TELKOM
NIKA Indonesia
n
Journal of
Electrical En
gineering
Vol.12, No.7, July 201
4, pp
. 5458 ~ 54
6
8
DOI: 10.115
9
1
/telkomni
ka.
v
12i7.577
6
5458
Re
cei
v
ed
Jan
uary 12, 201
4
;
Revi
sed Ma
rch 5, 2
014;
Acce
pted Ma
rch 2
8
, 2014
Performance Analysis of a Hybrid Mimo Technique for
High Da
ta Rate Wireless Communication System
Kehind
e Od
e
y
emi*
1
, Era
s
tus O
gunti
2
1
Departme
n
t of Electrical a
nd
Electron
ic E
ngi
neer
ing, Un
iver
sit
y
of Ibad
an,
Niger
ia
2
Departme
n
t of Electrical a
nd
Electron
ic Engi
neer
ing,
F
e
d
e
r
a
l Univ
ersit
y
of
T
e
chnol
og
y, A
k
ure, Nig
eria
*Corres
o
n
d
in
g
author, e-ma
il: keson
i
cs@
ya
h
oo.com
A
b
st
r
a
ct
T
he d
e
m
a
nd f
o
r h
i
gh
er
data
rate a
n
d
better
qu
alit
y
of serv
ice (QoS)
in
w
i
r
eless
co
mmu
n
icati
o
n
s
w
a
s grow
ing fast in the past
few
years.
Obtaini
ng
thes
e requ
ire
m
e
n
ts beco
m
es ch
all
e
ngi
ng for w
i
rel
e
ss
communic
a
tio
n
systems
du
e
to the pr
ob
le
ms of cha
n
n
e
l
mu
lti-pat
h fad
i
ng, hi
gh
er p
o
w
er and
ba
nd
w
i
dth
li
mitatio
n
s. On
e of th
e
most
pro
m
isi
n
g
sol
u
tions to
th
is
pr
obl
e
m
is
Multi
p
le I
nput M
u
lti
p
le O
u
tput (MI
MO)
system. This
p
aper
pro
pos
ed
a c
o
mbi
ned
s
patia
l
mu
ltipl
e
x
i
ng
MIMO sch
eme w
i
th
bea
mfor
mi
ng f
o
r h
i
g
h
data rate
w
i
rel
e
ss co
mmu
n
ic
ation. T
h
e
pro
pose
d
tr
ans
mi
ssion sc
he
me
combi
nes th
e
ben
efits of b
o
t
h
techniques and the syst
em
was able to transmit parallel data st
ream
s as well as providing beam
forming
gai
n. Actual
ly,
these
divers
e t
e
chn
i
qu
es, sh
are th
e
sa
me r
equ
ire
m
e
n
t of
mu
lt
ipl
e
anten
na ele
m
ents,
b
u
t
differ in the a
n
t
enna e
l
e
m
ent spaci
ng n
e
ces
s
ary for t
he different sche
m
e
s
to w
o
rk.
T
hus, smart anten
n
a
array w
a
s prop
osed as a pos
sible so
luti
on a
nd w
a
s adopt
e
d
at the both trans
mitter an
d the receiv
er. T
h
e
hybri
d
techni
q
ue prov
ides h
i
gher sp
ectral e
fficiency
an
d i
m
pr
ove b
e
tter Bit error rate (BER) of the system
than the co
nve
n
tion
al MIMO,
spatia
l multip
le
xing a
nd b
e
a
m
formi
ng tech
ni
ques u
n
d
e
r the
same si
mulati
o
n
envir
on
me
nt.
Ke
y
w
ords
: MIMO, spatial mu
ltiplex
i
n
g
, bea
mfor
mi
ng,
sma
r
t antenna, BE
R, spectral effi
ciency
Copy
right
©
2014 In
stitu
t
e o
f
Ad
van
ced
En
g
i
n
eerin
g and
Scien
ce. All
rig
h
t
s reser
ve
d
.
1. Introduc
tion
The dem
and
for high d
a
ta multimedia
service
s
i
s
fast
gro
w
ing in th
e past fe
w years
and
several tech
nique
s ap
pro
a
ch
es h
a
ve
been
studie
d
to enhan
ce
bit rate and
to improve
the
reliability of any wireless
system. The
availabl
e bandwidth and the maxi
mum radiated
power are
subj
ect to fundame
n
tal p
h
ysical con
s
t
r
aints a
s
well
as regul
atio
ns and a
r
e a
l
so limited. As a
large
in
cre
a
se in
ch
annel
cap
a
city a
nd
high tran
smi
s
sion
rate
s fo
r
wirel
e
ss
co
m
m
unication
s,
the
techn
o
logie
s
for the powe
r
saving a
n
d
efficient
freq
uen
cy usabili
ty are requi
red. In a Single
-
Input Single-Output (SISO
)
antenn
a system whe
r
e th
ere is o
n
ly on
e antenn
a at both tran
smitter
and re
ceive
r
suffers a
b
o
ttlene
ck
in
te
rms of
c
apa
cit
y
due to
the
Shanno
n-Nyq
u
ist
crite
r
ion
[1,
2]; while futu
re wi
rele
ss services
dem
an
d mu
ch hi
ghe
r data
bit-rate
tran
smissio
n
with
smalle
r
bit
error rate. In orde
r to increase the ca
p
a
city of
the SISO system
s to meet su
ch de
mand, t
h
e
band
width a
n
d
transmissio
n power have
to be incre
a
sed sig
n
ifica
n
tly.
Presently, a lot of research
develop
me
nts have
sho
w
n that u
s
ing Multiple Input Multiple
Output
(MIM
O) system
s could
im
p
r
ove
system
relia
bility and in
cr
ease
the cha
nnel cap
a
city
in
wirel
e
ss co
m
m
unication substa
ntially
without
in
cre
a
sing the
tran
smissi
on
po
we
r an
d b
and
wi
dth
[3, 4].
There are thre
e typical ap
pro
a
ch
es in the
MIMO system and the
s
e inclu
de sp
atial
multiplexing
(SM),
spatial
diversity a
nd b
eamfo
rming. Th
e
spatial dive
rsi
t
y techniq
u
e
is
pred
omin
antly aimed
at improving sy
stem reli
ability because it
is used to combat channel
fading, SM tech
niqu
e is
cap
able of i
n
crea
sing
da
ta transmissi
on rate
whil
e beamfo
rmi
n
g
provide
s
a
si
gnifica
nt increase in p
e
rfo
r
man
c
e
of wi
rele
ss comm
unication
syst
ems
by focu
sing
on the si
gnal
energy in a
particular
direction to
in
crease the received SNR a
nd also re
du
ce
interferen
ce [5]. Consi
d
e
r
i
ng the adva
n
tage
s of
these
variou
s MIMO tech
nique
s, there is a n
e
ed
to integrate th
em so that th
e whol
e wirel
e
ss
syste
m
can ben
efit from these te
ch
nique
s.
In
this pap
e
r
,
a hybrid
MIMO
te
chni
que
i
s
co
nceived
a
s
a promi
s
in
g so
lution
fo
r
spe
c
trally effi
cient tra
n
smi
ssi
on techniq
ue fo
r wi
rele
ss
comm
uni
cation syste
m
. These diverse
techni
que
s,
share
the
sam
e
re
qui
reme
n
t
of multip
le
a
n
tenna
elem
e
n
ts, but
differ in the
anten
na
element spa
c
ing ne
ce
ssary for the different sche
me
s to work. Th
a
t
is, under th
e beamfo
rmi
n
g
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
Perform
a
n
c
e
Analysis of a
Hybrid Mim
o
Tech
niqu
e for High Data Rate… (Kehi
nd
e Ode
y
em
i)
5459
techni
que, th
e antenna
sp
acin
g must b
e
small in
order to provid
e the requi
re
d high ch
ann
el
correl
ation
bu
t spatial
multi
p
lexing
and
spatial
dive
rsit
y techni
que
s
requi
re
the
a
n
tenna
spa
c
i
ng
to be large e
noug
h that the correlatio
n betwe
en t
he MIMO cha
n
n
e
ls is lo
w [6]. Thus, the use of
sma
r
t antenn
a arrays at transmitte
r and
/or re
ceive
r
termin
als p
r
ov
ides a p
o
ssibl
e
solutio
n
in this
work for the
antenn
a sp
acing pro
b
lem
so that
the system would
ha
ve high-co
rrel
a
tion and lo
w-
correl
ation scenari
o
s
simul
t
aneou
sly ne
ce
ssary for th
ese different tech
niqu
es.
Variou
s hyb
r
i
d
MIMO te
ch
nique
s
sche
me hav
e
bee
n propo
se
d i
n
the p
a
st to
improve
the pe
rform
a
nce
of wi
rel
e
ss commu
n
i
cation
sy
ste
m
s. Mo
st of
them focus on
combi
n
i
n
g
beamfo
rming
with
diversit
y techni
que
s [7-1
0]. Ho
wever, thi
s
co
mbined
te
ch
nique
can
o
n
ly
enabl
e a sy
stem to achie
v
e both diversity gain an
d
beamformin
g gain. Thi
s
can im
prove
the
system pe
rfo
r
man
c
e, with
out
im
proving
the sy
stem
spe
c
tru
m
e
ffi
cien
cy si
nce
both te
chniq
u
e
s
are mai
n
ly to comb
at fadin
g
. Based on t
h
is limitat
ion, a system of h
y
bridizi
ng be
amformin
g wi
th
spatial multip
lexing techni
que is p
r
op
o
s
ed.
Thi
s
propo
sed te
ch
nique imp
r
ov
ed the syste
m
spe
c
tral effici
ency si
gnifica
ntly, as well a
s
gua
rante
e
in
g the system
BER perfo
rm
ance.
2. Sy
stem Model
A
sma
r
t
a
n
te
nna
MIM
O
system wa
s
p
r
opo
se
d
fo
r high data rat
e
wi
rele
ss system as
illustrated in
Figure 1
and
was
configured i
n
such
a way th
at both the transmitter and the
receiver were
equi
ppe
d
with on
e o
r
m
o
re smart
anten
na a
r
rays. T
h
e tra
n
smitte
r
has M
T
an
te
nn
a
arrays with e
a
ch a
r
ray ha
ving N anten
na eleme
n
ts
and there are M
R
antenn
a arrays at the
receiver
with
each a
rray
having K antenna ele
m
en
ts. We assu
med that the chan
nel sta
t
e
informatio
n
(CSI) i
s
only
kno
w
n
to the
re
ceive
r
a
n
d
that the
cha
nnel
ha
s the
Rayleig
h
fadi
ng
distrib
u
tion; a
nd sp
atially unco
r
related complex
Gau
s
sian noi
se i
s
adde
d to the faded sig
nal
at
the receiver.
The sp
aci
ng
betwe
en the antenn
a arra
ys wa
s made
to be more than 10
λ
whil
e the
antenn
a elem
ent sp
aci
ng o
f
each
antenn
a array is a
h
a
lf wavelen
g
th (
λ
/2
). The v
e
ctors
W an
d
Z
are called the
transmit be
a
m
forming a
n
d
receive b
e
a
m
forming ve
ctors, re
sp
ecti
vely
Figure 1. Pro
posed MIMO
Wirel
e
ss Syst
em with Smart Antenna Array
3. Sy
stem Analy
s
is
Conve
n
tionall
y
, the MIMOchann
el impul
se
re
spo
n
se of
MIMO
sy
stems with
T
M
trans
m
it
antenn
as a
n
d
R
M
receive anten
nas i
s
given a
s
[11]:
11
12
1
2
21
2
2
12
(1
)
T
T
RT
RR
RT
M
M
MM
MM
MM
hh
h
h
hh
H
hh
h
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 12, No. 7, July 201
4: 5458 – 54
68
5460
Whe
r
e
,
()
ij
h
sho
w
s the cha
n
n
e
l impulse re
spo
n
se between the
th
j
trans
m
itter to the
th
i
receiver el
em
ent and is giv
en as:
,
1
(2
)
()
(
)
L
ij
n
n
n
h
Whe
r
e
,
()
ij
h
multi
path
chan
nel
impul
se
re
spon
se, L i
s
t
he nu
mbe
r
of
path
s
,
n
show
s the
amplitude
of
the
th
n
path an
d it obey
s in
depe
ndent
a
nd ide
n
tical
Rayleig
h
di
stribution
(i.i.d),
.
rep
r
e
s
ent
s the impulse fun
c
tion an
d
n
rep
r
esents the d
e
lay of the
th
n
arriving path.
Applying thi
s
to multipl
e
antenn
a a
r
ra
ys case
a
s
i
n
figure
1, th
is m
a
kes the
ch
ann
el
matrix to be
RT
KM
NM
a matrix.
12
11
1
12
22
2
12
(3)
T
T
T
RR
R
RT
M
M
M
MM
M
KM
NM
hh
h
hh
h
H
hh
h
Whe
r
e
1
1
h
is ch
annel fadin
g
vector from
th
j
the anten
na a
rray at the tra
n
smitter to
th
i
antenna
array at the receive
r
.
,
1
,1
,1
,1
,
2
,
,2
,2
,2
,1
,
2
,
,,
,
,1
,
2
,
(4)
jj
j
ii
i
K
jj
j
j
ii
iK
i
jN
jN
jN
ii
i
K
KN
hh
h
hh
h
h
hh
h
At the trans
m
itter, the trans
m
it
s
i
gnal split into
M
T
p
a
rallel
sign
als
12
()
,
(
)
(
)
T
M
Sn
S
n
S
n
through the
splitter (dem
ultiplexer
) an
d is sent to the different
antenn
a arra
y to perform
beamfo
rming,
thus the tran
smit sign
al be
come
s:
()
(
)
(
5
)
j
j
Sn
W
S
n
Whe
r
e
j
W
is the transmit be
amformin
g weight vector a
nd is given a
s
:
*
()
(
6
)
jT
j
Wa
2
/
,
.
.
....
....
.....
..
2
(
1
)
/
()
1
,
,
(
7
)
tj
tj
T
j
d
Si
n
j
N
d
Si
n
Tj
ae
e
Whe
r
e
j
is the angle of dep
arture (AO
D
),
t
d
is the distan
ce between t
he antenn
a e
l
ement a
t
th
j
transmitter a
n
tenna a
rray,
is the carri
e
r
wavele
ngth,
N is the nu
mber of elem
ent at the
th
j
transmitter
antenn
a arra
y and
()
Tj
a
is t
he tran
smit
array stee
rin
g
re
spo
n
se. After
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
Perform
a
n
c
e
Analysis of a
Hybrid Mim
o
Tech
niqu
e for High Data Rate… (Kehi
nd
e Ode
y
em
i)
5461
beamfo
rming,
()
sn
be
come
s
1
N
colum
n
v
e
ct
o
r
()
j
sn
. At the receiver
side, th
e re
ceive
sign
al at
th
i
arra
y element is denote
d
as v
e
ctor
X
n
and is
given as:
()
()
(
8
)
jj
Xn
H
W
S
n
The re
ceive beamfo
rming
is
the
n
wei
ghted on
X
n
a
nd the
outp
u
t sig
nal aft
e
r
beamfo
rming at
the
th
i
receiv
e element ant
enna a
r
ray is given as:
1
(9
)
(
)
()
()
T
M
H
ii
j
i
j
rt
Z
X
n
g
n
1
(
10)
(
)
()
()
T
M
HH
j
ii
i
i
j
rt
Z
H
S
n
Z
g
n
Whe
r
e
i
Z
is the received be
amformin
g weight vector a
nd is given a
s
:
[]
(1
1
)
iR
i
Za
2
/
,
.
.
.
.
.
.
.
.
...
..
..
..
2
(
1
)
/
(1
2
)
()
1
,
,
rj
rj
jd
S
i
n
j
K
d
S
i
n
i
R
ae
e
Whe
r
e
i
is the
AOA (Angle
of Arrival),
r
d
is the distan
ce
betwe
en
the
antenn
a elem
ent at
th
i
transmitter array,
is the ca
rrie
r
wavel
e
n
g
th, K is the
numbe
r of el
ement at the
th
i
receiv
e
r
antenn
a arra
y and
()
Ri
a
is the receive a
r
ray steeri
ng resp
onse.
1
()
()
()
(
1
3
)
T
M
H
j
ii
i
j
rn
Z
H
S
n
n
Whe
r
e
()
i
n
sp
atia
lly
uncorrelat
ed compl
e
x Gau
ssi
an noi
se
with entry is
di
strib
u
ted as ~CN
(0,
o
N
) and is give
n
as:
()
()
(
1
4
)
H
ii
i
nZ
g
n
Since,
()
()
j
jj
Sn
W
S
n
we then s
u
bs
titute for
()
j
Sn
in the Equation (13
)
.
1
11
(
)
(
)
(
)
..
..
..
...
(
)
(
)
(
)
(
1
5
)
T
TT
M
HH
ii
i
i
i
i
MM
r
t
Z
h
nW
S
n
Z
h
nW
S
n
n
In matrix form:
12
11
1
1
1
2
11
1
11
12
2
22
22
22
2
2
2
2
12
12
..
.....
()
()
..
......
.....
.......
...
()
....
....
T
T
T
T
T
T
T
RR
RR
RR
T
M
HH
H
M
M
H
HH
M
M
HH
H
M
M
MM
M
M
MM
M
Zh
W
Z
h
W
Zh
W
rS
n
rS
n
Zh
W
Zh
W
Z
h
W
rS
n
Zh
W
Z
h
W
Z
h
W
(1
6
)
T
M
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 12, No. 7, July 201
4: 5458 – 54
68
5462
(1
7
)
rH
S
Whe
r
e
H
is effective ch
ann
el matrix and
is define
d
as:
12
11
1
1
1
2
11
12
22
22
2
2
2
2
12
12
..
..
..
.
..
..
...
.
(1
8
)
..
..
..
...
..
..
..
.
.
.
...
..
T
T
T
T
T
RR
RR
RR
T
M
HH
H
M
M
H
HH
M
M
HH
H
MM
MM
MM
M
Zh
W
Z
h
W
Z
h
W
Zh
W
Zh
W
Z
h
W
H
Zh
W
Z
h
W
Z
h
W
This
sho
w
s that the ch
an
nel matrix co
nsi
s
ts of MI
MO ch
ann
el fading an
d in
formation
c
o
nc
er
n
i
ng
AO
D
a
n
d
AO
A. As
a re
su
lt,
H
is then
tran
sformed
from
a
RT
KM
NM
chan
nel
matrix to a
RT
M
M
chann
el matrix
H. Due to th
e stro
ng spat
ial correl
ation
existing in e
a
c
h
antenn
a a
rra
y, accordi
ng
to the fadin
g
of the fi
rst e
l
ement fo
r ea
ch
antenn
a a
rray, the
enti
r
e
steeri
ng resp
onse of the antenna a
r
ray is [12]:
1
,
0
(,
)
[
]
(
)
[
]
(
)
(
1
9
)
L
jT
iR
i
i
j
T
j
n
i
ht
a
t
a
Whe
r
e the
ch
annel fa
ding
vector
j
i
h
is a m
a
trix of
KN
acco
rdingto E
quat
ion (4
),
,
ij
t
is
the multipath
fading
comp
onent
s
cou
p
li
ng the
first e
l
ement of
th
e
th
j
antenn
a
array at the
transmitter to the
th
i
antenn
a array at the receiver an
d
it obeys indep
ende
nt and identi
c
all
y
Rayleig
h
-di
s
tribution (i.i.d).
Since the cha
nnel is a
s
sum
ed to be flat,
Equation (17) becom
es:
,
()
[
]
()
[
]
(
2
0
)
jT
iR
i
i
j
T
j
ht
a
t
a
Then, the effective ch
ann
el fading ele
m
ent
,
ij
H
can b
e
roughly obtai
n
ed as:
,
*
,
(2
1
)
[]
[]
[
]
[
]
ij
HT
ii
i
j
j
j
RR
T
T
H
aa
a
a
Since
Ri
aK
and
Tj
aN
Whe
r
e
.
is th
e Eucli
dea
n
Vector Norm,
thus
the
effective chan
nel
fading
elem
ent
,
ij
H
ca
n
be app
roxima
tely obtained
as:
,
,
.(
2
2
)
ij
ij
HK
N
Therefore, th
e corre
s
p
ondi
ng entire
cha
nnel matrix can be form
ed
as:
1,
1
1
,
2
1,
1,
1
1
,
1
2
,
,
,1
,
2
(23
)
.
....
..
.
....
..
.
.
....
..
..
...
.
.
....
..
T
T
RT
RR
M
M
MM
MM
HK
N
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
Perform
a
n
c
e
Analysis of a
Hybrid Mim
o
Tech
niqu
e for High Data Rate… (Kehi
nd
e Ode
y
em
i)
5463
Since the
ele
m
ent of
,
ij
and
,
ij
h
has the
sa
me dist
ributio
n (i.i
.d), then
the effective cha
nnel
matrix in Equation (16
)
be
comes:
.
(
24)
HK
N
H
To dete
c
t the
tran
smit si
gn
al
Sn
, Zero forci
ng (ZF)
and
Minimum M
e
an Squ
a
re
Error
(MMSE) dete
c
tion alg
o
rith
m were con
s
idere
d
and t
he re
ceive
r
wa
s de
sign
usin
g the lin
ear
matrix
G
accordin
g to ce
rtain algo
rithm.
Thus, the re
ceive sign
al is:
()
(2
5
)
rn
H
S
The dete
c
ted
sign
al is:
(26)
SG
r
(2
7
)
SG
H
S
G
For ZF dete
c
t
i
on algo
rithm:
1
()
(
2
8
)
HH
ZF
GH
H
H
For MMSE d
e
tection al
gorithm:
1
[
]
(29)
R
HH
KM
MMS
E
o
I
GH
H
H
If the system use
s
ZF o
r
M
M
SE detectio
n
algo
rithm, the effective d
e
tection S
N
R of the
th
q
data stre
ams with linear Z
F
or MMSE equali
z
er
at th
e receiver i
s
expre
s
sed a
s
[13, 14]:
1
,
(
30)
;
1
,
2
,
...
..
...
..
()
ZF
Bf
o
qT
H
qq
qM
HH
1
,
(3
1
)
1
;
1
,
2
,
..
....
....
[]
R
MM
S
E
B
f
o
q
T
H
KM
qq
o
qM
I
HH
Whe
r
e
o
is the average S
N
R at each re
ceiver
anten
na
array an
d is
obtaine
d as:
(
32)
q
o
o
P
KN
Whe
r
e
q
P
is the transmit po
wer at ea
ch
th
j
tr
ansmit ante
n
na array.
If the transmit powe
r
is eq
u
a
lly allocated
across the tra
n
smit anten
n
a
array,
(3
3
)
o
q
T
P
P
MN
Then,
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 12, No. 7, July 201
4: 5458 – 54
68
5464
(3
4
)
o
o
To
P
M
KNN
Whe
r
e
o
P
is the total transmit
t
ed power.
By s
ubs
tituting for
o
in the Equation (29
)
and (3
0),
1
,
(3
5
)
;
1
,
2
,
.
..
.....
..
ZF
Bf
o
q
T
H
o
T
qq
P
qM
HH
K
N
N
M
1
,
(3
6)
1
;
1
,
2
,
.
...
..
..
..
[]
R
MM
S
E
B
F
o
q
T
H
KM
oq
q
T
o
P
qM
I
M
KNN
H
H
Acco
rdi
ng to Equation (24), E
quation (3
5) and
(36
)
b
e
com
e
:
1
,
;
1
,
2
,
.
.
....
...
.
(
3
7
)
()
ZF
B
f
o
q
T
H
qq
o
T
PN
K
qM
HH
N
M
22
1
,
1
;
1
,
2
,
...
....
...
(38
)
[(
)
]
R
MM
SE
BF
o
q
T
To
K
M
H
To
q
q
o
P
qM
MK
N
N
I
M
KNN
H
H
K
N
P
Thus, the
system capa
city for wirele
ss
system is give
n by [14, 15]:
2
1
lo
g
(
1
)
(
3
9
)
T
NM
q
q
C
The
system
cap
a
city for t
he hyb
r
id
scheme
whe
n
ZF and
MM
SE were ad
opted a
s
receiver i
s
ob
tained a
s
:
2
1
1
,
lo
g
(
1
)
(
4
0
)
()
T
NM
o
ZF
B
f
H
q
qq
o
T
PN
K
C
HH
N
M
2
1
22
1
,
lo
g
(
4
1
)
[(
)
]
t
R
NM
o
MM
SE
B
f
q
To
K
M
H
To
q
q
o
P
C
M
KNN
I
MK
N
N
H
H
K
N
P
4. Simulation Resul
t
s
This pap
er provides
the si
mulation re
su
lts
of the
pro
posed hyb
r
id
MIMO te
chn
i
que fo
r
high data rat
e
wi
rele
ss co
mmuni
cation
system.
T
h
e performanc
e
metrics
in
terms
of
spec
tra
l
efficien
cy an
d BER of Conventional MI
MO, S
patial Multiplexing
scheme a
nd
the Beamforming
scheme
a
r
e
given to
com
pare
with the
propo
se
d
h
y
brid te
ch
niq
ue. Th
e tran
smitter
and
t
h
e
receiver a
r
e
assume
d to have 2 smart a
n
tenna
a
rray
s
at both ends and we exa
m
ine N and K
to
be eq
ual to 2
,
4 and 8
e
le
ments in
ea
ch array. The
spa
c
in
g bet
ween a
n
tenn
a
arrays i
s
larg
er
than 10
λ
, whi
l
e the spa
c
in
g betwe
en an
tenna elem
en
ts is
λ
/ 2. The angle sprea
d
at each of the
transmitter a
n
tenna a
r
ray is 30 de
gre
e
s
and 7
0
deg
re
e
s
at the re
ceiver
side. T
he ch
ann
el h
a
s
the Rayleig
h
fading di
strib
u
tion, and
sp
atially
uncorrelated compl
e
x
Gau
ssi
an noise
is adde
d
to
the fade
d
sig
nal at th
e
re
ceiver. 1
6
-QA
M
mod
u
lation
s a
r
e
u
s
ed
to
mod
u
late th
e sym
bol
s at
the
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
Perform
a
n
c
e
Analysis of a
Hybrid Mim
o
Tech
niqu
e for High Data Rate… (Kehi
nd
e Ode
y
em
i)
5465
transmitter a
nd ZF and M
M
SE detectio
n
are ado
pte
d
at the rece
iver for the entire schem
e
s
.
These dete
c
tions a
r
e fu
rther
enha
n
c
ed
by line
a
r nulli
ng a
nd succe
s
sive interferen
ce
can
c
ell
a
tion
a
l
gorithm
calle
d Verti
c
al
- Be
ll-Lab
s Laye
r
ed Sp
ace-Ti
me Archite
c
tu
re
(V-BLAST
) to
achi
eve bette
r perfo
rma
n
ce.
Spectral effici
ency is the capa
city of t
he system whi
c
h sho
w
s the amount of m
a
ximum
informatio
n that can b
e
sent by a wire
less
com
m
un
ication
syste
m
. Conventio
nally, this ca
n be
increa
sed
by the facto
r
of
mi
n
,
RT
M
M
without u
s
in
g
addition
al tra
n
smits
po
we
r or spe
c
tral
band
width. T
h
is
pape
r
sh
ows that m
a
ximum
spe
c
tra
l
efficien
cy is achievable
b
y
increa
sing
th
e
numbe
r of ele
m
ent in each antenn
a ar
ra
y at both ends of radio lin
k.
Figure 2. Spectral Efficien
cy for the Prop
ose
d
Wirele
ss MIMO System whe
n
2
T
M
,
2
R
M
,
K=2 and
N=2
Figure 2
sho
w
s the
sp
ectral efficie
n
cy
p
e
rf
orm
a
n
c
e
o
f
the propo
se
d sy
stem
with
ZF a
nd
MMSE detection wh
en
2
T
M
,
2
R
M
,
K=2 an
d N=2. This
re
sult
indicates th
at the hybri
d
scheme
ha
s the be
st pe
rfo
r
man
c
e
with
averag
e
spe
c
tral efficien
cy
of 38.86b/s/
Hz
whe
n
MM
ES
wa
s con
s
ide
r
ed a
s
dete
c
ti
on an
d 33.0
8
b
/s/Hz for Z
F
detectio
n
tha
n
sp
atial mult
iplexing
sche
me
with the average sp
ect
r
al efficien
cy of 21.
73b/s/
Hz a
nd 14.24b/
s/Hz for MMSE
and ZF dete
c
tion
respe
c
tively; and
beamfo
rming
schem
e
with th
e
av
erage sp
ect
r
al efficien
cy
of 11.62b/
s/Hz. The
result furthe
r sho
w
s that the
Co
nventio
nal MIMO sy
stem with
2
T
M
and
2
R
M
has a
n
averag
e spe
c
tral efficien
cy
of 4.38b/s/Hz
when MMS
E detection
wa
s u
s
ed a
n
d
3.54b/
s/Hz for
ZF dete
c
tion
whi
c
h o
b
viou
sly indi
cate th
at the
Conve
n
tional MIM
O
schem
e h
a
s
a poo
r
ca
pa
city
perfo
rman
ce
compa
r
e
d
to the other sch
eme. T
hus, this p
r
oves that spatial multipl
e
xing
techni
que
ca
n be u
s
ed to
achieve
d
hi
gh data
ra
te
than bea
mforming techniq
ue and it furt
her
sho
w
s that hybrid schem
e
make u
s
e o
f
the adv
antages of both tech
niqu
e to prod
uce high
er
s
p
ec
tral effic
i
enc
y.
Figure 3 sho
w
s the
avera
ge erro
r prod
uce by the
propo
sed te
chn
i
que
s wh
en
2
T
M
,
2
R
M
, K=2 an
d
N=2. It is cle
a
r t
hat MMSE d
e
tection
pe
rfo
r
med
better t
han the
ZF d
e
tection
in the
entire
schem
e. Th
e result
sh
ows t
hat spatial m
u
ltiplexing
scheme pe
rforms
b
e
tter at
l
o
we
r
SNR while be
amformin
g scheme pe
rform better at
high SNR. Thi
s
limitation of the two sche
mes
wa
s com
pen
sated fo
r wit
h
hybrid sch
e
me. Thu
s
, hybrid sche
me
has go
od BER perfo
rma
n
ce
and will p
r
ovi
de an ave
r
ag
e error of 0.0
017 for MMS
E detection a
nd 0.002
4 for ZF detectio
n
.
The re
sult al
so shows tha
t
beamformin
g scheme
on
ly perform
s b
e
tter than hybrid sch
e
me
at
high SNR of 14dB, but it will provid
e a
high aver
a
g
e
error of 0.0
326 than hyb
r
id sche
me. It is
also
clea
r fro
m
the re
sult that conve
n
tio
nal
MIMO ha
ve a poor p
e
r
forma
n
ce a
m
ong the e
n
tire
scheme.
0
2
4
6
8
10
12
14
16
18
20
0
10
20
30
40
50
60
S
NR [
d
B
]
S
p
ec
t
r
a
l
E
f
f
i
c
i
en
c
y
[
b
ps
/
H
z
]
Conv
e
n
t
i
o
nal
M
I
M
O
(M
M
S
E
)
Conv
e
n
t
i
o
nal
M
I
M
O
(Z
F
)
Hy
bri
d
S
c
hem
e
(
M
M
S
E
)
Hy
bri
d
S
c
hem
e
(
Z
F
)
SM
Sc
h
e
m
e
(
M
M
SE)
SM
Sc
h
e
m
e
(
Z
F
)
BF
Sc
h
e
m
e
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 12, No. 7, July 201
4: 5458 – 54
68
5466
Figure 3. BER Perfo
r
man
c
e when
N=2
and
K=
2
Figure 4. BER Perfo
r
man
c
e when
N=2
and
K=
2 for V-BLAST
The perfo
rma
n
ce of the system wa
s furt
her
en
han
ce
d by V-BLAST algorithm a
s
sho
w
n
in the result produ
ce
d in
Figure 4. V-BL
AST improves the pe
rforman
c
e of MMSE and ZF
detectio
n
in t
he enti
r
e
sch
e
mes. T
he
h
y
brid
scheme
outpe
rform
s
other sche
m
e
s with
MMSE
detectio
n
hav
ing a b
e
tter a
v
erage
BER
of 0.0008
65
and ZF
dete
c
tion with a
n
a
v
erage
BER
of
0.0012. Thi
s
prove
s
that h
y
brid sch
e
me
has im
pr
oved in pe
rform
a
nce i
n
term
s
of error
red
u
ction
by 49.6% and
50% for MMSE and ZF detection resp
ectively with the aid of V-BLAST.
Figure 5. Spectral Efficien
cy for the Prop
ose
d
Wirel
e
s
s
MIM
O
Sy
stem wh
en
2
T
M
,
2
R
M
,
K=
4 and N=
4
Figure 6. Spectral Efficien
cy of the Propose
d
Wirel
e
s
s
MIM
O
Sy
stem
2
T
M
,
2
R
M
,
K=8 and
N=8
Figure 5
sh
o
w
s that the
spe
c
tral
effici
ency
of the
system i
n
cre
a
se
s lin
ea
r
with the
numbe
r of el
ement in e
a
c
h a
n
tenn
a
array. With the
R
M
and
T
M
antenna arrays remai
n
con
s
tant an
d element K an
d N we
re in
creased from t
w
o to four, the s
i
mulation res
u
lt shows
t
hat
the cap
a
city perfo
rman
ce
of hybrid sch
e
me is
b
e
tter than individu
al scheme
an
d it will provi
de
an ave
r
ag
e
spectral effici
e
n
cy
of
126.5
9
b
/s/Hz
whe
n
MMSE dete
c
tion i
s
u
s
e
d
a
nd 1
04.93
b/s/
Hz
for ZF dete
c
tion. This
sho
w
s that hyb
r
i
d
sc
hem
e ha
s an in
creme
n
t of 87.73b/
s/Hz for MM
SE
detectio
n
an
d 71.87
b/s/Hz for
ZF d
e
tection
wh
en
the anten
na
array elem
en
t was in
cre
a
sed.
This
wa
s furt
her p
r
ove
d
b
y
incre
a
si
ng
K and
N to
eight elem
en
ts as th
e result wa
s given
in
Figure 6. T
he hybri
d
schem
e ha
s
an average
spe
c
tral
efficien
cy of 3
60.11b/
s/Hz
and
0
2
4
6
8
10
12
14
16
18
20
10
-6
10
-5
10
-4
10
-3
10
-2
10
-1
10
0
S
NR [
d
B
]
BER
C
onv
ent
i
ona
l
M
I
M
O
(Z
F
)
C
onv
ent
i
ona
l
M
I
M
O
(M
M
S
E
)
H
y
bri
d
S
c
hem
e(Z
F
)
H
y
br
i
d
S
c
hem
e(
M
M
S
E
)
BF
Sc
h
e
m
e
S
M
S
c
hem
e(
Z
F
)
SM Sc
h
e
m
e
(
MM
SE
)
0
2
4
6
8
10
12
14
16
18
20
10
-6
10
-5
10
-4
10
-3
10
-2
10
-1
10
0
SN
R
[d
B
]
BER
H
y
b
r
i
d
Sc
h
e
me
(
MMSE
)
H
y
b
r
i
d
S
c
hem
e(
Z
F
)
BF
S
c
h
e
m
e
S
M
S
c
hem
e(
M
M
S
E
)
S
M
S
c
hem
e(
Z
F
)
0
2
4
6
8
10
12
14
16
18
20
0
20
40
60
80
10
0
12
0
14
0
16
0
SN
R
[
d
B]
S
pec
t
r
al
E
f
f
i
c
i
enc
y
[
bps
/
H
z
]
H
y
b
r
i
d
Sc
h
e
m
e
(
M
M
SE)
Hy
b
r
i
d
S
c
he
m
e
(
Z
F
)
S
M
S
c
h
e
me
(
MMS
E)
SM
Sch
e
m
e
(
Z
F
)
BF
Sch
e
m
e
0
2
4
6
8
10
12
14
16
18
20
0
50
100
150
200
250
300
350
400
450
SN
R
[
d
B]
S
pec
t
r
al E
f
f
i
c
i
enc
y
[
bps
/
H
z
]
H
y
b
r
i
d
S
c
hem
e(
M
M
S
E
)
H
y
b
r
i
d
S
c
hem
e(
Z
F
)
SM Sc
h
e
m
e
(
M
M
SE
)
S
M
S
c
hem
e(
Z
F
)
BF
Sc
h
e
m
e
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
Perform
a
n
c
e
Analysis of a
Hybrid Mim
o
Tech
niqu
e for High Data Rate… (Kehi
nd
e Ode
y
em
i)
5467
285.78
b/s/Hz for MMSE a
nd ZF
dete
c
tion respe
c
tively than oth
e
r
sch
e
me. T
hus, thi
s
p
r
o
v
es
that the
cap
a
city of MIM
O
sy
stem
ca
n be
enh
an
ced by i
n
cr
ea
sing
the
nu
mber of a
n
te
nna
element in ea
ch array at the transmitter
and re
ceive
r
.
Similarly, Fig
u
re 6
sh
ows the BER pe
rfor
ma
nce of
the syste
m
; whe
n
N
and
K were
increa
sed
fro
m
2 to 4
wh
iles the
ante
nna a
r
ray
R
M
a
nd
T
M
remains
cons
tant. The res
u
lt
prove
s
that
the in
crea
se
in the
anten
na a
r
ray ele
m
ent
N a
n
d
K produ
ce
d
better sy
ste
m
perfo
rman
ce.
It was sho
w
n that hybrid
sch
eme ha
s a better BER perfo
rma
n
c
e with ave
r
ag
e
error
of 0.00
0
434 fo
r MMS
E detectio
n
a
nd 0.0
0029
7
for ZF
dete
c
tion tha
n
any
of the
schem
es.
This
sho
w
s a
signifi
cant i
m
provem
ent i
n
BER comp
are to
whe
n
N an
d K eq
u
a
l to 2 an
d when
the syste
m
was e
nha
nced
with V-BLAS
T as i
n
figure
3 and
4 resp
ectively. Figu
re 8
sh
ows t
h
e
enha
ncement
of the system in this case
wi
th V-BLAST and better pe
rformance of hybrid
scheme
was re
co
rded
wit
h
an
averag
e BER of
0.0000
719
and
0.0002
02 fo
r MMSE a
n
d
ZF
detectio
n
re
spectively.
Figure 6. BER Perfo
r
man
c
e when
N=4
and
K=
4
Figure 7. BER Perfo
r
man
c
e when
N=4
and
K=
4 for V-BLAST
5. Conclusio
n
The p
e
rfo
r
ma
nce
analy
s
is
of a hybri
d
M
I
MO tech
niqu
e wa
s p
r
o
p
o
s
ed for
high
d
a
ta rate
wirel
e
ss com
m
unication system in this
pape
r. This
schem
e involves the combi
nation of sp
a
t
ial
multiplexing a
nd beamfo
rmi
ng techni
que
and wa
s u
s
e
d
as a tran
sm
issi
on sche
m
e
which ma
kes
the syste
m
to
be a
b
le to transmit
parall
e
l data
stre
a
m
s a
s
well
a
s
obtai
ning
b
eamformi
ng
gain.
The MMSE a
nd ZF MIM
O
detectio
n
al
gorithm
wa
s employed at the
re
ceive
r
and wa
s
furt
her
enha
nced by
V-BLAST. Spectral
efficie
n
c
y an
d BER
were
the
two perfo
rman
ce
metrics used to
determi
ne th
e efficien
cy o
f
the scheme
s
. The
simu
l
a
tion re
sult
s
sho
w
that th
e hybrid
sch
e
me
outperfo
rm
s
the individ
ual
sp
atial multi
p
lexing
and
beamfo
rming
schem
e; an
d ea
ch
of th
e
scheme
s
is
better than the Co
nventi
onal MIMO
scheme. It was foun
d tha
t
the higher
the
antenn
a arra
y element the highe
r the
system s
p
e
c
tral efficie
n
cy and the better the system
reliability. The results also show that the
MMSE detection has a better performance in all the
scheme
s
tha
n
the ZF dete
c
tion an
d eve
n
whe
n
enh
a
n
ce
d by V-BLAST.
Referen
ces
[1]
Proakis JG. Digital Communications.
4th e
d
. Ne
w
Y
o
rk: McGra
w
-
Hil
l, 200
1.
[2]
Shannon CE.
A mathematic
al t
heor
y
of com
m
unic
a
tion
– P
a
rt I & II.
Bell Syst. Tech. J.,
1948
27: 3
7
9
–
423.
[3]
F
o
schin
i GJ. L
a
yere
d space-t
i
me architect
u
r
e
for
w
i
r
e
less
communic
a
tio
n
in a fadin
g
e
n
viro
nment
w
h
e
n
usin
g mu
lti-elem
ent ant
enn
as.
Bell Labs Tech J.,
1996; 1(2): 41-5
9
.
[4]
Gans G, Gans MJ. On the limits of
w
i
reless
communic
a
ti
ons in
a fadi
n
g
envir
onm
ent
w
h
en
usin
g
multipl
e
ante
n
n
a
s. W
i
reless Person
al Comm
unic
a
tions. 1
9
9
8
; 6(3): 311
–35
5.
0
2
4
6
8
10
12
14
16
18
20
10
-6
10
-5
10
-4
10
-3
10
-2
10
-1
10
0
SN
R
[
d
B
]
BER
Conv
ent
i
onal
M
I
M
O
(
Z
F
)
Conv
ent
i
onal
M
I
M
O
(
M
M
S
E
)
Hy
bri
d
S
c
he
m
e
(
Z
F
)
H
y
bri
d
S
c
hem
e(M
M
S
E
)
B
F
S
c
hem
e
SM
Sc
h
e
m
e
(
Z
F
)
SM
Sc
h
e
m
e
(
M
M
S
E
)
0
2
4
6
8
10
12
14
16
18
20
10
-6
10
-5
10
-4
10
-3
10
-2
10
-1
10
0
SN
R
[d
B]
BER
Hy
bri
d
S
c
hem
e(
M
M
S
E
)
H
y
br
i
d
S
c
he
m
e
(Z
F
)
BF
S
c
h
e
m
e
SM
Sch
e
m
e
(
M
M
SE
)
S
M
S
c
hem
e(Z
F
)
Evaluation Warning : The document was created with Spire.PDF for Python.