABEPjoint
simplify as
,
+

Nt log2 (Nt )
2
2
ABEPspatial =+ 2 log2 (Nt M ) R 40 0

,
Nt (M 1) log2 (Nt )
M (Nt 1) log2 (M )
ABEP
=
+
joint
2
log2 (Nt M )
2
log2 (Nt M )

R 402 20 .
(17)
Proof: ABEPsignal follows from Corollary 1 by using the
known results summarized in Table I. ABEPspatial in (16) can
(s) = (1 + 402 s)Nr
be obtained from Corollary 1 with MSSK
[34, eq. (2.8)] and by computing the related integral with [34,
eq. (5A.4b)]. ABEPjoint in (16) can be computed from Corol" r
2 d
lary 1, i.e., (nt ,l ,nt ,l) = N
t ,nr
nr =1 |n
l nt ,nr l | =
"Nr
2
t ,nr
nr =1 |n
l nt ,nr l | , which for i.i.d. Rayleigh fading leads to M(l ,l) (s) = (1 + 202 (2l + 2l )s)Nr . The final integral can be computed using [34, Eq. (5A.4b)]. Finally,
"Mfollows from (16) with 2l = 0 for l = 1, 2, . . . , M ,
"M (17)
l=1
l l ) = (M /2) log2 (M ), and simple
l=1 NH (
algebraic manipulations. This concludes the proof.

The formulas in (16) and (17) provide important considerations about the performance of SM. For example, (16) shows
that, regardless of the signal constellation diagram, the bit
mapping on the spatial constellation diagram has no influence
on the performance of SM. On the other hand, the bit mapping
on the signal constellation diagram plays an important role
in ABEPjoint . In particular, while conventional bit mappings
(e.g., Gray coding) based on the Euclidean distance of the
signal constellation points turn out to be optimal to minimize
ABEPsignal , additional constraints might be introduced on the
optimal choice of the signal constellation diagram and on the
related bit mapping to minimize ABEPjoint (see Corollary 5 as
well). On the other hand, for a constant-modulus modulation,
we notice that ABEPjoint is independent of the properties

of the signal constellation diagram and only depends on its
cardinality M . Thus, the optimization criterion based on the
Euclidean distance, which is optimal for ABEPsignal , turns out
to be optimal for the overall ABEP. Finally, we remark that
ABEPspatial and ABEPjoint are independent of the phases of
the complex points of the signal constellation diagram. Only
the moduli of these points play a role. This result suggests
that, to minimize ABEPspatial and ABEPjoint , we can focus our
attention only on the moduli and neglect the phases.
To enable a deeper understanding of the achievable performance and a simpler comparison with other transmission
technologies, in Corollary 5, we provide a tight high-SNR
approximation of (16) and (17).
2) Corollary 5: For high-SNR, ABEPspatial and ABEPjoint
in (16) can be simplified as in (18), shown at the bottom
" "M
(M,N ,H)
= M
of the page, where joint r
[NH (
l=1
l l )
"M "M l=1 2
(M,Nr )
2
2 Nr
2 Nr
], joint = l=1 l=l=1 (l + l )
, and
(l + l )
"M 2Nr
(M,Nr )
spatial = l=1 l
. For a constant-modulus modula(M,Nr ,H)
= (220 )Nr (M 2 /
(M,Nr )
(M,N )
2) log2 (M ), joint = M (M 1)(220 )Nr , and spatialr =
r
. Formulas for ABEPsignal can be found in Table I.
M 2N
0
1 Nr
tion, they simplify as follows: joint
Proof:
N Equation (18) follows from R() = 2

2Nr 1
r
[40,
eq.
(14.4.18)]
and
some
algebra.

Nr
The high-SNR framework in (18) is simple and accurate and
sheds light on the performance of SM. By using [43], it enables
us to compute coding and diversity gains. In particular, the
diversity gain is Nr , whereas the coding gain depends on the
MIMO setup, i.e., Nt , M , and the spatial constellation diagram.
The impact of the signal constellation diagram comes into play
(M,N )
(M,N )
(M,N ,H)
only through spatialr , joint r , and joint r . More specifically, (18) provides the criterion to choose the points of the
signal constellation diagram, i.e., the moduli 2l that minimize
(M,N )
(M,N )
(M,N ,H)
the ABEP: spatialr , joint r , and joint r
should be kept
as small as possible for a given average energy constraint.
Thus, the Euclidean distance criterion used for ABEPsignal
along with the minimization of the coefficients above give the
cost functions that need to be jointly considered to optimize
the performance of SM. In Section VI, we show the very
log2 (M )
ABEPsignal = log
ABEPRayleigh
MOD
2 (Nt M )

"
1 log2 (Nt ) Nt
ABEPspatial = M
R 402 2l
log2 (Nt M ) 2
l=1
)),
) (
(
M +(
M "
"
Nt log2 (Nt )
1
1
2
2
2
=
+
(N
1)N
(
)
R
2
ABEP
joint
t
H
l
0
l
l
M
log
(N
M
)
2
t
l
2
(16)
l=1
l=l=1
,
+
Nr
1 Nt 1 log2 (Nt ) Nr 2Nr 1 (M,Nr )
ABEPspatial =
402
2
spatial
Nr
2 M log2 (Nt M )
+
2Nr 1 (M,Nr ,H) , 2 Nr
1 Nt 1 log2 (Nt ) 2Nr 1 (M,Nr )

Nt 1
1
ABEP
40
joint
+
joint =
joint
Nr
Nr
2 M log (Nt M )
M log (Nt M )
2
(18)
1135
TABLE II
SNR D IFFERENCE ( IN D ECIBELS ) B ETWEEN T RANSMISSION T ECHNOLOGY X AND Y . PSK AND QAM D ENOTE S INGLE -A NTENNA S CHEMES W ITH
G RAY C ODING . SM-PSK AND SM-QAM D ENOTE SM W ITH PSK AND QAM M ODULATION W ITH G RAY C ODING IN THE S IGNAL C ONSTELLATION
QAM
QAM
D IAGRAM. (M, Nt ) I S R EFERRED TO SM, W ITH M = IM JM FOR SM-QAM. M PSK AND M QAM = IM
JM
A RE R EFERRED
SSK
TO PSK AND QAM, R ESPECTIVELY . Nt
I S R EFERRED TO SSK. T HE C OMPARISON I S M ADE BY C ONSIDERING THE
S AME DATA R ATE R FOR E ACH T RANSMISSION T ECHNOLOGY, W HICH I MPLIES log2 (Nt M ) = log2 (M QAM ) =
log2 (M PSK ) = log2 (NtSSK ). OTHER S YMBOLS A RE D EFINED IN TABLE I
interesting, and apparently unexpected, result that, for some

MIMO setups and data rates, SM with PSK-modulated points
(SM-PSK) outperforms SM with QAM-modulated points
(SM-QAM) for the same average energy constraint. On the
other hand, it is well known that ABEPsignal with QAM modulation is never worse than ABEPsignal with PSK modulation.
This result can be well understood with the help of (18): unlike
PSK, QAM has points with moduli that can be either smaller or
(M,N )
(M,N )
larger than 1, which has an impact on spatialr , joint r , and
(M,N ,H)
joint r . Since the ABEP of SM is a weighted summation

of all these terms, it turns out that SM-PSK may outperform
SM-QAM. This leads to two important conclusions: 1) The best
modulation scheme (between PSK and QAM) to use depends
on M and Nt for a given data rate, and 2) neither PSK nor
QAM seem to be optimal signal modulation schemes for SM.
However, (18) provides the criterion to compute the optimal
modulation scheme that minimizes the ABEP.
A. Comparison With Single-Antenna PSK/QAM and
SSK Modulations
By exploiting Corollary 5, in this section, we aim at computing in closed-form the SNR difference between SM and
other transmission technologies with similar complexity, such
as single-antenna PSK/QAM and SSK modulations. The highSNR framework for PSK/QAM can be found in Table I,
1
whereas for SSK, we get ABEPSSK = (Nt /2)R(402 ) =
2Nr 1
Nr
2 Nr
(Nt /2) Nr (40 )
from ABEPspatial in (18).
2
Due to space constraints, we cannot report all the details of
the analytical derivation, but we can only summarize the main
procedure used to compute the formulas in Table II. From (18)
and Table I, for any transmission technology X, the ABEP is
ABEPX = KX (02 X )Nr = KX (SNRX )Nr . Then, for any
pair ABEPX and ABEPY , we have ABEPX = ABEPY
KX (SNRX )Nr = KY (SNRY )Nr . If we define the SNR dif(X/Y )

ference (in decibels) as SNR = 10 log10 (SNRX /SNRY ),
(X/Y )
then we get SNR = (10/Nr ) log10 (KY /KX ) = (10/
(X/Y )
(X/Y )
Nr ) log10 (SNR ). If SNR > 0, then for the same ABEP,
(X/Y )
Y needs SNR decibels less transmit energy than X, i.e.,
(X/Y )
SNR is the energy gain of Y with respect to X.
Using Table II, in Table III, we show some examples of the
SNR advantage/disadvantage of SM with respect to SSK and
single-antenna PSK/QAM. Further comments are postponed to
Section VI.
VI. N UMERICAL AND S IMULATION R ESULTS
The aim of this section is to substantiate frameworks and
claims through Monte Carlo simulations. Two case studies are
considered: 1) i.i.d. Rayleigh fading (see Section V), and 2)
identically distributed Nakagami-m fading (see Section IV-B).
In the first case study, we focus our attention on the better
accuracy provided by our upper bound, on the comparison of
SM with other modulations, and on understanding the role
played by the signal and spatial constellation diagrams. In the
second case study, we turn our attention to analyze the effect
of fading correlation and fading severity on the achievable
diversity. Without loss of generality, we consider the identically
distributed setup to keep the chosen parameters and variables
reasonably low to maintain a sensible set of simulation results.
This allows us to focus our attention on fundamental behaviors
and to show the main trends. In particular, in the presence
of channel correlation, we consider the constant correlation
model [41]. The reason of this choice is twofold: 1) to reduce
the number of parameters needed to identify the correlation
profile and 2) to study a worst-case scenario, which arises when
assuming that the constant correlation coefficient corresponds
to the pair of antennas that are most closely spaced.
1136
TABLE III
SNR (
IN D ECIBELS ) D IFFERENCE (S EE TABLE II FOR D EFINITION ) B ETWEEN SM-PSK/QAM AND S INGLE -A NTENNA PSK/QAM M ODULATION , AS
W ELL AS SM-QAM AND SSK M ODULATION . SM O UTPERFORMS ( I . E ., I T R EQUIRES L ESS T RANSMIT E NERGY PER S INGLE T RANSMISSION ) THE
(X/Y )
C OMPETING T RANSMISSION T ECHNOLOGY IF SNR > 0. F OR A G IVEN R ATE R IN bpcu, THE C ONSTELLATION S IZE I S 1) M (PSK,QAM) = 2R
FOR S INGLE -A NTENNA PSK/QAM M ODULATION ; 2) NtSSK = 2R FOR SSK M ODULATION ; AND 3) M Nt = 2R FOR SM, W HERE
Nt = 2, 4, 8 IN THE F IRST /S ECOND /T HIRD L INE OF E ACH ROW, R ESPECTIVELY. N.A. M EANS N OT AVAILABLE
A. Better Accuracy of the Improved Upper Bound

In Figs. 1 and 2, we study the accuracy of the improved upper
bound in Section III-A against Monte Carlo simulations and the
conventional union bound. The frameworks for single-antenna
PSK/QAM are obtained from Table I. It can be noticed that our
framework is, in general, more accurate than the conventional
union bound, and that it well overlaps with the Monte Carlo
simulations. In particular, our bound is more accurate than
the conventional union bound for large M and small Nr . In
addition, the figures compare the ABEP of SM and singleantenna PSK/QAM. In particular, the worst-case scenario with
only Nt = 2 is considered. We observe two different trends:
1) in Fig. 1, SM-PSK always outperforms PSK, regardless of
Nr , and the gain increases with Nr ; on the other hand, 2) in
Fig. 2, SM-QAM is worse than QAM if Nr = 1, and it outperforms QAM if Nr = 3. This result is substantiated by the
high-SNR framework in Table II. The general outcome of our
study for i.i.d. Rayleigh fading is the following: 1) SM-QAM
never outperforms QAM for Nr = 1, and 2) SM-QAM
never outperforms QAM for data rates R less than R =

2 bpcu. Further comments about this outcome are given in
Section VI-B.
B. Comparison With PSK, QAM, and SSK Modulations
Motivated by Fig. 2, we exploit the framework in Table II
to understand deeper the possible performance advantage of
SM with respect to SSK and single-antenna PSK/QAM. The
accuracy of the frameworks in Table II has been validated
through Monte Carlo simulations, and a perfect match has
been found. In particular, the interested reader might verify the
accuracy of Table II by looking at the SNR difference estimated
through Monte Carlo simulations in Figs. 38. Table II provides
the following outcomes.
1) If Nr = 1, SM-QAM never outperforms QAM, and the
gap increases with the data rate.
2) Whatever Nr is and if R < 3 bpcu, SM-PSK and SMQAM never outperform PSK and QAM, respectively.
1137
Fig. 1. ABEP of PSK (M PSK = 64) and SM-PSK (M = 32, Nt = 2)

against Em /N0 . Accuracy of proposed analytical framework (denoted by
improved union-bound in the legend) and conventional union bound (denoted
by union-bound in the legend) for unit-power (02 = 1) i.i.d. Rayleigh fading
(the rate is R = 6 bpcu).
Fig. 3. ABEP of SM-PSK against Em /N0 . Performance comparison for

various sizes of signal and spatial constellation diagrams. Accuracy of proposed
analytical frameworks for unit-power (02 = 1) i.i.d. Rayleigh fading (the rate
is R = 4 bpcu). The setup (M = 2, Nt = 8) is not shown as it overlaps with
the setup (M = 4, Nt = 4).
Fig. 2. ABEP of QAM (M QAM = 64) and SM-QAM (M = 32, Nt = 2)

against Em /N0 . Accuracy of proposed analytical framework (denoted by
improved union-bound in the legend) and conventional union bound (denoted
by union-bound in the legend) for unit-power (02 = 1) i.i.d. Rayleigh fading
(the rate is R = 6 bpcu).
Fig. 4. ABEP of SM-QAM against Em /N0 . Performance comparison for

3) Except the former setups, SM always outperforms PSK

and QAM, and the gain increases with R, and if more
antennas are available at the transmitter, i.e., more information bits can be sent through the spatial constellation
diagram.
4) The SNR gain increases with Nr , which means that SM
is inherently able to exploit receiver diversity much better
than PSK/QAM.
It is important to emphasize here that in Section V-A, we
have pointed out that QAM might not be the best modulation
scheme for SM. This means that the optimal signal constellation
diagram for SM is still unknown, and thus, the ABEP of SM
might be reduced further by looking for the signal constellation
diagram that optimizes the coefficients spatialr , joint r , and

(M,N ,H)
joint r . In other words, the noticeable gain offered by SM
might be increased further, and possibilities of improvement for
those setups where SM is worse than state-of-the-art might be
found as well. Further comments about the impact of the signal
modulation scheme on the performance of SM are available in
Section VI-C. This study corroborates our analytical findings
and confirms that an adaptive multimode modulation scheme
might be a very good choice. Finally, in Table III, we compare
SM-QAM with SSK as well. It can be noticed that, particularly
for high data rates, SSK outperforms SM-QAM. This result
shows that, when R increases, it is convenient to transmit the
information bits only through the spatial constellation diagram
(M,N )
(M,N )
1138


as this minimizes the ABEP over i.i.d. fading channels. However, the price to pay for this additional improvement is the need
of larger antenna arrays at the transmitter. Therefore, there is
a clear tradeoff between the achievable performance and the
number of antennas that can be put on a transmitter and still
being able to keep the i.i.d. assumption. In any case, these
numerical examples corroborate the potential performance and
energy gain benefits of exploiting SSK for low-complexity
massive MIMO implementations [20].
C. Interplay of Signal- and Spatial-Constellation Diagrams
In this section, we wish to give a deeper look at the performance of SM for various configurations of signal and spatial


constellation diagrams, as well as at the effect of the adopted

modulation scheme. More specifically, we seek to answer two
fundamental questions. 1) Is there, for a given data rate R,
an optimal pair (Nt , M ) that minimizes the ABEP, and 2) is
the optimal modulation scheme for single-antenna systems still
optimal for SM? The results shown in Figs. 38 provide a sound
answer to both questions. In particular, if R = 4 bpcu, we have
the following.
1) The ABEP decreases by increasing Nt , but the improvement is negligible for Nt > 4 (thus, Nt = 4 can be seen
as the optimal choice in this scenario).
2) The SNR gain with Nt is higher in SM-QAM than in
SM-PSK.
1139
3) For Nt = 2, SM-PSK outperforms SM-QAM, which substantiates the claims in Section V, whereas there is no
difference between them for Nt 4. In fact, in this latter
case, PSK and QAM lead to the same signal constellation
diagram. Thus, since PSK modulation is, in general,
simpler to be implemented as the power amplifiers at the
transmitter have less stringent linearity requirements [46],
then SM-PSK seems to be preferred to SM-QAM in all
cases.
If R = 5 bpcu, we have the following.
1) Nt = 8 is the best choice to minimize both the
ABEP and the size of the antenna array at the
transmitter.
2) For SM-PSK, the setup Nt = 4 is a very appealing configuration as the ABEP is close to the optimal value, but
the complexity of the transmitter is very low.
3) For Nt = 2, SM-QAM is definitely superior to SM-PSK
as the spatial constellation diagram has a low impact on
the overall performance.
4) For Nt = 4, SM-PSK is much better than SM-QAM, and
in particular, for SM-QAM, the net improvement when
moving from Nt = 2 to Nt = 4 is negligible.
5) For Nt 8, there is no difference between SM-PSK and
SM-QAM since they have the same signal constellation
diagram, and thus, SM-PSK is the best choice because it
is simpler to implement.
In addition, if R = 6 bpcu, we have a behavior similar to
R = 4 bpcu and R = 5 bpcu. Thus, we focus only on two
main aspects: 1) The best ABEP is obtained when Nt = 16.
By comparing the best MIMO setup for different rates, we
conclude that the best Nt increases with the rate, and the rule of
thumb seems to be double the number of transmit antennas for
each 1 bpcu increase of the data rate. Although this increase
of the rate might appear to be small for every doubling of the
number of antennas at the transmitter, this multiplexing gain is
obtained with a single active RF chain and with low (singlestream) decoding complexity. These two features agree with
current trends in MIMO research [20], [22], as mentioned in
Section I. 2) If Nt = 8, SM-PSK is a very appealing choice
to achieve very good performance with low complexity. In
addition, we emphasize the good accuracy of our framework
in all the analyzed scenarios.
Finally, we close this section by mentioning that the good
performance offered by SM-PSK against SM-QAM for some
MIMO setups and rates brings to our attention that SM-PSK
might be a good candidate for energy efficient applications. As
a matter of fact, in [46], it is mentioned that a non-negligible
percentage of the energy consumption at the base stations of
current cellular networks is due to the linearity requirements
of the power amplifiers, which are needed to use high-order
modulation schemes (such as QAM), and which result in
the low power efficiency of the amplifiers. Furthermore, in
[47, p. 12], it is clearly stated that this power inefficiency significantly contributes to the so-called quiescent energy, which is
independent of the amount of transmitted data and thus should
be reduced as much as possible.
Fig. 9. ABEP against Em /N0 over i.i.d. Nakagami-m fading (mNak = 1.0,
i.e., Rayleigh, Nr = 2, and rate R = 6 bpcu). Performance comparison and
accuracy of the analytical framework for SM-QAM, QAM, and SSK.
Fig. 10. ABEP against Em /N0 over i.i.d. Nakagami-m fading (mNak = 0.5
and mNak = 1.5, Nr = 2, and rate R = 6 bpcu). Performance comparison
and accuracy of the analytical framework for SM-QAM, QAM, and SSK.
D. Impact of Fading Severity

In Figs. 9 and 10, we study the impact of fading severity on
the performance of QAM, SM, and SSK modulations. Fig. 9
shows the basic scenario with i.i.d. Rayleigh fading (mNak =
1.0), where from Section IV-B4, we know that all modulations
have the same diversity. Fig. 10 highlights the effect of more
(mNak = 0.5) and less (mNak = 1.5) severe fading. The figures provide the following three important outcomes, which are
well captured by the framework in Section IV-B4.
1) Overall, the ABEP gets better for increasing values of
mNak .
2) The SNR gain of SM with respect to QAM increases
if mNak = 0.5, as a consequence of the steeper slope
of some components of the ABEP of SM. Furthermore,
we notice that SSK is the only modulation scheme with
no reduction of the diversity gain. If Nt = 32, SM has
1140
Fig. 11. ABEP of SM-QAM against Em /N0 over correlated (at the transmitter) and identically distributed Nakagami-m fading (mNak = 0.5 and
mNak = 1.5, Nr = 2, and rate R = 6 bpcu). Performance comparison and
accuracy of the analytical framework for M = 2 and Nt = 32.
Fig. 12. ABEP of SM-QAM against Em /N0 over correlated (at the transmitter) and identically distributed Nakagami-m fading (mNak = 0.5 and
mNak = 1.5, Nr = 2, and rate R = 6 bpcu). Performance comparison and
accuracy of the analytical framework for M = 32 and Nt = 2.
performance very close to SSK, but the different slope is

noticeable even for moderate SNRs.
3) If mNak = 1.5, QAM provides the best diversity gain, but
at low SNR, the high coding gain introduced by SM and
SSK is still advantageous. However, a crossing point can
be observed for high SNR, which shows that QAM should
be preferred in this case.
In conclusion, these results substantiate the diversity analysis
conducted in Section IV-B4 and show, once again, that the
characteristics of fading are of paramount importance to assess
the superiority of a modulation scheme with respect to another
one. An adaptive multimode modulation scheme might be an
appealing choice to use always the best modulation scheme for
any fading scenario.
E. Impact of Fading Correlation
Finally, in Figs. 1114, we study the impact of fading correlation at the transmitter and receiver over Nakagami-m fading.
The analytical framework is available in Section IV-B4, and in
particular, in the analyzed scenario, M(nt ) (s) = M (s) can
be found in [34, eq. (9.173)]. We use a constant correlation
model, and Nak denotes the correlation coefficient of pairs of
Nakagami-m envelopes. We consider two case studies: 1) channel correlation only at the transmitter (see Figs. 11 and 12),
and 2) channel correlation only at the receiver (see Figs. 13
and 14). The rationale of this choice is to investigate the
different effects that correlation might have at either ends of the
communication link. In fact, according to (5), correlation might
have a different impact at the transmitter and at the receiver:
correlation at the transmitter affects the distance of points in
the spatial constellation diagram, whereas correlation at the
receiver reduces the diversity gain of maximal ratio combining
(MRC) at the destination.
In Figs. 11 and 12, we study the impact of correlation at the
transmitter. It can be noticed, as expected, that performance
Fig. 13. ABEP of SM-QAM against Em /N0 over correlated (at the receiver)
and identically distributed Nakagami-m fading (mNak = 0.5 and mNak =
1.5, Nr = 2, and rate R = 6 bpcu). Performance comparison and accuracy
of the analytical framework for M = 2 and Nt = 32.
degrades with channel correlation. In addition, the impact of

correlation increases with Nt , which is a reasonable outcome
in our scenario. However, the SNR degradation with increasing
values of Nak is tolerable if Nak < 0.6, whereas for higher
values, a few decibel loss can be observed. Very interestingly,
Fig. 12 shows that channel correlation has a negligible effect
if mNak = 0.5. This result is very interesting, particularly if
compared to the same curves in Fig. 11 and with the ABEP
of QAM in Fig. 10 (QAM uses just one transmit antenna, and
thus, it is not affected by fading correlation at the transmitter).
In particular, we note the following: 1) If Nt = 2, SM is always
superior to QAM, regardless of fading correlation; and 2) if
Nt = 32, SM is much better than QAM, even for a high fading
correlation (Nak = 0.9). The net outcome is the following: For
1141
in all scenarios. Only in some figures there are negligible errors,

which are mainly due to the Green approximation described
in Section IV-B. Thus, our frameworks can be exploited for
accurate system optimization.
VII. C ONCLUSION
Fig. 14. ABEP of SM-QAM against Em /N0 over correlated (at the receiver)
and identically distributed Nakagami-m fading (mNak = 0.5 and mNak =
1.5, Nr = 2, and rate R = 6 bpcu). Performance comparison and accuracy
of the analytical framework for M = 32 and Nt = 2.
severe fading channels, correlation degrades the ABEP, but it

does not offset the SNR gain that, for independent fading, SM
has with respect to QAM. On the other hand, if mNak = 1.5,
the superiority of QAM becomes even more pronounced if
compared to the independent fading scenario. In conclusion,
fading correlation at the transmitter poses no problems to SM in
severe fading channels, whereas it should be carefully managed
in other fading scenarios, particularly if we want to keep
the performance advantage over single-antenna QAM (whose
ABEP is not affected by this correlation). For SM, solutions
to counteract fading correlation have recently been proposed
in [9] and [14]. Once again, we emphasize that, because of the
constant correlation model, Figs. 11 and 12 show the worst case
effect of fading correlation, particularly for large Nt .
In Figs. 13 and 14, we study the impact of correlation at
the receiver. Overall, the ABEP degrades for increasing Nak .
A higher robustness to fading correlation can be noticed for
mNak = 1.5. If mNak = 0.5, the diversity advantage of SM
with respect to QAM if kept in the presence of channel correlation too. For large antenna arrays at the transmitter (e.g., Nt =
32), the diversity loss in ABEPsignal has a negligible impact
even for high correlated channels. If mNak = 1.5, we observe
that the SNR degradation gets smaller for larger antenna arrays
at the transmitter. In other words, transmitting more information
bits through the spatial constellation diagram (e.g., increasing
Nt ) can mitigate the effect of channel correlation at the receiver.
However, Figs. 11 and 12 point out a clear tradeoff: increasing
Nt degrades the ABEP if we have channel correlation at the
transmitter. We believe that the exploitation of the proposed
frameworks for an end-to-end system optimization by taking
into account all these tradeoffs might be a very important
research issue: How do we find the optimal SM setup providing the best performance/complexity tradeoff as a function of
fading correlation, fading severity, etc.?
Finally, we wish to emphasize the good accuracy of our
framework for the very complicated fading scenario under
analysis. Our framework agrees with Monte Carlo simulations
In this paper, we have proposed a comprehensive framework for the analysis of SM-MIMO over generalized fading
channels. The framework is applicable to a large variety of
correlated fading models and MIMO setups. Furthermore, and
more importantly, by carefully analyzing the obtained formulas,
we have derived important information about the performance
of SM over fading channels, including the effect of fading
severity, the achievable diversity gain, along with the impact
of the signal constellation diagram. It has been shown that the
modulation scheme used in the signal constellation diagram
significantly affects the performance, and for i.i.d. Rayleigh
fading, closed-form expressions for its optimization have been
proposed. Finally, we have conducted an extensive simulation
campaign to validate the analytical derivation and have showcased important trends about the performance of SM for a large
variety of fading scenarios and MIMO setups. We believe that
our frameworks can be very useful to understand fundamental
behaviors and tradeoffs of SM as well as can be efficiently used
for system optimization.
A PPENDIX A
P ROOF OF P ROPOSITION 1
Before going into the details of the proof, let us analyze
the Hamming distance NH ((
nt , l ) (nt , l )) of messages
nt , l ) (nt , l ))
(
nt , l ) and (nt , l ). In particular, NH ((
is equal to the number of different bits between the messages.
Since a bit error might occur when 1) only the antenna index is wrongly detected; 2) only the signal-modulated point
is wrongly detected; or 3) both antenna index and signalmodulated point are wrongly detected, then we conclude that
nt , l )
the total number of bits in error is given by NH ((
nt nt ) + NH (l l ), where NH (
nt
(nt , l )) = NH (
nt ) and NH (l l ) are defined in Proposition 1. This remark
is used to compute (6)(8), and it is important to highlight the
role played by the bit mapping in each constellation diagram.
Proposition 1 can be obtained as follows:
1) ABEPsignal is obtained from (4) by grouping together
all the terms for which n
t = nt and l = l, and by
noticing the following: a) NH (
nt nt ) = 0 if n
t = nt ;
and b) (5)reduces to APEP((
nt , l ) (nt , l )) =
" r
2
E (nt ) {Q( |l l |2 N
nr =1 |nt ,nr | ) }. Then,
ABEPsignal = ABEPbound
signal in (4) reduces to (19), shown
at the bottom of the next page.
2) It can readily be noticed that ABEPbound
MOD (nt ) is the union
bound of a conventional modulation scheme [34], where
a) only the nt th transmit antenna is active, and b) we have
the same constellation diagram as the signal constellation diagram of SM. More specifically, ABEPbound
MOD (nt )
is the ABEP of a single-inputmultiple-output system
1142
with MRC. This ABEP is known in closed-form for

many modulation schemes and bit mappings without the
need of using union-bound methods. Thus, to get more
accurate estimates of the ABEP, ABEPbound
MOD () can be
replaced by ABEPMOD (), as shown in (8), which is
the exact ABEP of a single-inputmultiple-output system
with MRC.
3) Likewise, ABEPspatial is obtained from (4) by grouping
together all the terms for which n
t = nt and l = l,
and by noticing that a) NH (l l ) = 0 if l = l,
and b) (5)
reduces to APEP((
nt , l ) (nt , l )) =
"Nr
2
E(nt ,nt ) {Q( l nr =1 |n t ,nr nt ,nr |2 )}. Finally,
from [34, eq. (4.2)], we havee l (nt , n
t) =
APEP((
nt , l ) (nt , l )), where l (, ) is defined in
Section III-A.
4) ABEPjoint in (7) collects all the terms that are neither in ABEPsignal nor in ABEPspatial . More specifically, (7) can be obtained from [34, eq. (4.2)]:
t , l) = APEP((
nt , l ) (nt , l )) =
(nt , l, n
"Nr
E(nt ,nt ){Q( nr =1 |n t ,nr l nt ,nr l |2 )}, where
(, , , ) is defined in Section III-A.
A PPENDIX B
P ROOF OF P ROPOSITION 2
= E exp s
=E
Nr
|n t ,nr nt ,nr |

exp s|n t ,nr nt ,nr |2
= E
Nr
1
Nr
1
I0 (2sn t ,nr nt ,nr )
exp
nr =1
sn2 t ,nr
Nr
1

exp sn2 t ,nr
nr =1
Nr
1
(21)
nr =1
where the first equality is due to the independence of the

channel phases, and the second equality is obtained from [35,
p. 339, eqs. (366) and (367)] and [24, eq. (14)]. Accordingly,
M(nt ,n t ) () simplifies as
M(nt ,n t ) (s)
N
r
1

= E
exp sn2 t ,nr exp sn2 t ,nr
nr =1
I0 (2sn t ,nr nt ,nr )]

1
Nr

exp sn2 t ,nr exp sn2 t ,nr
nr =1
(22)
*
2
nr =1
E {exp [2sn t ,nr nt ,nr cos(n t ,nr nt ,nr )]}
nr =1
nr =1
Nr
1
Nr
1
I0 (2sn t ,nr nt ,nr )] f () d
M(nt ,n t ) (s)

'
in
J(s; n t ,nr , nt ,nr )
By definition, M(nt ,n t ) () is given by
0 r
E { N
t ,nr nt ,nr cos(n
t ,nr nt ,nr )]}
nr =1 exp[2sn
(20). It can be obtained as

exp 2sn t ,nr nt ,nr
nr =1

cos(n t ,nr nt ,nr )
(20)
where the last equality explicitly shows the conditioning

over fading envelopes and channel phases, and , are
shorthands to denote the set of all fading envelopes and channel
phases, respectively. Let us compute J(s; n t ,nr , nt ,nr ) =
bound
ABEPsignal =
1 log2 (M )
Nt log2 (Nt M )
bound
ABEPMOD (nt ) =
where f () is the multivariate Nakagami-m pdf in [41, eq. (2)].

As an example, and without loss of generality, let us consider
t = 2.
Nr = 2. For ease of notation, we set nt = 1 and n
Accordingly, (22) reduces to (23), shown on the top of the next
page. Finally, by using the infinite series representation of Iv ()
in [33, eq. (9.6.10)], and after lengthy algebraic manipulations,
M(nt ,n t ) () can be rewritten as shown in (13), where the
integrals in (24), shown at the top of the next page, have
been introduced. These latter integrals can be computed in
closed-form from [24, Sec. III-B], thus obtaining the final result
in (14). More specifically, the analytical procedure we have
used to compute (14) is as follows: 1) first, the integral on
variable 2,1 is solved in closed-form by using the identities in
[32, eq. (8.4.3.1)] and [32, eq. (8.4.22)], as well as by applying the MellinBarnes theorem in [32, eq. (2.24.1.1)] on the
obtained integral; and 2) second, the obtained single integral
on variable 1,1 is solved in closed-form by using again the
identity in [32, eq. (8.4.3.1)] and by applying the MellinBarnes
theorem in [32, eq. (2.24.1.1)].
Nt
"
ABEPbound
MOD (nt )
'
23
4 *
Nr
M
M "
"
"
|l l |2
|nt ,nr |2
NH (l l )E(nt ) Q
nt =1
1
1
M log2 (M )
l=1
l=1
nr =1
(19)
1143

2
2
2
2
M(nt ,n t ) (s) =
exp s1,1
exp s2,1
I0 (2s1,1 2,1 ) exp s1,2
exp s2,2
I0 (2s1,2 2,2 ) f ()d
%
&
mNak

p44 2

|1
trid |
2
2
2
ImNak 1 (|p12 |1,1 1,2 )
f () = 2(mNak 1) (m ) 1,1 2,2 exp 2 2,2 |p12 |(mNak 1) 1,1 exp p211 1,1
Nak

ImNak 1 (|p23 |1,2 2,1 )

|p23 |(mNak 1) 1,2 exp p222 1,2

2
ImNak 1 (|p34 |2,1 2,2 )
|p34 |(mNak 1) 2,1 exp p233 2,1
(23)
+
2

+
2mNak +2k1 1 2mNak +2k2 +2k3 1
(p11 ,p33 )
F
(s)
=
1,1
2,1
exp s + p211 1,1
0
0
2

I0 (2s1,1 2,1 )d1,1 d2,1

exp s + p233 2,1
+
+
2

(p ,p )
2mNak +2k1 +2k2 1 2mNak +2k3 1
Fk 22 44 (s) =
1,2
2,2
exp s + p222 1,2
0
0
2

I0 (2s1,2 2,2 )d1,2 d2,2

exp s + p244 2,2
The analytical development can be generalized to arbitrary

Nr by simply inserting in (22) the general pdf in [41, eq. (2)]
and solving the integrals as in (21)(24).
Finally, a few comments about the Green approximation
= trid in (23). 1) The pdf in (23) requires the correlation matrix of the Gaussian RVs associated to the fading
envelopes. This matrix can be computed from the amplitude
(nt ,nr ,
nt ,
nr )
by using the procedure in
correlation coefficient Nak
(nt ,nr )
[42, Sec. III]. 2) For arbitrary and unequal values of Nak
,
the Green method in [41], which is given under the assumption
(nt ,nr )
= 1 for nt = 1, 2, . . . , Nt and nr = 1, 2, . . . , Nr ,
that Nak
must be generalized. More specifically, the coefficients ui in
[41, eq. (9)], which are needed to compute trid , take the form
ui = (i, i)/vi , where (i, i) is the entry of located in the
ith row and in the ith column, and vi are the coefficients to
be computed by solving the nonlinear system of equations in
[41, eq. (10)].
R EFERENCES
[1] Y. Chau and S.-H. Yu, Space modulation on wireless fading channels,
in Proc. IEEE Veh. Technol. Conf., Oct. 2001, vol. 3, pp. 16681671.
[2] S. Song, Y. Yang, Q. Xiong, K. Xie, B.-J. Jeong, and B. Jiao, A channel
hopping technique I: Theoretical studies on band efficiency and capacity,
in Proc. IEEE Int. Conf. Commun., Circuits Syst., Jun. 2004, vol. 1,
pp. 229233.
[3] R. Y. Mesleh, H. Haas, C. W. Ahn, and S. Yun, Spatial modulationA
new low complexity spectral efficiency enhancing technique, in Proc.
IEEE Int. Conf. Commun. Netw. China, Oct. 2006, pp. 15.
[4] M. Di Renzo, H. Haas, and P. M. Grant, Spatial modulation for multipleantenna wireless systemsA survey, IEEE Commun. Mag., vol. 49,
no. 12, pp. 182191, Dec. 2011.
[5] R. Y. Mesleh, H. Haas, S. Sinanovic, C. W. Ahn, and S. Yun, Spatial
modulation, IEEE Trans. Veh. Technol., vol. 57, no. 4, pp. 22282241,
Jul. 2008.
[6] J. Jeganathan, A. Ghrayeb, and L. Szczecinski, Spatial modulation:
Optimal detection and performance analysis, IEEE Commun. Lett.,
vol. 12, no. 8, pp. 545547, Aug. 2008.
[7] Y. Yang and B. Jiao, Information-guided channel-hopping for high
data rate wireless communication, IEEE Commun. Lett., vol. 12, no. 4,
pp. 225227, Apr. 2008.
[8] J. Jeganathan, A. Ghrayeb, L. Szczecinski, and A. Ceron, Space shift
keying modulation for MIMO channels, IEEE Trans. Wireless Commun.,
vol. 8, no. 7, pp. 36923703, Jul. 2009.
(24)
[9] T. Handte, A. Muller, and J. Speidel, BER analysis and optimization of

generalized spatial modulation in correlated fading channels, in Proc.
IEEE Veh. Technol. Conf., Sep. 2009, pp. 15.
[10] S. U. Hwang, S. Jeon, S. Lee, and J. Seo, Soft-output ML detector for
spatial modulation OFDM systems, IEICE Electron. Exp., vol. 6, no. 19,
pp. 14261431, Oct. 2009.
[11] M. M. Ulla Faiz, S. Al-Ghadhban, and A. Zerguine, Recursive leastsquares adaptive channel estimation for spatial modulation systems, in
Proc. IEEE Malaysia Int. Conf. Commun., Dec. 2009, pp. 785788.
[12] M. Di Renzo and H. Haas, Performance comparison of different spatial
modulation schemes in correlated fading channels, in Proc. IEEE Int.
Conf. Commun., May 2010, pp. 16.
[13] N. Serafimovski, M. Di Renzo, S. Sinanovic, R. Y. Mesleh, and H. Haas,
Fractional bit encoded spatial modulation (FBE-SM), IEEE Commun.
Lett., vol. 14, no. 5, pp. 429431, May 2010.
[14] R. Y. Mesleh, M. Di Renzo, H. Haas, and P. M. Grant, Trellis coded
spatial modulation, IEEE Trans. Wireless Commun., vol. 9, no. 7,
pp. 23492361, Jul. 2010.
[15] M. Di Renzo and H. Haas, Performance analysis of spatial modulation,
in Proc. IEEE Int. Conf. Commun. Netw. China, Aug. 2010, pp. 17.
[16] S. Sugiura, S. Chen, and L. Hanzo, Coherent and differential spacetime shift keying: A dispersion matrix approach, IEEE Trans. Commun.,
vol. 58, no. 11, pp. 32193230, Nov. 2010.
[17] M. Di Renzo and H. Haas, Transmit-diversity for spatial modulation
(SM): Towards the design of high-rate spatially-modulated space-time
block codes, in Proc. IEEE Int. Conf. Commun., Jun. 2011, pp. 16.
[18] E. Basar, U. Aygolu, E. Panayirci, and H. V. Poor, Space-time block
coded spatial modulation, IEEE Trans. Commun., vol. 59, no. 3, pp. 823
832, Mar. 2011.
[19] S. Sugiura, S. Chen, and L. Hanzo, Generalized space-time shift keying
designed for flexible diversity-, multiplexing- and complexity-tradeoffs,
IEEE Trans. Wireless Commun., vol. 10, no. 4, pp. 11441153, Apr. 2011.
[20] T. L. Marzetta, Noncooperative cellular wireless with unlimited numbers
of base station antennas, IEEE Trans. Wireless Commun., vol. 9, no. 11,
pp. 35903600, Nov. 2010.
[21] G. Wright, GreenTouch initiative: Large scale antenna systems demonstration, in Proc. Spring Meeting, Seoul, Korea, 2011.
[22] A. Mohammadi and F. M. Ghannouchi, Single RF front-end MIMO
transceivers, IEEE Commun. Mag., vol. 49, no. 12, pp. 104109,
Dec. 2011.
[23] M. Di Renzo and H. Haas, Improving the performance of space shift
keying (SSK) modulation via opportunistic power allocation, IEEE Commun. Lett., vol. 14, no. 6, pp. 500502, Jun. 2010.
[24] M. Di Renzo and H. Haas, A general framework for performance analysis
of space shift keying (SSK) modulation for MISO correlated Nakagamim fading channels, IEEE Trans. Commun., vol. 58, no. 9, pp. 25902603,
Sep. 2010.
[25] M. Di Renzo and H. Haas, Space shift keying (SSK) modulation with
partial channel state information: Optimal detector and performance
analysis over fading channels, IEEE Trans. Commun., vol. 58, no. 11,
pp. 31963210, Nov. 2010.
1144
[26] M. Di Renzo and H. Haas, Space shift keying (SSK-) MIMO over correlated Rician fading channels: Performance analysis and a new method for
transmit-diversity, IEEE Trans. Commun., vol. 59, no. 1, pp. 116129,
Jan. 2011.
[27] M. Di Renzo and H. Haas, Space shift keying (SSK) modulation: On
the transmit-diversity/multiplexing trade-off, in Proc. IEEE Int. Conf.
Commun., Jun. 2011, pp. 16.
[28] M. Di Renzo and H. Haas, Bit error probability of space shift keying
MIMO over multiple-access independent fading channels, IEEE Trans.
Veh. Technol., vol. 60, no. 8, pp. 36943711, Oct. 2011.
[29] M. Di Renzo, D. De Leonardis, F. Graziosi, and H. Haas,
Space shift keying (SSK-) MIMO with practical channel estimates,
IEEE Trans. Commun., 2012, to be published. [Online]. Available:
http://arxiv.org/pdf/1201.4793v1.pdf
[30] A. Alshamali and B. Quza, Performance of spatial modulation in correlated and uncorrelated Nakagami fading channel, J. Commun., vol. 4,
no. 3, pp. 170174, Apr. 2009.
[31] M. Di Renzo, R. Y. Mesleh, H. Haas, and P. M. Grant, Upper bounds
for the analysis of trellis coded spatial modulation over correlated fading
channels, in Proc. IEEE Veh. Technol. Conf., May 2010, pp. 15.
[32] A. P. Prudnikov, Y. A. Brychkov, and O. I. Marichev, Integrals and Series:
More Special Functions. Moscow, Russia: Fizmatlit, 2003.
[33] M. Abramowitz and I. A. Stegun, Handbook of Mathematical Functions
with Formulas, Graphs, and Mathematical Tables, 9th ed. New York:
Dover, 1972.
[34] M. K. Simon and M.-S. Alouini, Digital Communication over Fading
Channels, 1st ed. Hoboken, NJ: Wiley, 2000.
[35] H. L. Van Trees, Detection, Estimation, and Modulation Theory, Part I.
Hoboken, NJ: Wiley, 2001.
[36] M. Irshid and I. Salous, Bit error probability for coherent M-ary PSK
systems, IEEE Trans. Commun., vol. 39, no. 3, pp. 349352, Mar. 1991.
[37] A. Hedayat, H. Shah, and A. Nosratinia, Analysis of space-time coding in
correlated fading channels, IEEE Trans. Wireless Commun., vol. 4, no. 6,
pp. 28822891, Nov. 2005.
[38] J. Lassing, E. G. Strom, E. Agrell, and T. Ottosson, Computation of the
exact bit-error rate of coherent M-ary PSK with Gray code bit mapping,
IEEE Trans. Commun., vol. 51, no. 11, pp. 17581760, Nov. 2003.
[39] K. Cho and D. Yoon, On the general BER expression of one- and twodimensional amplitude modulations, IEEE Trans. Commun., vol. 50,
no. 7, pp. 10741080, Jul. 2002.
[40] J. Proakis, Digital Communications, 4th ed. New York: McGraw-Hill,
2000.
[41] G. K. Karagiannidis, D. A. Zogas, and S. A. Kotsopoulos, An efficient
approach to multivariate Nakagami- m distribution using Greens matrix
approximation, IEEE Trans. Wireless Commun., vol. 2, no. 5, pp. 883
889, Sep. 2003.
[42] Q. T. Zhang, A decomposition technique for efficient generation of correlated Nakagami fading channels, IEEE J. Sel. Areas Commun., vol. 18,
no. 11, pp. 23852392, Nov. 2000.
[43] Z. Wang and G. B. Giannakis, A simple and general parameterization
quantifying performance in fading channels, IEEE Trans. Commun.,
vol. 51, no. 8, pp. 13891398, Aug. 2003.
[44] M. Di Renzo and H. Haas, Bit error probability of space modulation
over Nakagami- m fading: Asymptotic analysis, IEEE Commun. Lett.,
vol. 15, no. 10, pp. 10261028, Oct. 2011.
[45] J. Cheng and T. Berger, Capacity of a class of fading channels with channel state information (CSI) feedback, in Proc. Allerton Conf. Commun.
Control, Comput., Oct. 2001, pp. 11521160.
[46] Z. Hasan, H. Boostanimehr, and V. K. Bhargava, Green cellular networks: A survey, some research issues and challenges, IEEE Commun.
Surveys Tuts., vol. 13, no. 4, pp. 524540, Fourth Quarter, 2011.
[47] S. D. Gray, Theoretical and practical considerations for the design of
green radio networks, in Proc. IEEE Veh. Technol. Conf., Budapest,
Hungary, May 2011. [Online]. Available: http://www.ieeevtc.org/
conf-admin/vtc2011spring/5.pdf
Marco Di Renzo (SM05A07M09) was born

in LAquila, Italy, in 1978. He received the Laurea (cum laude) and Ph.D. degrees in electrical
and information engineering from the University
of LAquila, in April 2003 and January 2007,
respectively.
From August 2002 to January 2008, he was
with the Center of Excellence for Research Design methodologies of Embedded controllers, Wireless interconnet and Systems-on-chip, University of
LAquila. From February 2008 to April 2009, he was
a Research Associate with the Telecommunications Technological Center of
Catalonia, Barcelona, Spain. From May 2009 to December 2009, he was an
Engineering and Physical Sciences Research Council Research Fellow with the
Institute for Digital Communications, The University of Edinburgh, Edinburgh,
U.K. Since January 2010, he has been a Tenured Researcher (Charg de
Recherche Titulaire) with the French National Center for Scientific Research
(CNRS), as well as a research staff member of the Laboratory of Signals and
Systems (L2S): a joint research laboratory of the CNRS, the cole Suprieure
dlectricit (SUPLEC), and the University of Paris-Sud XI, Paris, France.
His main research interests are in the areas of wireless communications theory,
signal processing, and information theory.
Dr. Di Renzo was the recipient of a special mention for the outstanding five-year (19972003) academic career from the University of LAquila;
the THALES Communications fellowship for doctoral studies (20032006),
University of LAquila; and the Torres Quevedo Award for his research on
ultra-wideband systems and cooperative localization for wireless networks
(20082009), Ministry of Science and Innovation, Spain.
Harald Haas (S98A00M03) received the Ph.D.

degree from the University of Edinburgh in 2001.
He currently holds the Chair of mobile communications with the Institute for Digital Communications,
University of Edinburgh, Edinburgh, U.K. His main
research interests are in the areas of wireless system
design and analysis, as well as digital signal processing, with a particular focus on interference coordination in wireless networks, spatial modulation, and
optical wireless communication. He holds more than
15 patents. He has published more than 50 journal
papers, including a science article and more than 140 peer-reviewed conference
papers. Nine of his papers are invited papers. He has co-authored a book entitled
Next Generation Mobile Access Technologies: Implementing TDD (Cambridge
University Press). Since 2007, he has been a Regular High Level Visiting
Scientist supported by the Chinese 111 program with the Beijing University
of Posts and Telecommunications.
Prof. Haas was an invited speaker at the 2011 Technology Entertainment
Design (TED) Global Conference. He was shortlisted for the World Technology
Award for communications technology (individual) in 2011.

SM Mimo Mod

Uploaded by

Copyright:

Available Formats

SM Mimo Mod

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

SM Mimo Mod

Uploaded by

Copyright:

Available Formats

1124

IEEE TRANSACTIONS ON VEHICULAR TECHNOLOGY, VOL. 61, NO. 3, MARCH 2012

Bit Error Probability of SM-MIMO Over

AbstractIn this paper, we study the performance of spatial

Index TermsLarge-scale antenna systems, massive

PATIAL MODULATION (SM) is a digital modulation

0018-9545/$31.00 2012 IEEE

IEEE TRANSACTIONS ON VEHICULAR TECHNOLOGY, VOL. 61, NO. 3, MARCH 2012

single-antenna PSK/QAM modulation, the performance

antenna nt is denoted by s(t|(nt , l )) = Em l w(t).

antenna to the nr th receive antenna. nt ,nr = nRt ,nr +

Em n t ,nr l w(t) for n

1 We emphasize that symbols with

znr (t)sch,n (t|(nt , l )) dt

|sch,nr (t|(nt , l ))|2 dt . (2)

The outcome of (2) is the estimated message (

IEEE TRANSACTIONS ON VEHICULAR TECHNOLOGY, VOL. 61, NO. 3, MARCH 2012

where ABEPsignal , ABEPspatial , and ABEPjoint are defined as

ABEPsignal = N1t log

ABEPjoint = Nt1M log (N

ABEPSSK (l) = Nt log (Nt )

Proof: See Appendix A.

IV. S IMPLIFIED E XPRESSIONS OF THE

computation of simple single integrals and summations. More

To our best knowledge, and according to Section I, such a

1) to compute closed-form expressions of M(nt ,n t ) () and

2) to provide simplified formulas of the ABEP in (7) and (8)

A. Identically Distributed Fading at the Transmitter

IEEE TRANSACTIONS ON VEHICULAR TECHNOLOGY, VOL. 61, NO. 3, MARCH 2012

considered, i.e., l = 0 for l = 1, 2, . . . , M , then ABEPspatial

t because it is the same for each pair

l ) = nt ,nr , then nt ,nr l = [nt ,nr exp(jnt ,nr )]

Likewise, if a constant-modulus modulation, i.e., l = 0 for

Proof: ABEPsignal in (9) follows immediately from (7) by

[l exp(jl )] = nt ,nr l exp(j(nt ,nr + l )) = nt ,nr

the integral in (11).

2) ABEPspatial : For Nakagami-m fading, M(nt ,n t ) () is

B. Nakagami-m Envelopes With Uniform Phases

E{[nt ,nr E{nt ,nr }] [n t ,nr E{n t ,nr }]}

In addition, the channel phases nt ,nr are independent and

1) is the 2Nr 2Nr correlation matrix of the Gaussian

Fk 11 33 (s) = (1/4)sp33 Nak 2 3

G1,2 s2  1mNak k2 k3 1mNak k1

+ k1 +k2 +k3

|p12 |2k1 |p23 |2k2 |p34 |2k3 Fk 11 33 (s)Fk 22 44 (s)

IEEE TRANSACTIONS ON VEHICULAR TECHNOLOGY, VOL. 61, NO. 3, MARCH 2012

It is worth mentioning that (14) gives an exact result when

Proof: The equality in law nt ,nr = nt ,nr exp(jnt ,nr )

ABEPsignal and ABEPjoint ABEPsignal ), and that the

In conclusion, the performance of SM in Nakagami-m

ident the connection established between M(nt ,n t ) () and

the signal constellation diagram. Thus, the optimality of usual

l sin(l ). In addition, Re{nt ,nr } and Im{nt ,nr } are

jnI t ,nr ) = Re{nt ,nr } + jIm{nt ,nr } with Re{nt ,nr } =

m fading, in Rician fading, the ABEPs in (6) have the same

l nRt ,nr cos(l ) l nI t ,nr sin(l ) and

V. AVERAGE B IT E RROR P ROBABILITY

IEEE TRANSACTIONS ON VEHICULAR TECHNOLOGY, VOL. 61, NO. 3, MARCH 2012

Let us consider the channel model in Section IV-C, which

G1,2 s2 1mNak k2 k3 1mNak k1

N Equation (18) follows from R() = 2

1 Nt 1 log2 (Nt ) 2Nr 1 (M,Nr )