G Model

AEUE-50649; No. of Pages 6 ARTICLE IN PRESS

Int. J. Electron. Commun. (AEÜ) xxx (2010) xxx–xxx

Contents lists available at ScienceDirect

International Journal of Electronics and

Communications (AEÜ)
journal homepage: www.elsevier.de/aeue

Second-order statistics of SC macrodiversity system operating over Gamma

shadowed Nakagami-m fading channels
Dušan M. Stefanović a , Stefan R. Panić b,∗ , Petar Ć. Spalević c
a
High Technical College, Aleksandra Medvedeva 20, 18000 Niš, Serbia
b
Department of Telecommunications, Faculty of Electronic Engineering, University of Niš, Aleksandra Medvedeva 14, 18000 Niš, Serbia
c
Department of Telecommunications, Faculty of Technical Science, University of Priština, Knjaza Miloša 7, 40000 Kosovska Mitrovica, Serbia

a r t i c l e i n f o a b s t r a c t

Article history: This paper derives the second-order statistics of SC (selection combining) macrodiversity operating over
Received 11 January 2010 the Gamma shadowed Nakagami-m fading channels. Macrodiversity system of SC type consists of two
Received in revised form 28 April 2010 microdiversity systems and selection (switching) is based on their output signal power values. Each
Accepted 25 May 2010
microdiversity system is of MRC (maximal ratio combining) type with arbitrary number of branches in
the presence of correlative Nakagami-m fading. We have derived the infinite-series expressions for LCR
Keywords:
(level crossing rate) and AFD (average fading duration) at the output of this system. Numerical results
Level crossing rate
are also presented in order to show the influence of various parameters such as number of the diversity
Average fade duration
Shadowed fading channel
branches at the microcmbiners, fading severity and level of correlation between those branches on the
Microdiversity system’s statistics, and then compared to the previously published results from this area.
Macrodiversity © 2010 Elsevier GmbH. All rights reserved.

1. Introduction amount of channel state information available at the receiver. Com-

bining techniques like maximal ratio combining (MRC) and equal
Rapid growth of mobile communications as well as the emer- gain combining (EGC) and require all or some of the amount of the
gence of wireless local area network (LAN) technologies has channel state information of received signal, and separate receiver
recently increased the interest in the wireless communications. chain for each branch of the diversity system, which increases the
Wireless channels are simultaneously affected by short-term fad- complexity of system. In opposition to previous combining tech-
ing and long-term fading (shadowing) [1]. The short-term signal niques, selection combining (SC) receiver processes only one of the
variation is described by several distributions such as Hoyt, diversity branches, and is much simpler and cheaper for practical
Rayleigh, Rice, Nakagami-m, and Weibull. Nakagami-m fading realization.
describes multipath scattering with relatively large delay-time Generally, SC selects the branch with the highest signal-to-noise
spreads, with different clusters of reflected waves [2]. It provides ratio (SNR), that is the branch with the strongest signal [1–3],
good fits to collected data in indoor and outdoor mobile–radio assuming that noise power is equally distributed over branches.
environments and is used in many wireless communications appli- While short-term fading is mitigated through the use of diver-
cations. sity techniques typically at the single base station (microdiversity),
Various techniques for reducing short-term fading effect are use of such microdiversity approaches alone will not be sufficient
used in wireless communication systems [3]. An efficient method to mitigate the overall channel degradation when shadowing is
for amelioration system’s quality of service (QoS) with using mul- also concurrently present. In cellular networks, long-term fading
tiple receiver antennas in is called space diversity. Upgrading known as shadowing can put a heavy limit on system performance.
transmission reliability without increasing transmission power and Shadowing is the result of the topographical elements and other
bandwidth while increasing channel capacity is the main goal structures in the transmission path such as trees, tall buildings.
of space diversity techniques. Several principal types of combin- Now, we must simultaneously take short- and long-term fading
ing techniques can be generally performed by their dependence conditions into account since they both coexist in wireless systems
on complexity restriction put on the communication system and [4]. Macrodiversity is used to alleviate the effects of shadowing,
where multiple signals are received at widely located base sta-
tions, ensuring that different long-term fading is experienced by
∗ Corresponding author. Tel.: +381 63 470 649. these signals [5]. The simultaneous use of multiple base stations
E-mail addresses: dusan.stefanovic@itcentar.co.rs (D.M. Stefanović), and the processing of signals from these base stations will provide
stefanpnc@yahoo.com (S.R. Panić), petarspalevic@yahoo.com (P.Ć. Spalević). the framework for both macro- and microdiversity techniques to

1434-8411/$ – see front matter © 2010 Elsevier GmbH. All rights reserved.
doi:10.1016/j.aeue.2010.05.001

Please cite this article in press as: Stefanović DM, et al. Second-order statistics of SC macrodiversity system operating over Gamma
shadowed Nakagami-m fading channels. Int J Electron Commun (AEÜ) (2010), doi:10.1016/j.aeue.2010.05.001
G Model
AEUE-50649; No. of Pages 6 ARTICLE IN PRESS
2 D.M. Stefanović et al. / Int. J. Electron. Commun. (AEÜ) xxx (2010) xxx–xxx

improve the performance in shadowed fading channels [6–10]. At Treating the correlation between the branches as exponential, the
the macrolevel, SC is basically a fast response hand-off mechanism expression for the pdf of the SNR at the outputs of microdiversity
that instantaneously or, with minimal delay chooses the best base systems follows [18]:
station to serve mobile based on the signal power received [11]. z  1
 N m Mi  Nm 
The level crossing rate (LCR) and the average fading duration p i
= i i
zi Mi −1 exp − i i
z (3)
˝i  (Mi ) ri ˝i ri ˝i
(AFD) are second-order statistical quantities, which complement
the static probabilistic description of the fading signal (the first- In pervious equation,  (x) denotes the Gamma function, mi is
order statistics), and have found several applications in the well-known Nakagami-m fading severity parameter. Ni denotes
modelling and design of practical systems and designing wireless the number of identically assumed channels at each microlevel.
communication systems. Actually, these second-order statistical Number of diversity branches at the microlevel can be arbitrary.
measures are related to criterion used to assess error probability of However, since channels are considered correlated, microdiversity
packets of distinct length and to determinate parameters of equiv- system is applied on small terminals where spacing between the
alent channel, modelled by a Markov chain with defined number diversity branches is small [3]. Also there is no need to increase
of states [12]. significantly number of diversity branches, because achieved out-
Macrodiversity over shadowed fading channels was discussed put performance improvement with few diversity branches would
by several researches [4–11,13]. In Ref. [6] the first-order statis- not increase very much with the appliance of more branches. So
tics analysis of macrodiversity system operating over the shadowed the limitation for the number of diversity branches in the microdi-
Nakagami-m fading channels was carried out. In Ref. [11] analytical versity system is the trade-off between the complexity of practical
expressions for the pointwise outage probability were derived and realization and requested performance improvement. Parameter ri
averaged over the desired region in order to determine estimated is related to the exponential correlation i among the branches and
outage for that region. Also the discussions about the second- is given with:
order statistics of various diversity systems can be easily found
2i

1 − i Ni

in the literature [12,14–16]. In Ref. [17], first an approximation ri = Ni + Ni − (4)
of Nakagami-Gamma (GMSM) modified Suzuki composite fading 1 − i 1 − i
model was made, and then LCR and AFD are analyzed in absence Parameter Mi is defined as:
in any diversity. Moreover, to the best knowledge of authors, no
analytical study investigating the second-order statistics of macro- mi Ni 2
Mi = (5)
diversity system operating over Gamma shadowed fading channel ri
has been reported in the literature.
In this paper the second-order statistical measures (AFD and Since the outputs of a MRC system and their derivatives follow
LCR) operating at the output of macrodivesity system over the [14]:
correlated Nakagami-m fading channels are presented. We have

Ni

Ni
focused on MRC combining at each base station, and selection com- 2 2
. z ik ·
zi = zik zi = z ik i = 1, 2 (6)
bining, based on received signal power, between base stations. zi
k=1 k=1
Received signal powers of the over diversity output signals are
.
modelled by statistically independent Gamma distributions. Some then z i is a Gaussian random variable and with zero mean:
numerical results of the system’s LCR and the AFD are presented,
. 
in order to show the influence of various parameters such as num- . 1 zi 2
ber of the diversity branches at the microcmbiners, fading severity p(zi ) = √ . exp − . (7)
2zi 2zi 2
and level of correlation between those branches on the system’s
received statistics. Obtained results are then discussed from the
and variance given with [15]:
previously published results in this area point of view.
.2
Ni .2
z2  z ik
2. System model  zi = ik
(8)
zi 2
k=1
.
Let z be the received signal envelope, and z its derivative with
For the case of equivalently assumed channels, when stands:
respect to time, with joined probability density function (pdf) .2 .2 .2
. .
pz z(z z). The level crossing rate (LCR) at the envelope z is defined  zi1 =  zi2 = .... zik , k = 1, ..., N pervious reduces into:
as the rate at which a fading signal envelope crosses level z in a .
. . 2
positive or a negative direction and is mathematically defined by fd ˝i
z2i = z2ik = (9)
formula [14]: mi
 ∞
.  .  . where fd is a Doppler shift frequency.
NZ (z) = zp . z, z dz (1)
zz Conditioned on ˝i , the joint PDF p(zi , żi /˝i ) can be calculated
0
as:
The average fade duration (AFD) is defined as the average time  .    · zi Mi −1
 N m Mi  Nm 
i i i i
over which the signal envelope ratio remains below a specified level p zi , zi /˝i = p zi /˝i ×p(zi )= exp − zi
 (Mi ) ri ˝i ri ˝i
after crossing that level in a downward direction, and is determi-

. 
nated as [14]:
1 zi 2
×√ . exp − . ; i=1, 2 (10)
Fz (z ≤ Z) 2zi
TZ (z) = (2) 2zi 2
NZ (z)

Our macrodiversity system is of SC type and consists of two It is already quoted that our macrodiversity system is of SC
microdiversity systems with selection based on their output signal type and that selection based on the microcombiners output signal
power values. Each microdiversity system is of MRC type with arbi- power values. This selection can be written through the first-order
trary number of branches in the presence ofNakagami-m fading. statistical parameters PDF and cumulative distribution function

Please cite this article in press as: Stefanović DM, et al. Second-order statistics of SC macrodiversity system operating over Gamma
shadowed Nakagami-m fading channels. Int J Electron Commun (AEÜ) (2010), doi:10.1016/j.aeue.2010.05.001
G Model
AEUE-50649; No. of Pages 6 ARTICLE IN PRESS
D.M. Stefanović et al. / Int. J. Electron. Commun. (AEÜ) xxx (2010) xxx–xxx 3

3. Second-order statistics

After substituting (10), (11) and (14) into (1), and following the
procedure explained in Appendix, we can easily derive the infinite-
series expression for the system output LCR, in the form of:

Nz (z) z M1 −1
 N m M1  2
1 1
=
fd  (M1 ) (c1 ) (c2 ) r1 m1
 (M1 +c1 +c2 +k−1/2)/2

∞
N1 m1 z/r1 ((1/˝01 ) + (1/˝02 ))
× c k+c2
c2 (1 + c2 )k ˝01 1 ˝02
k=0

 
N1 m1 z(˝01 + ˝02 )
× K(M1 +c1 +c2 +k−1/2) 2
r1 ˝01 ˝02

z M2 −1
 N m M2  2
2 2
+
 (M2 ) (c1 ) (c2 ) r2 m2
 (M2 +c1 +c2 +k−1/2)/2

∞
N2 m2 z/r2 ((1/˝01 ) + (1/˝02 ))
Fig. 1. Normalized average LCR of our macrodiversity structure for various values × c k+c2
of correlation level and diversity order. c2 (1 + c2 )k ˝01 1 ˝02
k=0

 
N2 m2 z(˝01 + ˝02 )
(CDF) at the macrodiversity output in the form of: × K(M2 +c1 +c2 +k−1/2) 2 (15)
r2 ˝01 ˝02
 
.
∞ ˝1
 .   
p . (z, z) = d˝1 d˝2 p . z, z/˝1 P˝1 ˝2 ˝1 ˝2
ZZ z1 z1 In the similar manner from (12) we can obtain an infinite-series
0 0
  expression for the output AFD, in the form of:
∞ ˝2
 .   
+ d˝2 d˝1 p . z, z/˝2 P˝1 ˝2 ˝1 ˝2 Fz (z ≤ Z)
z2 z2 Tz (z) =
0 0
Nz (z)
(11)
 M1 ∞ ∞ 
2 N1 m1 /r1 N1 m1 k
and Pz (z) =
 (M1 ) (c1 ) (c2 )M1 r1
  k=0 l=0
∞ ˝1
     (c1 +c2 +l−k−M1 )/2
Fz (z) = d˝1 d˝2 Fz1 z/˝1 P˝1 ˝2 ˝1 ˝2 N1 m1 z/r1 ((1/˝01 ) + (1/˝02 ))
0 0 × c l+c2
  ˝01 1 ˝02 c2 (1 + c2 )l (1 + M1 )k
∞ ˝2
   
 
+ d˝2 d˝1 Fz2 z/˝2 P˝1 ˝2 ˝1 ˝2 (12)
N1 m1 z(˝01 + ˝02 )
0 0 × K(c1 +c2 +l−k−M1 ) 2
r1 ˝01 ˝02
Here F(zi /˝i ) defines the CDF of the SNR, at the outputs of macro-  M2 ∞ ∞ 
diversity systems given with: 2 N2 m2 /r2 N2 m2 k
+
  (M2 ) (c1 ) (c2 )M2 r2
  zi
  k=0 l=0
F zi /˝i = p ti /˝i dti (13)  (c1 +c2 +l−k−M2 )/2
0 N2 m2 z/r2 ((1/˝01 ) + (1/˝02 ))
× l+c2 c
˝01 ˝02 1 c1 (1 + c1 )l (1 + M2 )k
Since base stations at the macrodiversity level are widely
 
located, due to sufficient spacing between antennas, signal pow- N2 m2 z(˝01 + ˝02 )
ers at the outputs of the base stations are modelled as statistically × K(c1 +c2 +l−k−M2 ) 2 (16)
r2 ˝01 ˝02
independent. Here long-term fading is as in Ref. [13] described with
Gamma distributions, which are as above mentioned independent,
as:
      4. Numerical results
p˝1 ,˝2 ˝1 , ˝2 = p˝1 ˝1 × p˝2 ˝2
In order to show the influence of various parameters such as
1 ˝1c1 −1 −˝1 /˝01 number of the diversity branches at the microcmbiners, fading
= c1 e
 (c1 ) ˝01 severity and level of correlation between those branches on the
system’s statistics, numerical results are given and graphically
1 ˝2c2 −1 −˝2 /˝02
× c2 e (14) presented. Normalized values of LCR, by maximal Doppler shift
 (c2 ) ˝02
frequency fd are presented at Figs. 1 and 2.
We can observe from Fig. 1, that lower levels are crossed with
In previous equation c1 and c2 denote the order of Gamma distri- the higher number of diversity branches at each microcombiner,
bution, the measure of the shadowing present in the channels. ˝01 and lower level of correlation between the branches. From Fig. 2
and ˝02 are related to the average powers of the Gamma long-term is obvious that larger values of Nakagami-m fading and shadowing
fading distributions. severity parameters mi and c2 provide smaller LCR values.

Please cite this article in press as: Stefanović DM, et al. Second-order statistics of SC macrodiversity system operating over Gamma
shadowed Nakagami-m fading channels. Int J Electron Commun (AEÜ) (2010), doi:10.1016/j.aeue.2010.05.001
G Model
AEUE-50649; No. of Pages 6 ARTICLE IN PRESS
4 D.M. Stefanović et al. / Int. J. Electron. Commun. (AEÜ) xxx (2010) xxx–xxx

Fig. 4. Normalized AFD of our macrodiversity structure for various values of fading
Fig. 2. Normalized average LCR of our macrodiversity structure for various values and shadowing severity level.
of fading and shadowing severity level.

reduced to [[14], Eq.27] by setting  = 0, and by adequate trans-

Normalized AFD for various values of system’s parameters is formation of parameters, obtained results are indirectly compared
presented at Figs. 3 and 4. Similarly, with higher number of diversity to those in Ref. [14]. Like in AFD case, there is an improvement in
branches, smaller correlation level and higher number of fading and obtained LCR values by proposed macrodiversity structure. Finally,
shadowing severity, better performances of system are achieved it would be interesting to discuss obtained LCR and AFD values from
(lower values of AFD). the point of view of Ref. [21]. Since Inverse Gaussian shadowed
In Figs. 5–6 normalized AFD and LCR values of our macrodiver- Nakagami-m composite fading channel has similar or better per-
sity system are compared to AFD and LCR values of non-diversity formances than Gama shadowed (for the value m = 2, AFD and LCR
cases. Considering [[20], Eq. (8)] and [[6], Eq. (3)], with respect to values for Inverse Gaussian shadowing from Figs. 1 and 2 are simi-
(9)–(10) and (2) AFD can be efficiently obtained for the case of Gama lar or better compared to Gamma shadowed channel), it is shown
shadowed Nakagami-m composite fading channel, when neither that by applying our model not for the best case, improvement is
macro- or microdiversity is applied. It is visible from Fig. 5 that AFD achieved. This arises assumption, that this method could be suc-
are lower for the case of simultaneous fading/shadowing cancella- cessfully applied for other composite models of fading/shadowing
tion, which are proposed in our paper, and that there is high margin (not just Inverse Gaussian, but also GMSM model, because compar-
of performance improvement comparing to the non-diversity case. ing with Fig. 2 from Ref. [17] similar conclusion arises).
Comparation of LCR values is presented at Fig. 6. Based on [6, Eq. In summary, numerical results support the assertion that the
(3)], with respect to (9)–(10) and (1) AFD can be efficiently obtained gain in performances is made by using this macrodiversity struc-
for the case of single Gama shadowed Nakagami-m composite fad- ture. Since better results are obtained for the cases of higher values
ing channel. Also by considering (3), (9)–(10) and (1), LCR values can of fading/shadowing parameters mi and ci and lower values of cor-
be obtained for the case when microdiversity is applied, consider- relation i , higher number of diversity branches should be used in
ing no shadowing. Since (3) can be considered as the generalization opposite cases, so there must be the trade-off between the desired
of [[14], Eq.27] for non-zero values of correlation, because it can be performances and complexity.

Fig. 3. Normalized AFD of our macrodiversity structure for various values of corre- Fig. 5. Normalized AFD of our macrodiversity structure compared to normalized
lation level and diversity order. AFD of non-diversity cases.

Please cite this article in press as: Stefanović DM, et al. Second-order statistics of SC macrodiversity system operating over Gamma
shadowed Nakagami-m fading channels. Int J Electron Commun (AEÜ) (2010), doi:10.1016/j.aeue.2010.05.001
G Model
AEUE-50649; No. of Pages 6 ARTICLE IN PRESS
D.M. Stefanović et al. / Int. J. Electron. Commun. (AEÜ) xxx (2010) xxx–xxx 5

we can write I1 in the form of:

 ∞ M1 −1/2+c1 −1  N m 
˝1 1 1
I1 = G1 c1 exp − z
˝01
0   r1 ˝1
˝1 ˝1
× exp −  c2 , d˝1 (A3)
˝01 ˝01
 N m M1 
1 1 fd z M1 −1 
G1 =
r1  (M1 ) (c1 ) (c2 ) 2m1

After presenting the lower incomplete Gamma function and the

1 F1 Hypergeometric function in the manner of [19]:
(a, x) = a−1 xa exp(−x) 1 F1 (1; 1 + a; x)

∞
(a)k xk (A4)
1 F1 (a; b; x) =
(b)k k!
k=0

and by using (1)k = k! pervious expression can be written as:

Fig. 6. Normalized average LCR of our macrodiversity structure compared to nor- ∞ 
∞ c1 +c2 +k+M1 −1/2−1  N m 
malized average LCR of non-diversity cases. ˝1 1 1
I1 = G1 c1 k+c2 exp − z
0
˝01 ˝02 c2 (1 + c2 )k r1 ˝1
k=0
  1 1

× exp −˝1 + d˝1 (A5)
5. Conclusion ˝01 ˝02

In this paper, the second-order statistics of SC macrodiver- Applying the property of Kv (x), the modified Bessel function of
sity system operating over Gamma shadowed Nakagami-m fading the second kind [19]:
channels with arbitrary parameters were analyzed. Useful infinite-  ∞  b   b v/2   
series expressions for LCR and AFD at the output of this system xv−1 exp − − ax dx = 2 Kv 2 ab (A6)
were derived. The effects of various parameters such as number of 0
x a
diversity branches at the microcmbiners, fading severity and level
finally we get:
of correlation between those branches on the system’s statistics  M1
were presented. Comparation with previous results from this area z M1 −1 fd N1 m1
I1 =
 (M1 ) (c1 ) (c2 ) r1
is included.
   (M1 +c1 +c2 +k−1/2)/2
2

∞
N1 m1 z/r1 (1/˝01 ) + (1/˝02 )
× c k+c2
m1 c2 (1 + c2 )k ˝01 1 ˝02
k=0
Appendix A. Appendix
 
N1 m1 z(˝01 + ˝02 )
After substituting (10), (11) and (14) into (1), we obtain the × K(M1 +c1 +c2 +k−1/2) 2 (A7)
r1 ˝01 ˝02
following expression:
since I2 is the same type of integral like I1 , it is easily to show that:
   M2
 N m M1 z M1 −1 
∞
I2 =
z M2 −1 fd N2 m2
1 1 M1 −1/2
Nz (z) = fd ˝1  (M2 ) (c1 ) (c2 ) r2
r1  (M1 ) (c1 ) (c2 ) 2m1   (M2 +c1 +c2 +k−1/2)/2

0

∞
 N m  ˝ c1 −1  ˝1 2 N2 m2 z/r2 (1/˝01 ) + (1/˝02 )
1 1 1 ˝1 × k+c2 c1
× exp − z exp − d˝1 m2 c1 (1 + c1 )k ˝01 ˝02
c1
r1 ˝1 ˝01 ˝01 k=0


0
 
˝2 2
c −1
˝2
 N m M2 N2 m2 z(˝01 + ˝02 )
2 2 × K(M2 +c1 +c2 +k−1/2) 2 (A8)
× c exp − d˝2 + fd r2 ˝01 ˝02
˝02 2 ˝02 r1
  ∞ Using (18), (20) and (22) in the similar way makes (12) looks as:
z M2 −1  M2 −1/2 N M1 ∞ ∞ 
×
 (M1 ) (c1 ) (c2 ) 2m2
˝2
2 1 m1
r1
N1 m1 k
0 Pz (z) =
 N m  ˝ c2 −1   ˝2  (M1 ) (c1 ) (c2 )M1 r1
2 2 2 ˝2 k=0 l=0
× exp − z c2 exp − d˝2
r2 ˝2 ˝02 ˝02 0   (c1 +c2 +l−k−M1 )/2
c1 −1
 N1 m1 z/r1 (1/˝01 ) + (1/˝02 )
˝1 ˝1 × c l+c2
× c exp − d˝1 = I1 + I2 (A1) ˝01 1 ˝02 c2 (1 + c2 )l (1 + M1 )k
˝01 1 ˝01
 
N1 m1 z(˝01 + ˝02 )
× K(c1 +c2 +l−k−M1 ) 2
r1 ˝01 ˝02
Now, by using the well-known definition of lower incomplete
Gamma function:
 M2 ∞ ∞ 
2 N2 m2 /r2 N2 m2 k
 x +
 (M2 ) (c1 ) (c2 )M2 r2
t a−1 exp(−t) dt = (a, x) (A2) k=0 l=0
0

Please cite this article in press as: Stefanović DM, et al. Second-order statistics of SC macrodiversity system operating over Gamma
shadowed Nakagami-m fading channels. Int J Electron Commun (AEÜ) (2010), doi:10.1016/j.aeue.2010.05.001
G Model
AEUE-50649; No. of Pages 6 ARTICLE IN PRESS
6 D.M. Stefanović et al. / Int. J. Electron. Commun. (AEÜ) xxx (2010) xxx–xxx

  (c1 +c2 +l−k−M2 )/2

N2 m2 z/r2 (1/˝01 ) + (1/˝02 ) [13] Shankar PM. Outage analysis in wireless channels with multiple interferers sub-
× l+c2 c1 ject to shadowing and fading using a compound pdf model. AEU—International
˝01 ˝02 c1 (1 + c1 )l (1 + M2 )k Journal of Electronics and Communications 2007;61(4):255–61.

  [14] Iskander CD, Mathiopoulos PT. Analytical level crossing rate and average fade
N2 m2 z(˝01 + ˝02 ) duration in Nakagami fading channels. IEEE Transactions on Communications
× K(c1 +c2 +l−k−M2 ) 2 (A9) 2002;50(8):1301–9.
r2 ˝01 ˝02 [15] Dong X, Beaulieu NC. Average level crossing rate and average fade duration of
selection diversity. IEEE Communication Letters 2001;10(5):396–9.
[16] Yacoub MD, Bautista J, Guedes RL. On the higher order statistic of the Nak-
agami distribution. IEEE Transactions on Vehicular Technology 1999;48(1):
References 790–4.
[17] Ramos FA, Kontorovitch V, Lara M. On the generalized and modified Suzuki
model (GMSM): approximations and level crossing statistics. In: Proceedings
[1] Stuber GL. Mobile communication. 2nd ed USA: Kluwer; 2003.
of the second international conference on electrical and electronics engineering
[2] Lee WCY. Mobile communications engineering. 1st ed New York: Mc-Graw Hill;
(ICEEE). 2005. p. 110–3.
1992.
[18] Aalo VA. Performance of maximal-ratio diversity systems in a corre-
[3] Simon MK, Alouini MS. Digital communication over fading channels. 1st ed
lated Nakagami fading environment. IEEE Transactions on Communications
New York, USA: Wiley; 2000.
1995;43(2):2360–9.
[4] Abu-Dayya AA, Beaulieu CN. Micro- and macrodiversity MDPSK on shad-
[19] Gradshteyn I, Ryzhik I. Tables of integrals, series, and products. 1st ed New
owed frequency-selective channels. IEEE Transactions on Communications
York: Academic Press; 1980.
1995;43(8):2334–42.
[20] Vetelino FS, Young C, Andrews L. Fade statistics and aperture averaging
[5] Adinoyi A, Yanikomeroglu H, Loyka S. Hybrid macro- and generalized selection
for Gaussian beam waves in moderate-to-strong turbulence. Applied Optics
combining microdiversity in lognormal shadowed Rayleigh fading channels.
2007;46(18):3780–9.
In: Proceedings of the IEEE international conference on communications, vol.
[21] Trigui I, et al. On the level crossing rate and average fade duration of composite
1. 2004. p. 244–8.
multipath/shadowing channels. In: Proceedings of the global telecommunica-
[6] Shankar PM. Analysis of microdiversity and dual channel macrodiversity in
tions conference, IEEE GLOBECOM 2008. 2008. p. 1–5.
shadowed fading channels using a compound fading model. AEU—International
Journal of Electronics and Communications 2008;62(6):445–9. Dušan M. Stefanović was born in Niš, Serbia, in 1979. He received M.Sc. degree in
[7] Shankar PM. Performance analysis of diversity combining algorithms in shad- electrical engineering from Faculty of Electronic Engineering, Niš, Serbia, in 2004.
owed fading channels. Wireless Personal Communications 2006;37:61–72. His primary research interests are statistical communication theory, optical and
[8] Turkmani AMD. Performance evaluation of composite microscopic plus macro- satellite communications. He is currently working on his Ph.D. thesis.
scopic diversity system. IEE Proceedings, Part I: Communications, Speech and
Vision 1991;138:15–20. Stefan R. Panić was born in Pirot, Serbia, in 1983. He received M.Sc. degree in elec-
[9] Jeong WC, Chung JM. Analysis of macroscopic diversity combining of MIMO trical engineering from Faculty of Electronic Engineering, Niš, Serbia, in 2007. He
signals in mobile communications. IAEU—International Journal of Electronics joined the Department of Telecommunication, Faculty of Electronic Engineering,
and Communications 2005;59:454–62. Niš in 2008 as Research Assistant on joint project between Faculty of Electronic
[10] Al-Hussaini EK, Al-Bassiouni AM, Mourad HM, Al-Shennawy H. Compos- Engineering and Ministry of Science Republic of Serbia. He has published several
ite macroscopic and microscopic diversity of sectorized macrocellular and journal publications on the above subject.
microcellular mobile radio systems employing RAKE receiver over Nakagami
fading plus lognormal shadowing channel. Wireless Personal Communications Petar Č. Spalević received B.Sc. and M.Sc. degrees in electrical engineering from the
2002;21:309–28. Faculty of Electronic Engineering, University of Priština, Serbia, and Ph.D. degree
[11] Mukherjee S, Avidor D. Effect of Microdiversity and Correlated Macrodiversity from the Faculty of Electronic Engineering, University of Niš, Serbia. His research
on Outages in a Cellular System. IEEE Transactions on Wireless Technology interests are statistical communication theory, optical and satellite communications
2003;2(1):50–9. and optimal receiver design. He has published several journal publications on the
[12] Mitic A, Jakovljevic M. Second-Order Statistics in Weibull-Lognormal Fading above subject.
Channels. Telecommunications in Modern Satellite Cable and Broadcasting
Services 2007:26–8.

Please cite this article in press as: Stefanović DM, et al. Second-order statistics of SC macrodiversity system operating over Gamma
shadowed Nakagami-m fading channels. Int J Electron Commun (AEÜ) (2010), doi:10.1016/j.aeue.2010.05.001