T
E
L
KO
M
N
I
KA
T
e
l
e
c
om
m
u
n
icat
ion
,
Com
p
u
t
i
n
g,
E
lec
t
r
on
ics
an
d
Cont
r
ol
Vol.
18,
No.
1,
F
e
br
ua
r
y
2020,
pp.
4
19
~
4
26
I
S
S
N:
1693
-
6930,
a
c
c
r
e
dit
e
d
F
ir
s
t
G
r
a
de
by
Ke
me
nr
is
tekdikti
,
De
c
r
e
e
No:
21/E
/KP
T
/2018
DO
I
:
10.
12928/
T
E
L
KO
M
NI
KA
.
v18i1.
8720
419
Jou
r
n
al
h
o
mepage
:
ht
tp:
/
/j
our
n
al.
uad
.
ac
.
id/
index
.
php/T
E
L
K
OM
N
I
K
A
M
IM
O
c
h
an
n
e
ls
:
o
p
t
imiz
in
g
t
h
r
o
u
gh
p
u
t
a
n
d
r
e
d
u
c
i
n
g
o
u
t
age
b
y
i
n
c
r
e
asi
n
g
m
u
ltip
l
e
xi
n
g
g
ai
n
Ob
oye
r
u
lu
Agboj
e
,
Ns
ik
an
Nkordeh
,
Uz
airue
S
t
an
ley
I
d
iak
e
,
Ol
olad
e
Ol
ad
oyin
,
Kenn
e
d
y
Ok
ok
p
u
j
ie,
I
b
in
ab
o
B
ob
-
M
an
u
e
l
Co
v
e
n
an
t
U
n
i
v
er
s
i
t
y
,
N
i
g
er
i
a
Ar
t
icle
I
n
f
o
AB
S
T
R
ACT
A
r
ti
c
le
h
is
tor
y
:
R
e
c
e
ived
J
a
n
27
,
2019
R
e
vis
e
d
De
c
6
,
20
19
Ac
c
e
pted
De
c
2
1,
20
19
T
h
e
t
w
o
mai
n
ai
m
s
o
f
d
e
p
l
o
y
i
n
g
mu
l
t
i
p
l
e
i
n
p
u
t
m
u
l
t
i
p
l
e
o
u
t
(MIMO
)
are
t
o
ach
i
e
v
e
s
p
at
i
al
d
i
v
er
s
i
t
y
(i
m
p
ro
v
es
ch
an
n
el
rel
i
a
b
i
l
i
t
y
)
an
d
s
p
at
i
al
mu
l
t
i
p
l
e
x
i
n
g
(i
n
crea
s
e
d
a
t
a
t
h
ro
u
g
h
p
u
t
).
A
c
h
i
e
v
i
n
g
b
o
t
h
i
n
a
g
i
v
en
s
y
s
t
em
i
s
i
mp
o
s
s
i
b
l
e
fo
r
n
o
w
,
an
d
a
t
rad
e
-
o
ff
h
as
t
o
b
e
reac
h
ed
as
t
h
e
y
may
b
e
co
n
f
l
i
c
t
i
n
g
o
b
j
ec
t
i
v
es
.
T
h
e
b
as
i
c
co
n
ce
p
t
o
f
mu
l
t
i
p
l
ex
i
n
g
:
d
i
v
i
d
e
(m
u
l
t
i
p
l
e
x
)
t
ran
s
mi
t
a
d
at
a
s
t
ream
s
ev
era
l
b
ran
c
h
es
a
n
d
t
r
an
s
m
i
t
v
i
a
s
ev
era
l
(i
n
d
ep
e
n
d
e
n
t
)
ch
a
n
n
e
l
s
.
In
t
h
i
s
p
a
p
er,
w
e
fo
cu
s
ed
mai
n
l
y
o
n
ach
i
e
v
i
n
g
s
p
a
t
i
a
l
mu
l
t
i
p
l
e
x
i
n
g
b
y
m
o
d
e
l
i
n
g
t
h
e
ch
a
n
n
e
l
u
s
i
n
g
t
h
e
d
i
ag
o
n
a
l
Bel
l
L
ab
s
s
p
ace
t
i
me
s
c
h
eme
(D
-
BL
A
ST
)
an
d
t
h
e
v
ert
i
cal
Bel
l
L
ab
s
s
p
ace
t
i
me
arch
i
t
ect
u
re
(V
-
BL
A
ST
)
Mat
l
ab
s
i
m
u
l
a
t
i
o
n
s
res
u
l
t
s
w
ere
a
l
s
o
g
i
v
e
n
t
o
fu
rt
h
er
co
mp
are
t
h
e
ad
v
an
t
a
g
es
o
f
s
p
a
t
i
a
l
mu
l
t
i
p
l
e
x
i
n
g
.
K
e
y
w
o
r
ds
:
Dive
r
s
it
y
M
I
M
O
M
ult
ipl
e
xing
R
e
li
a
bil
it
y
S
pa
ti
a
l
T
hr
oughput
Th
i
s
i
s
a
n
o
p
en
a
c
ces
s
a
r
t
i
c
l
e
u
n
d
e
r
t
h
e
CC
B
Y
-
SA
l
i
ce
n
s
e.
C
or
r
e
s
pon
din
g
A
u
th
or
:
Uz
a
ir
ue
S
tanle
y
I
diake
,
C
ove
na
nt
Unive
r
s
it
y,
Nige
r
ia.
E
mail:
S
tanle
y
.
uz
a
ir
ue
@c
ove
na
ntuni
ve
r
s
it
y.
e
du.
n
g
,
uz
a
ir
ue
.
s
tanle
y@gmail.
c
om
1.
I
NT
RODU
C
T
I
ON
T
he
ne
e
d
f
or
a
nd
da
ta
r
a
tes
a
nd
a
high
qua
li
ty
of
s
e
r
vice
(
QoS)
.
Ove
r
the
ye
a
r
s
,
the
ubiqu
it
y
of
f
e
r
e
d
by
wir
e
les
s
c
omm
unica
ti
on
ha
s
made
it
the
mo
r
e
pr
e
f
e
r
r
e
d
mea
ns
ove
r
wir
e
d;
he
nc
e
,
ther
e
ha
s
be
e
n
a
n
incr
e
a
s
e
in
r
e
s
e
a
r
c
h
on
how
to
im
pr
ove
the
modul
a
ti
on
s
c
he
mes
u
s
e
d
ove
r
the
a
ir
int
e
r
f
a
c
e
.
M
ult
ipl
e
input
mul
ti
ple
output
(
M
I
M
O)
o
f
f
e
r
s
de
s
ir
a
ble
p
r
ope
r
ti
e
s
that
mee
t
mos
t
o
f
the
r
e
quir
e
ment
s
tate
d
a
bove
.
B
y
us
ing
mul
ti
ple
output
mul
t
ipl
e
in
put
(
M
I
M
O
)
s
ys
tems
,
diver
s
it
y
ga
in
mi
ti
ga
tes
f
a
ding,
incr
e
a
s
e
s
c
ove
r
a
ge
a
nd
im
pr
ove
s
QoS.
M
ult
ipl
e
xing
ga
in
inc
r
e
a
s
e
s
c
a
pa
c
it
y
a
nd
s
pe
c
tr
a
l
e
f
f
icie
nc
y
wi
th
no
a
ddit
ional
p
owe
r
or
ba
ndwidth
e
xpe
ndit
ur
e
[
1]
.
T
he
c
or
e
idea
und
e
r
the
M
I
M
O
s
ys
tems
is
the
a
bil
it
y
to
tur
n
m
ult
i
-
pa
th
pr
opa
ga
ti
on,
whic
h
is
typi
c
a
ll
y
a
n
obs
tac
le
in
c
onve
nti
ona
l
wir
e
les
s
c
omm
unica
ti
on,
int
o
a
be
ne
f
it
f
or
us
e
r
s
[
2]
.
W
it
h
M
I
M
O,
the
c
a
pa
c
it
y
of
a
c
om
muni
c
a
ti
on
s
ys
tem
incr
e
a
s
e
s
li
ne
a
r
ly
with
the
n
umber
of
a
ntenna
s
,
ther
e
by
a
c
hieving
a
n
incr
e
a
s
e
i
n
s
pe
c
tr
a
l
e
f
f
icie
nc
y,
without
r
e
qui
r
ing
mor
e
r
e
s
our
c
e
s
in
t
e
r
ms
of
ba
ndwidth
a
nd
powe
r
[
3
-
5]
.
F
r
om
F
igur
e
1
s
hows
that
M
I
M
O
tec
hnology
ha
s
t
wo
main
objec
ti
ve
s
whic
h
it
a
im
s
to
a
c
hieve
:
high
s
pa
ti
a
l
mul
ti
plexing
ga
in
a
nd
high
s
pa
ti
a
l
diver
s
i
ty.
T
o
a
tt
a
in
s
pa
ti
a
l
mul
ti
plexing,
the
s
ys
tem
is
made
to
c
a
r
r
y
mul
ti
p
le
da
ta
s
tr
e
a
m
ove
r
one
f
r
e
que
nc
y,
s
im
ult
a
ne
ous
ly
-
f
or
m
mul
ti
ple
indepe
nde
nt
li
nks
(
on
s
a
me
c
ha
nne
l)
be
twe
e
n
t
r
a
ns
mi
tt
e
r
a
nd
r
e
c
e
iver
to
c
o
mm
unica
te
a
t
higher
da
ta
r
a
tes
.
I
n
low
S
NR
e
nv
ir
onment,
s
pa
ti
a
l
d
iver
s
it
y
tec
hniques
a
r
e
a
ppli
e
d
to
mi
ti
ga
te
f
a
ding
a
nd
the
pe
r
f
or
manc
e
ga
in
is
typ
ica
ll
y
e
xpr
e
s
s
e
d
a
s
diver
s
it
y
ga
in
(
in
dB
)
[
6
]
;
f
or
higher
S
NR
f
a
c
il
it
a
tes
the
us
e
of
s
pa
ti
a
l
mul
ti
plexing
(
S
M
)
,
i.
e
.
,
the
tr
a
ns
mi
s
s
ion
of
pa
r
a
ll
e
l
da
ta
s
tr
e
a
ms
,
a
nd
in
f
or
ma
ti
on
theor
e
ti
c
c
a
pa
c
it
y
in
bit
s
pe
r
s
e
c
ond
pe
r
He
r
tz
(
bit
s
/s
/Hz
)
is
the
pe
r
f
or
manc
e
mea
s
ur
e
o
f
c
hoice
[
7]
.
S
pa
ti
a
l
diver
s
it
y
wor
ks
on
the
pr
inciple
of
t
r
a
n
s
mi
s
s
ion
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
1693
-
6930
T
E
L
KO
M
NI
KA
T
e
lec
omm
un
C
omput
E
l
C
ontr
o
l
,
Vol.
18
,
No
.
1
,
F
e
br
ua
r
y
2020
:
4
19
-
4
26
420
of
s
tr
uc
tur
e
d
r
e
dunda
nc
y
im
pr
ove
s
s
ignal
qua
li
ty,
a
nd
is
a
ble
to
a
c
hieve
a
high
s
ignal
to
nois
e
r
a
ti
o
(
S
NR
)
,
whic
h
tr
a
ns
late
s
to
high
r
e
li
a
bil
it
y
[
8]
.
F
ig
ur
e
1
.
S
c
he
matic
r
e
pr
e
s
e
ntation
of
M
I
M
[
8]
2.
RE
S
E
AR
CH
M
E
T
HO
D
I
n
today’
s
inf
or
mation
a
nd
c
omm
unica
ti
on
tec
hnology
wor
ld
,
ba
ndwidth
a
nd
c
a
pa
c
it
y,
a
pa
r
t
f
r
om
s
e
c
ur
it
y,
a
r
e
the
mos
t
im
po
r
tant
p
he
nomena
f
or
a
ny
da
ta
r
e
late
d
a
c
ti
vit
y
.
M
os
t
I
T
r
e
late
d
pr
oc
e
s
s
e
s
ha
ve
a
high
a
f
f
ini
ty
f
o
r
lar
ge
ba
ndwidth,
a
nd
thi
s
plac
e
s
a
s
tr
ingent
r
e
quir
e
ment
on
the
c
a
p
a
c
it
y
of
the
c
omm
unica
ti
on
c
ha
nne
l.
Ac
c
or
ding
to
the
S
ha
nnon
-
Ha
r
tl
e
y
th
e
or
e
m:
the
C
a
pa
c
it
y
C
o
f
a
r
a
dio
c
ha
nne
l
is
de
pe
nde
nt
on
the
ba
ndwidth
B
a
nd
the
s
ignal
to
no
is
e
r
a
ti
on
.
(
1)
M
I
M
O
tec
hnology
im
pr
ove
wir
e
les
s
c
omm
unica
ti
on
c
ha
nne
l
by
two
main
pr
oc
e
s
s
e
s
:
−
C
ombating
mul
ti
pa
th
s
c
a
tt
e
r
ing
in
the
c
omm
unica
t
ion
c
ha
nne
l.
−
E
xploi
ti
ng,
mul
ti
pa
th
s
c
a
tt
e
r
ing
in
the
c
omm
unica
t
ion
c
ha
nne
l.
I
n
thi
s
pa
pe
r
,
we
li
mi
ted
our
wor
k
to
s
pa
ti
a
l
mul
ti
plexing
-
a
mathe
matica
l
model
f
or
S
pa
ti
a
l
M
ult
ipl
e
xing
is
p
r
e
s
e
nted
f
oll
owe
d
by
s
e
r
ies
of
s
im
ulations
us
ing,
M
AT
L
AB
.
T
he
mul
ti
plexin
g
ga
in
is
r
e
s
pons
ibl
e
f
or
M
I
M
O
s
ys
tems
of
f
e
r
ing
a
li
ne
a
r
incr
e
a
s
e
in
the
a
c
hieva
ble
da
ta
r
a
te.
I
nde
e
d,
in
a
M
I
M
O
c
ha
nne
l,
mul
ti
ple
indepe
nde
nt
da
ta
s
tr
e
a
ms
c
a
n
be
tr
a
ns
mi
tt
e
d
withi
n
the
ba
ndwidth
of
ope
r
a
ti
on
a
n
d,
unde
r
s
uit
a
ble
c
ha
nne
l
c
ondit
ions
,
thes
e
c
a
n
be
s
e
pa
r
a
ted
a
t
the
r
e
c
e
iver
.
F
ur
ther
mor
e
,
e
a
c
h
da
ta
s
tr
e
a
m
e
xpe
r
ienc
e
s
a
t
lea
s
t
the
s
a
me
c
ha
nne
l
qua
li
ty
that
would
be
e
xpe
r
ienc
e
d
by
a
S
I
S
O
s
ys
tem,
e
f
f
e
c
ti
ve
ly
e
nha
nc
ing
the
c
a
pa
c
it
y
by
a
mul
t
ipl
ica
ti
ve
f
a
c
tor
e
qua
l
to
the
number
of
e
s
tablis
he
d
s
tr
e
a
ms
[
8]
.
T
o
be
mo
r
e
s
pe
c
if
i
c
,
we
f
oc
us
on
the
high
S
NR
r
e
gim
e
,
a
nd
thi
nk
of
a
S
c
he
me
a
s
a
f
a
mi
ly
of
c
ode
s
,
one
f
or
e
a
c
h
S
NR
leve
l
.
A
s
c
he
me
is
s
a
id
to
ha
ve
a
s
pa
ti
a
l
mul
t
ipl
e
xing
ga
in
r
a
nd
a
diver
s
it
y
a
dva
ntage
d
if
the
r
a
te
of
the
s
c
he
me
s
c
a
les
li
ke
r
log
S
NR
a
nd
the
a
ve
r
a
ge
e
r
r
o
r
pr
oba
bil
it
y
de
c
a
ys
li
ke
1/
S
NR
[
8]
.
is
the
nt
x
nt
de
ter
mi
n
is
ti
c
matr
ix
.
F
o
r
a
M
I
M
O
S
ys
te
m,
the
tr
a
ns
mi
t
-
r
e
c
e
ive
s
ys
tem
is
r
e
pr
e
s
e
nted
by:
(
2)
w
he
r
e
:
,
white
Ga
us
s
ian
nois
e
the
matr
ix
T
o
ha
ve
a
n
ins
ight
o
f
s
pa
ti
a
l
mul
ti
plexing
pr
ope
r
t
y,
we
de
c
ouple
e
qua
ti
on
(
9
)
us
ing
s
ome
c
omm
on
matr
ix
tr
a
ns
f
or
mation.
Us
ing
s
ingul
a
r
va
lue
de
c
om
pos
it
ion
(
S
VD
)
,
c
a
n
be
wr
it
ten
a
s
:
(
3)
w
he
r
e
a
r
e
uni
tar
y
mat
r
ice
s
.
is
a
r
e
c
tangula
r
matr
ix
whos
e
diagona
l
e
leme
nts
a
r
e
non
-
ne
ga
ti
ve
r
e
a
l
number
s
a
nd
whos
e
of
f
-
diagona
l
e
leme
nts
a
r
e
z
e
r
o.
T
he
diagona
l
e
lem
e
nts
.
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KO
M
NI
KA
T
e
lec
omm
un
C
omput
E
l
C
ontr
o
l
M
I
M
O
c
hanne
ls
:
opti
miz
ing
thr
oughput
and
r
e
duc
ing
outage
.
.
.
.
(
Oboy
e
r
ulu
A
gbo
je)
421
a
r
e
the
e
igenva
lues
of
the
mat
r
ix
a
nd
a
ls
o
of
.
T
he
S
VD
c
a
n
be
r
e
wr
it
ten
a
s
:
(
4)
the
c
ha
r
a
c
ter
is
ti
c
of
a
M
I
M
O
s
ys
tem
is
de
ter
m
in
e
d
by
it
s
mul
ti
p
lexing
ga
in
a
nd
it
s
diver
s
it
y
ga
in
.
A
S
c
he
me
is
s
a
id
to
a
c
hieve
s
pa
ti
a
l
mul
ti
plexing
ga
i
n
a
nd
diver
s
it
y
ga
in
if
the
da
ta
r
a
te:
(
5)
a
nd
a
ve
r
a
ge
e
r
r
or
pr
oba
bil
it
y:
(
6)
T
he
diver
s
it
y
or
de
r
mea
s
ur
e
s
how
many
s
tatis
ti
c
a
ll
y
indepe
nde
nt
c
opies
o
f
the
s
a
me
s
ymbol
the
r
e
c
e
iver
is
a
ble
to
ge
t
in
or
de
r
to
r
e
pr
oduc
e
a
r
e
li
a
ble
e
s
ti
mate
of
the
tr
a
ns
mi
tt
e
d
s
y
m
b
ol
.
As
s
ume
the
f
a
ding
c
oe
f
f
icie
nt
matr
ix
H
is
known
to
the
r
e
c
e
iver
,
the
c
ha
nne
l
c
a
pa
c
it
y
(
bps
/Hz
)
of
a
s
ys
tem
with
tr
a
ns
mi
t
a
nd
r
e
c
e
ive
a
ntenna
s
is
given
by:
=
(
7)
int
r
oduc
ing
2
bounda
r
ies
f
or
li
mi
ts
.
F
or
the
c
a
s
e
of
,
a
t
high
(
8)
ther
e
is
a
de
c
r
e
a
s
e
in
e
r
r
o
r
p
r
oba
bil
it
y
a
s
we
incr
e
a
s
e
the
number
of
a
ntenna
s
,
whic
h
mea
n
a
n
in
c
r
e
a
s
e
in
S
NR
will
r
e
s
ult
in
a
de
c
r
e
a
s
e
in
e
r
r
or
P
r
oba
bi
li
ty
;
we
a
ls
o
a
c
hi
ev
e
d
e
c
r
e
a
s
e
e
r
r
or
pr
oba
bil
it
y
by
d
e
ployi
ng
higher
modul
a
ti
on
s
c
he
mes
.
A
plot
o
f
the
s
ymbol
e
r
r
or
r
a
te
(
S
E
R
)
to
the
S
NR
g
ives
a
lot
of
im
por
tant
a
nd
r
e
ve
a
li
ng
inf
o
r
mation;
the
s
lope
o
f
thi
s
c
ur
ve
giv
e
s
us
the
va
lue
o
f
the
diver
s
it
y
ga
in
.
T
he
m
u
l
ti
p
lexing
ga
in
gives
a
mea
s
ur
e
of
how
f
a
s
t
s
pe
c
tr
a
l
e
f
f
icie
nc
y
c
a
n
incr
e
a
s
e
with
incr
e
a
s
e
of
while
ke
e
ping
the
s
a
me
e
r
r
or
r
a
te;
the
c
or
r
e
s
ponds
to
the
maxi
mum
number
o
f
indepe
nde
nt
laye
r
s
of
pa
r
a
ll
e
l
c
h
a
nne
l
[
9]
,
a
nd
is
li
mi
te
d
by:
(
9)
the
mul
ti
plexing
ga
in
o
f
a
M
I
M
O
s
ys
tems
de
pe
nds
on
the
type
of
modul
a
ti
on
s
c
he
me
a
nd
the
S
NR
.
T
he
p
r
oba
bil
i
ty
of
e
r
r
o
r
f
or
a
n
N=
M
=
1
s
ys
tem
us
ing
P
S
K
modul
a
ti
on
is
given
by
[
8]
a
s
:
(
10)
f
or
a
s
ys
tem
w
i
th
two
r
e
c
e
iver
s
tr
a
ns
mi
tt
ing
the
s
a
me
s
ignal,
the
e
r
r
or
pr
oba
bil
it
y
is
given
a
s
:
(
11)
it
c
a
n
be
s
hown
f
r
om
s
pe
c
tr
a
l
a
na
lys
i
s
that
the
da
t
a
r
a
te
(
mul
ti
plexing
ga
in)
of
a
S
I
S
O
s
ys
tem
us
ing
4
-
P
AM
modul
a
ti
on
s
c
he
me
(
2bit
/s
/Hz
)
c
a
n
be
i
mpr
ove
d
withou
t
a
de
p
r
e
c
iation
in
e
r
r
or
r
a
te
(
ke
e
ping
the
s
a
me
e
r
r
or
)
by
s
witch
to
a
n
8
-
P
AM
modul
a
ti
o
ns
s
c
he
me
(
3bit
s
/s
/Hz
)
is
the
S
NR
is
incr
e
a
s
e
d
by
6
dB
;
a
f
ur
ther
6
dB
incr
e
a
s
e
in
the
S
NR
with
the
us
e
o
f
64
-
QA
M
modul
a
ti
on
s
c
he
me
(
3
bit
s
/s
/Hz
)
will
incr
e
a
s
e
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
1693
-
6930
T
E
L
KO
M
NI
KA
T
e
lec
omm
un
C
omput
E
l
C
ontr
o
l
,
Vol.
18
,
No
.
1
,
F
e
br
ua
r
y
2020
:
4
19
-
4
26
422
the
mul
tip
lex
ing
ga
in
while
ke
e
ping
the
e
r
r
o
r
r
a
te
the
s
a
me
a
s
whe
n
4
-
P
AM
modul
a
ti
on
s
c
h
e
me
wa
s
us
e
d
[
9]
.
E
r
godic
c
a
pa
c
it
y
r
e
pr
e
s
e
nts
the
maximum
a
c
hieva
ble
thr
oughput
a
ve
r
a
ge
d
a
c
r
os
s
a
ll
f
a
ding
c
ondit
ions
.
M
ult
i
-
pa
th
f
a
ding
pr
ope
r
ti
e
s
of
a
c
ha
nne
l
c
ha
nge
ove
r
r
e
la
tiv
e
ly
longer
ti
mes
c
a
les
(
c
om
pa
r
e
d
to
r
e
c
e
iver
nois
e
)
a
nd
the
f
a
ding
s
tate
s
tays
f
a
ir
ly
c
ons
tant
ove
r
a
n
int
e
r
va
l
c
a
ll
e
d
the
c
ha
nne
l
c
ohe
r
e
nc
e
int
e
r
va
l.
R
e
li
a
ble
c
omm
unica
ti
on
a
t
r
a
tes
a
r
bit
r
a
r
i
ly
c
los
e
to
the
e
r
god
ic
c
a
pa
c
it
y
r
e
quir
e
s
a
ve
r
a
gin
g
a
c
r
os
s
man
y
i
nde
pe
nde
nt
r
e
a
li
z
a
ti
ons
o
f
the
c
ha
nne
l
ga
i
ns
ove
r
ti
me.
S
ince
the
c
ha
nne
l
c
a
pa
c
it
y
incr
e
a
s
e
s
li
ne
a
r
ly
with
log
S
NR
,
in
or
de
r
to
a
c
hieve
a
c
e
r
tain
f
r
a
c
ti
o
n
of
the
c
a
pa
c
it
y
a
t
high
S
NR
,
we
s
hould
c
ons
ider
s
c
he
mes
that
s
uppor
t
a
da
ta
r
a
te
whic
h
a
ls
o
inc
r
e
a
s
e
s
wi
th
S
NR
.
He
r
e
,
we
thi
nk
of
a
s
c
he
me
a
s
a
f
a
mi
ly
of
c
ode
s
f
C
(
S
NR
)
g
of
block
length
T
,
one
a
t
e
a
c
h
S
NR
leve
l.
L
e
t
R
(
S
NR
)
(
bit
s
/s
ymbol
)
be
the
r
a
te
of
t
he
c
ode
C
(
S
NR
)
.
W
e
s
a
y
that
a
s
c
he
me
a
c
hieve
s
a
s
pati
al
multi
plex
ing
gain
o
f
r
i
f
the
s
uppor
ted
da
t
a
r
a
te
[
1
0
,
11]
.
(
12)
S
pa
ti
a
l
mul
ti
plexing
ga
in
c
a
n
a
ls
o
be
thought
a
s
the
da
ta
r
a
te
nor
malize
d
with
r
e
s
pe
c
t
to
the
S
NR
leve
l.
A
c
omm
on
wa
y
to
c
ha
r
a
c
ter
ize
the
pe
r
f
or
manc
e
of
a
c
omm
unica
ti
on
s
c
he
me
is
to
c
omput
e
the
e
r
r
or
pr
oba
bil
it
y
a
s
a
f
unc
ti
on
o
f
S
NR
f
or
a
f
ixed
da
ta
r
a
te.
A
plot
of
e
r
r
or
p
r
oba
bil
it
y
a
ga
ins
t
nor
malize
d
S
NR
wa
s
pr
opos
e
d
by
F
or
ne
y
in
other
to
c
ompar
e
thes
e
s
c
he
mes
f
a
ir
ly
.
(
13
)
w
he
r
e
is
the
c
a
pa
c
it
y
of
the
c
ha
nne
l
a
s
a
fu
nc
t
ion
of
S
NR
.
mea
s
ur
e
s
how
f
a
r
the
is
a
bove
the
mi
nim
a
l
r
e
quir
e
d
to
s
uppor
t
the
tar
ge
t
da
ta
r
a
te.
Anothe
r
wa
y
to
c
ha
r
a
c
t
e
r
ize
the
pe
r
f
or
manc
e
o
f
a
M
I
M
O
s
ys
tem
is
to
plot
e
r
r
or
pr
oba
bil
it
y
a
ga
ins
t
nor
malize
d
da
ta
r
a
te,
a
f
ixed
.
(
14)
Notice
that
a
t
high
S
NR
,
the
c
a
pa
c
it
y
of
the
mul
ti
ple
a
ntenna
c
ha
nne
l
is
C
(
S
NR
).
K
log
S
N
R
;
he
nc
e
the
s
pa
ti
a
l
mul
ti
plexing
ga
in
.
(
15)
T
wo
of
the
wa
ys
mul
t
ipl
e
xing
ga
in
c
a
n
be
a
c
hieve
d
a
r
e
:
V
-
B
LA
S
T
Ve
r
ti
c
a
l
B
e
ll
L
a
bs
S
pa
c
e
T
im
e
a
r
c
hit
e
c
tur
e
a
nd
D
-
B
L
AST
,
Dia
gona
l
B
e
ll
L
a
b
S
p
a
c
e
T
im
e
S
c
he
me
[
10
,
11
]
.
2.
1.
V
-
B
L
AST
T
he
ve
r
ti
c
a
l
B
e
ll
L
a
bs
s
p
a
c
e
ti
me
a
r
c
hit
e
c
tur
e
(
V
-
B
L
AST
)
a
tt
e
mpt
s
to
maximi
z
e
s
ys
tem
thr
oughput
a
t
the
c
os
t
of
a
ve
r
a
ging
a
c
r
os
s
f
e
we
r
f
a
d
in
g
c
oe
f
f
icie
nts
.
Unde
r
V
-
B
L
AST
,
the
mes
s
a
g
e
is
s
pli
t
a
c
r
os
s
two
c
ode
d
s
tr
e
a
ms
.
X
is
r
e
pr
e
s
e
nted
a
s
f
oll
ows
:
(
16)
whe
r
e
a
r
e
c
omponents
of
two
indepe
nde
nt
a
c
hi
e
ving
c
ode
with
r
a
tes
r
e
s
pe
c
ti
ve
ly.
F
or
both
s
tr
e
a
ms
,
the
tr
an
s
mi
t
a
ntenn
a
r
e
mains
f
ixed
a
nd
c
oding
is
pe
r
f
or
med
only
a
c
r
os
s
ti
me.
Unlike
in
the
Ala
mout
i
s
c
he
me
whe
r
e
two
inf
or
mation
s
ymbol
s
a
ppe
a
r
or
thogonal
a
t
the
r
e
c
e
ive
r
due
to
the
s
pe
c
ial
c
ons
tr
uc
ti
on
of
X.
I
n
V
-
B
L
AST
,
s
uc
h
a
c
ons
tr
uc
ti
on
is
not
us
e
d
a
nd
t
he
re
c
e
iver
e
s
ti
mat
e
s
a
nd
f
r
om
it
s
r
e
c
e
ived
s
ignal
us
ing
s
ignal
pr
oc
e
s
s
ing
tec
hniques
.
F
or
a
2
x
2
M
I
M
O
c
ha
nne
l
with
V
-
B
L
AST
de
c
ompos
e
s
the
a
na
lys
is
two
S
I
S
O
pa
r
a
ll
e
l
s
ub
-
c
ha
nne
ls
ha
ving
c
ompl
e
x
ga
ins
a
nd
r
es
pe
c
ti
ve
ly.
T
he
outage
p
r
oba
bil
it
y
f
or
thi
s
s
c
he
me
is
given
a
s
[
12
-
15]
:
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KO
M
NI
KA
T
e
lec
omm
un
C
omput
E
l
C
ontr
o
l
M
I
M
O
c
hanne
ls
:
opti
miz
ing
thr
oughput
and
r
e
duc
ing
outage
.
.
.
.
(
Oboy
e
r
ulu
A
gbo
je)
423
(
17)
2.
2.
D
-
B
L
AST
T
he
diagona
l
B
e
ll
L
a
bs
s
pa
c
e
ti
me
s
c
he
me
(
D
-
B
L
AST
)
us
e
s
a
d
iagona
l
s
tr
uc
tur
e
a
nd
pe
r
f
o
r
ms
c
oding
a
c
r
os
s
a
ntenna
s
a
s
we
ll
a
s
s
pa
c
e
.
T
he
s
e
q
ue
nc
e
f
o
r
ms
one
c
ode
wor
d
be
longi
ng
to
a
c
a
pa
c
it
y
a
c
hieving
outer
c
ode
o
f
r
a
te
.
E
a
c
h
c
ode
wor
d
is
then
s
pli
t
int
o
two
blocks
of
s
ymbol
s
a
n
d
,
e
a
c
h
ha
ving
s
ymbol
s
[
15
-
18]
.
T
he
tr
a
ns
mi
tt
e
r
ther
e
f
or
e
s
e
nds
out
the
f
ol
lowin
g
c
ode
w
o
r
d
given
a
s
:
(
18)
in
or
de
r
to
ini
ti
a
l
ize
the
diagona
l
block
s
tr
uc
tur
e
s
hown
a
bove
,
one
ini
ti
a
l
block
of
.
S
ymbol
s
i.
e
.
is
s
e
t
to
.
I
f
one
c
ons
ider
s
the
inf
or
mation
r
a
te
f
o
r
D
-
B
L
AST
ove
r
a
s
uf
f
icie
nt
ly
lar
ge
number
of
c
od
e
wor
ds
,
thi
s
one
-
ti
me
ove
r
he
a
d
ha
s
ne
gli
gibl
e
e
f
f
e
c
t.
T
he
diagona
l
s
tr
uc
tur
e
a
ls
o
a
ll
ows
s
uc
c
e
s
s
ive
int
e
r
f
e
r
e
nc
e
c
a
nc
e
ll
a
ti
on;
us
ing
D
-
B
L
AST
,
the
M
I
M
O
c
ha
nne
l
de
c
ompos
e
s
int
o
two
pa
r
a
ll
e
l
S
I
S
O
s
ub
-
c
ha
nne
ls
ha
ving
ga
ins
a
nd
[
19]
.
T
he
outage
p
r
oba
bil
it
y
f
or
D
-
B
L
A
S
T
is
given
a
s
f
oll
ows
[
14
-
16]
:
(
19)
the
joi
nt
dis
tr
ibut
ion
of
a
nd
is
de
ter
mi
ne
d
by
th
e
r
e
c
e
iver
a
r
c
hit
e
c
tur
e
.
Outa
ge
p
r
oba
bil
it
y
is
a
nother
s
tanda
r
d
pe
r
f
or
manc
e
c
r
it
e
r
i
on
de
n
ot
e
d
by
P
out
a
nd
de
f
ined
a
s
the
pr
oba
bil
i
ty
that
the
ins
tanta
ne
ous
c
ha
nn
e
l
c
a
pa
c
it
y
be
low
a
s
p
e
c
i
f
ied
va
lue,
or
e
quivale
ntl
y
,
the
pr
oba
bil
it
y
that
th
e
output
S
NR
(
or
S
I
NR
)
f
a
ll
s
be
low
a
p
r
e
-
de
f
ined
a
c
c
e
ptable
thr
e
s
hold.
M
a
thema
ti
c
a
ll
y
s
pe
a
king,
the
outage
p
r
oba
bil
it
y
is
the
c
.
d.
f
.
of
S
NR
e
va
luate
d
a
t
the
s
pe
c
if
ied
thr
e
s
hold,
i.
e
.
(
20)
is
the
pr
e
de
f
ined
th
r
e
s
hold
a
nd
is
the
p.
d
.
f
o
f
S
N
R
.
T
he
a
ve
r
a
ge
S
E
R
,
de
noted
by
P
S
E
R
,
is
the
one
th
a
t
is
mos
t
r
e
ve
a
li
n
g
a
bou
t
the
na
tu
r
e
of
the
s
ys
tem
be
ha
vior
a
nd
is
ge
ne
r
a
ll
y
the
mos
t
di
f
f
icult
pe
r
f
o
r
manc
e
c
r
it
e
r
ion
to
c
omput
e
.
I
t
is
de
f
ined
a
s
the
pr
oba
bil
it
y
that
a
tr
a
ns
mi
tt
e
d
da
ta
s
ymbol
is
de
tec
ted
in
e
r
r
o
r
a
t
the
r
e
c
e
iver
.
T
he
S
E
R
is
typi
c
a
ll
y
modul
a
ti
on/d
e
tec
ti
on
s
c
he
me
de
pe
n
de
nt,
a
nd
is
dir
e
c
tl
y
r
e
late
d
to
the
ins
tanta
ne
ous
S
NR
(
or
S
I
NR
f
or
mul
ti
us
e
r
s
ys
tems
)
[
20
-
25]
.
3.
S
I
M
UL
AT
I
ON
RE
S
UL
T
S
F
igur
e
2
s
hows
the
s
im
ulation
r
e
s
ult
of
the
a
c
hi
e
va
ble
r
a
te
tr
a
ns
mi
s
s
ion
to
the
a
ve
r
a
ge
S
NR
of
M
I
M
O
r
e
c
e
iver
s
unde
r
f
a
s
t
f
a
ding
c
ha
nn
e
l
c
on
di
ti
on.
F
r
om
th
is
gr
a
ph
we
c
a
n
s
e
e
that
the
S
I
C
ha
s
th
e
highes
t
S
NR
(
s
ignal
to
Nois
e
r
a
ti
o)
while
M
M
S
E
a
nd
Z
e
r
o
-
f
or
c
ing
a
r
e
o
f
thi
s
va
lue
of
S
NR
.
F
igur
e
3
s
hows
how
the
pe
r
f
or
manc
e
im
p
r
ove
s
with
incr
e
a
s
ing
mul
ti
plexing
ga
ins
,
whe
r
e
the
r
e
c
e
ivi
n
g
e
nd
us
e
s
S
IS
O,
qua
s
i
-
or
thogonal
,
S
T
B
C
a
nd
Or
thogonal
S
T
B
;
th
e
B
E
R
is
be
s
t
f
or
the
r
e
c
e
ivi
ng
s
ys
tem
us
ing
o
r
t
hogona
l
S
T
B
C
a
nd
wor
s
t
f
or
the
S
I
S
O
.
F
igur
e
4
s
ho
ws
the
s
im
ulation
r
e
s
ult
o
f
the
outage
p
r
oba
bil
it
y
f
o
r
1
-
s
tr
e
a
m
-
4x4
M
I
S
O
R
a
yleigh
to
the
a
ve
r
a
ge
S
N
R
unde
r
s
low
f
a
ding
c
ondit
ion.
T
ha
t
is
whe
n
the
f
a
d
ing
r
a
te
of
the
S
NR
is
be
low
nor
mal
.
T
he
s
im
ulations
in
F
igur
e
5
a
nd
F
igur
e
6
s
hows
that
by
incr
e
a
s
ing
the
a
ntenna
s
ys
tem,
ther
e
is
r
e
duc
ti
on
in
e
r
r
or
r
a
tes
whic
h
he
lps
to
im
pr
ove
the
a
va
il
a
bil
it
y
of
the
c
h
a
nne
l:
t
he
s
ys
tem
us
ing
the
4x4
M
I
M
O
pe
r
f
or
ms
be
s
t
c
ompar
e
d
to
the
s
ys
tems
u
s
ing
1x4
S
I
M
O,
4x1
M
I
S
O
a
nd
1x
1
S
I
S
O,
the
1x1
S
I
S
O
s
ys
tem
pe
r
f
o
r
ms
wor
s
t.
I
t
is
int
e
r
e
s
ti
ngly
noti
c
e
d
that
the
1x4
S
I
M
O
s
ys
tem
pe
r
f
or
ms
be
tt
e
r
than
the
4x1
M
I
S
O
s
ys
tem.
T
hi
s
mea
n
s
in
a
cas
e
whe
r
e
a
c
hoice
of
de
ployi
ng
mul
ti
ple
a
ntenna
a
t
th
e
input
or
output
is
to
be
made
,
it
is
be
tt
e
r
to
de
ploy
the
mul
ti
ple
a
ntenna
a
t
the
output
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
1693
-
6930
T
E
L
KO
M
NI
KA
T
e
lec
omm
un
C
omput
E
l
C
ontr
o
l
,
Vol.
18
,
No
.
1
,
F
e
br
ua
r
y
2020
:
4
19
-
4
26
424
-
2
0
-
1
0
0
10
20
30
40
0
5
10
15
20
25
30
35
40
45
A
c
h
i
e
v
a
b
l
e
r
a
t
e
o
f
M
I
M
O
r
e
c
e
i
v
e
r
s
f
o
r
i
.
i
.
d
.
f
a
s
t
f
a
d
i
n
g
c
h
a
n
n
e
l
A
v
e
r
a
g
e
S
N
R
(
d
B
)
A
c
h
i
e
v
a
b
l
e
r
a
t
e
(
b
i
t
s
/
t
r
a
n
s
m
i
s
s
i
o
n
)
M
a
t
c
h
e
d
F
i
l
t
e
r
Z
e
r
o
-
F
o
r
c
i
n
g
M
M
S
E
S
I
C
F
igur
e
2
.
Ac
hieva
ble
r
a
te
o
f
M
I
M
O
r
e
c
e
iver
s
f
o
r
i
.
i.
d
f
a
s
t
f
a
ding
c
ha
nne
l
0
5
10
15
20
25
30
10
-7
10
-6
10
-5
10
-4
10
-3
10
-2
10
-1
10
0
B
i
t
e
r
r
o
r
r
a
t
e
f
o
r
a
Q
P
S
K
i
n
p
u
t
-
i
.
i
.
d
.
R
a
y
l
e
i
g
h
s
l
o
w
f
a
d
i
n
g
c
h
a
n
n
e
l
A
v
e
r
a
g
e
S
N
R
(
d
B
)
BER
S
I
S
O
Q
u
a
s
i
-
O
r
t
h
o
g
o
n
a
l
S
T
B
C
O
r
t
h
o
g
o
n
a
l
S
T
B
C
F
igur
e
3
.
B
it
e
r
r
o
r
r
a
te
f
or
a
QPS
K
input
-
i.
i.
d
r
a
yle
igh
s
low
f
a
ding
c
ha
nne
l
-
1
0
-5
0
5
10
15
10
-5
10
-4
10
-3
10
-2
10
-1
10
0
O
u
t
a
g
e
p
r
o
b
a
b
i
l
i
t
y
f
o
r
1
s
t
r
e
a
m
-
4
x
1
M
I
S
O
i
.
i
.
d
.
R
a
y
l
e
i
g
h
s
l
o
w
f
a
d
i
n
g
c
h
a
n
n
e
l
A
v
e
r
a
g
e
S
N
R
(
d
B
)
O
u
t
a
g
e
p
r
o
b
a
b
i
l
i
t
y
O
S
T
B
C
Q
O
S
T
B
C
,
M
M
S
E
A
l
a
m
o
u
t
i
S
T
B
C
S
I
S
O
M
I
S
O
F
igur
e
4
.
Outa
ge
pr
oba
bil
it
y
f
or
1
s
tr
e
a
m
-
4
x
1
M
I
S
O
i.
i
.
d
r
a
yleigh
s
low
f
a
ding
c
ha
nne
l
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KO
M
NI
KA
T
e
lec
omm
un
C
omput
E
l
C
ontr
o
l
M
I
M
O
c
hanne
ls
:
opti
miz
ing
thr
oughput
and
r
e
duc
ing
outage
.
.
.
.
(
Oboy
e
r
ulu
A
gbo
je)
425
0
5
10
15
20
25
30
35
40
45
50
10
-6
10
-5
10
-4
10
-3
10
-2
10
-1
10
0
A
v
e
r
a
g
e
S
N
R
(
d
B
)
C
a
p
a
c
i
t
y
w
i
t
h
o
u
t
a
g
e
(
b
i
t
s
/
t
r
a
n
s
m
i
s
s
i
o
n
)
C
a
p
a
c
i
t
y
w
i
t
h
1
0
%
o
u
t
a
g
e
f
o
r
i
.
i
.
d
.
R
a
y
l
e
i
g
h
s
l
o
w
f
a
d
i
n
g
c
h
a
n
n
e
l
4
x
4
M
I
M
O
1
x
4
S
I
M
O
4
x
1
M
I
S
O
1
x
1
S
I
S
O
F
igur
e
5
.
C
a
pa
c
it
y
with
10
%
outage
f
or
i.
i
.
d
r
a
ylei
gh
f
a
ding
c
ha
nne
l
0
5
10
15
0
0
.
1
0
.
2
0
.
3
0
.
4
0
.
5
0
.
6
0
.
7
0
.
8
0
.
9
1
S
e
l
e
c
t
e
d
R
a
t
e
R
P
r
o
b
a
b
i
l
i
t
y
C
u
m
u
l
a
t
i
v
e
D
e
n
s
i
t
y
F
u
n
c
t
i
o
n
a
t
5
d
B
4
x
4
M
I
M
O
2
x
2
M
I
M
O
S
I
S
O
F
igur
e
6
.
C
umul
a
ti
v
e
de
ns
it
y
f
unc
t
ion
a
t
5dB
4.
CONC
L
USI
ON
S
im
ulations
we
r
e
c
a
r
r
ied
out
in
M
a
tl
a
b
to
de
mo
ns
tr
a
te
dif
f
e
r
e
nt
s
c
e
na
r
ios
o
f
M
I
M
O
a
nd
a
ls
o
to
s
how
how
the
pe
r
f
or
manc
e
im
p
r
ove
s
with
incr
e
a
s
i
ng
mul
ti
plexing
ga
ins
.
T
he
s
im
ulation
r
e
s
ult
s
a
r
e
s
hown
in
the
gr
a
phs
of
F
i
gu
r
e
2
t
o
F
igur
e
6
.
T
he
s
im
ulatio
ns
s
how
how
outage
is
r
e
duc
e
d
by
mul
ti
plexing.
F
igur
e
3
s
hows
how
the
pe
r
f
o
r
manc
e
im
pr
ove
s
with
incr
e
a
s
ing
mul
ti
plexing
ga
ins
,
whe
r
e
the
r
e
c
e
ivi
ng
e
nd
us
e
s
S
I
S
O,
Qua
s
i
-
Or
thogonal,
S
T
B
C
a
nd
Or
thogonal
S
T
B
;
the
B
E
R
is
be
s
t
f
o
r
the
r
e
c
e
ivi
ng
s
ys
tem
us
ing
Or
thogonal
S
T
B
C
a
nd
wor
s
t
f
o
r
the
S
I
S
O
.
T
he
s
i
mul
a
ti
ons
in
F
igu
r
e
5
s
hows
that
by
incr
e
a
s
ing
th
e
a
ntenna
s
ys
tem,
ther
e
is
r
e
duc
ti
on
in
e
r
r
o
r
r
a
tes
whic
h
he
lps
to
im
pr
ove
the
a
va
il
a
bil
it
y
o
f
the
c
ha
nne
l:
th
e
s
ys
tem
us
ing
the
4x4
M
I
M
O
pe
r
f
or
ms
be
s
t
c
ompar
e
d
to
the
s
ys
tems
u
s
ing
1x4
S
I
M
O,
4x1
M
I
S
O
a
nd
1x1
S
I
S
O.
T
he
1x1
S
I
S
O
s
ys
tem
pe
r
f
or
ms
wo
r
s
t.
I
t
is
int
e
r
e
s
ti
ngly
noti
c
e
d
that
the
1x4
S
I
M
O
s
ys
tem
pe
r
f
o
r
ms
be
tt
e
r
than
the
4x1
M
I
S
O
s
ys
tem.
T
his
mea
ns
in
a
c
a
s
e
whe
r
e
a
c
hoice
of
de
ployi
ng
mul
ti
pl
e
a
n
tenna
a
t
th
e
input
or
output
is
to
be
made
,
it
is
be
tt
e
r
to
de
ploy
the
mut
i
ple
a
ntenna
a
t
the
output
.
RE
F
E
RE
NC
E
S
[1
]
J
.
R.
H
amp
t
o
n
,
“
In
t
ro
d
u
c
t
i
o
n
t
o
MIMO
Co
mm
u
n
i
cat
i
o
n
s
,
”
Camb
ri
d
g
e
U
n
i
v
ers
i
t
y
Pres
s
,
2
0
1
3
.
[2
]
K
ei
t
h
l
y
,
“A
d
v
a
n
ced
Mea
s
u
reme
n
t
T
e
ch
n
i
q
u
e
s
fo
r
O
F
D
M
a
n
d
MIMO
-
b
as
e
d
Rad
i
o
S
y
s
t
ems
:
D
em
y
s
t
i
f
y
i
n
g
W
L
A
N
an
d
W
i
MA
X
T
e
s
t
i
n
g
,
”
1
st
ed
i
t
i
o
n
.
[3
]
M
.
Sh
arma
,
“E
ffect
i
v
e
ch
a
n
n
e
l
s
t
a
t
e
i
n
fo
rma
t
i
o
n
(CS
I)
feed
b
ack
f
o
r
MIMO
Sy
s
t
em
s
i
n
W
i
re
l
es
s
Bro
a
d
b
a
n
d
co
mmu
n
i
ca
t
i
o
n
,
”
M.
E
n
g
T
h
es
i
s
t
o
Sch
o
o
l
o
f
E
l
ec
t
ri
cal
E
n
g
i
n
eer
i
n
g
an
d
Co
m
p
u
t
er
Sci
e
n
ce
an
d
E
n
g
i
n
eer
i
n
g
Facu
l
t
y
,
Q
u
een
s
l
a
n
d
U
n
i
v
ers
i
t
y
o
f
T
ec
h
n
o
l
o
g
y
,
2
0
1
4
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
1693
-
6930
T
E
L
KO
M
NI
KA
T
e
lec
omm
un
C
omput
E
l
C
ontr
o
l
,
Vol.
18
,
No
.
1
,
F
e
br
ua
r
y
2020
:
4
19
-
4
26
426
[4
]
L
.
Z
h
en
g
,
“
D
i
v
ers
i
t
y
-
Mu
l
t
i
p
l
ex
i
n
g
T
rad
e
o
ff:
A
Co
mp
reh
en
s
i
v
e
V
i
e
w
o
f
Mu
l
t
i
p
l
e
A
n
t
en
n
a
Sy
s
t
ems
,
”
G
rad
u
at
e
D
i
v
i
s
i
o
n
o
f
t
h
e
U
n
i
v
ers
i
t
y
o
f
Cal
i
fo
r
n
i
a
a
t
Berk
el
e
y
,
2
0
0
2
.
[5
]
R
.
Tia
n
,
e
t
al
.
,
“
Ch
aract
eri
za
t
i
o
n
o
f
MIMO
A
n
t
e
n
n
a
s
w
i
t
h
Mu
l
t
i
p
l
ex
i
n
g
E
ffi
c
i
en
c
y
,”
E
l
ec
t
ro
ma
g
n
e
t
i
c
T
h
e
o
ry
D
ep
ar
t
men
t
o
f
E
l
ect
r
i
cal
a
n
d
In
f
o
rmat
i
o
n
T
ec
h
n
o
l
o
g
y
L
u
n
d
U
n
i
v
er
s
i
t
y
Sw
ed
e
n
,
2
0
1
0
.
[6
]
A
k
p
ai
d
a
V
O
A
,
et
al
.
,
“
D
et
ermi
n
at
i
o
n
o
f
an
o
u
t
d
o
o
r
p
at
h
l
o
s
s
mo
d
el
a
n
d
s
i
g
n
a
l
p
en
et
ra
t
i
o
n
l
e
v
el
i
n
s
o
me
s
e
l
ec
t
ed
mo
d
er
n
res
i
d
en
t
i
a
l
an
d
o
ff
i
ce
ap
ar
t
men
t
s
i
n
O
g
b
o
m
o
s
h
o
,
O
y
o
St
at
e,
N
i
g
er
i
a
,
”
Jo
u
r
n
a
l
o
f
E
n
g
i
n
eer
i
n
g
R
e
s
e
a
r
c
h
a
n
d
R
ep
o
r
t
s
,
v
o
l
.
2
,
n
o
.
1
,
p
p
.
1
-
2
5
,
2
0
1
8
.
[7
]
F
.
A.
T.
B
.
N
.
Mo
n
t
ei
r
o
,
“
L
at
t
i
ces
i
n
MIMO
S
p
at
i
a
l
Mu
l
t
i
p
l
e
x
i
n
g
:
D
e
t
ect
i
o
n
an
d
G
eo
met
r
y
,
”
U
n
i
v
er
s
i
t
y
o
f
Camb
ri
d
g
e,
2
0
1
2
.
[8
]
J.
G
.
Pro
ak
i
s
,
“
D
i
g
i
t
a
l
Co
mmu
n
i
ca
t
i
o
n
s
,
”
McG
raw
H
i
l
l
4
th
E
d
i
t
i
o
n
.
[9
]
Jr
.
G
.
D
.
Fo
rn
ey
an
d
G
.
U
n
g
erb
o
eck
,
“
Mo
d
u
l
at
i
o
n
a
n
d
co
d
i
n
g
fo
r
l
i
n
ear
g
au
s
s
i
an
c
h
an
n
el
s
,”
IE
E
E
Tr
a
n
s
.
In
f
o
.
Th
eo
r
y
,
v
o
l
.
4
4
,
n
o
.
6
,
p
p
.
2
3
8
4
-
2
4
1
5
,
1
9
9
8
.
[1
0
]
V
.
Na
g
p
al
,
“
Co
o
p
era
t
i
v
e
mu
l
t
i
p
l
e
x
i
n
g
i
n
W
i
rel
es
s
Re
l
ay
N
et
w
o
rk
,
”
U
n
i
v
er
s
i
t
y
o
f
Cal
i
f
o
rn
i
a
,
2
0
1
2
.
[1
1
]
L
.
G
.
O
rd
o
n
ez,
et
al
.
,
“O
n
t
h
e
D
i
v
ers
i
t
y
,
Mu
l
t
i
p
l
ex
i
n
g
an
d
A
rra
y
G
ai
n
T
rad
e
o
ff
i
n
MIMO
Ch
an
n
el
s
,
”
2
0
1
0
I
E
E
E
In
t
e
r
n
a
t
i
o
n
a
l
S
y
m
p
o
s
i
u
m
o
n
In
f
o
r
m
a
t
i
o
n
Th
e
o
r
y
,
2
0
1
0
.
[1
2
]
H
.
N
i
s
h
i
mo
t
o
,
“St
u
d
i
es
o
n
MIMO
s
p
a
t
i
a
l
mu
l
t
i
p
l
ex
i
n
g
fo
r
h
i
g
h
-
s
p
ee
d
co
mm
u
n
i
cat
i
o
n
,
”
Ph
D
d
i
s
s
er
t
at
i
o
n
t
o
H
o
k
k
a
i
d
o
U
n
i
v
er
s
i
t
y
,
2
0
0
7
.
[1
3
]
C
.
U
.
N
d
u
j
i
u
b
a,
e
t
al
.
,
“
MIMO
D
ef
i
ci
e
n
c
i
es
D
u
e
t
o
A
n
t
e
n
n
a
Co
u
p
l
i
n
g
,”
I
n
t
e
r
n
a
t
i
o
n
a
l
Jo
u
r
n
a
l
o
f
Net
wo
r
ks
a
nd
Co
m
m
u
n
i
c
a
t
i
o
n
s
,
v
o
l
.
5
,
n
o
.
1
,
p
p
.
10
-
17
,
2
0
1
5
.
[1
4
]
C
.
N
d
u
u
j
i
b
a,
et
al
.
,
“
Ch
arac
t
eri
za
t
i
o
n
o
f
M
u
l
t
i
p
l
e
I
n
p
u
t
Mu
l
t
i
p
l
e
O
u
t
p
u
t
:
In
v
es
t
i
g
at
i
n
g
Sp
at
i
al
D
i
v
ers
i
t
y
,”
In
t
e
r
n
a
t
i
o
n
a
l
Jo
u
r
n
a
l
o
f
A
p
p
l
i
e
d
E
n
g
i
n
ee
r
i
n
g
R
e
s
ea
r
ch
,
v
ol
.
1
1
,
n
o
.
2
,
pp
.
7
6
6
2
-
76
65
,
2
0
1
6
.
[1
5
]
O
k
o
k
p
u
j
i
e
K
.
O
.
,
et
al
.
,
“Perfo
rman
ce
A
n
al
y
s
i
s
an
d
Mo
d
el
i
n
g
o
f
Mi
mo
Sy
s
t
em
s
,
”
In
t
e
r
n
a
t
i
o
n
a
l
Jo
u
r
n
a
l
o
f
A
p
p
l
i
e
d
E
n
g
i
n
eer
i
n
g
R
es
e
a
r
c
h
,
v
o
l
.
1
1
,
n
o
.
2
3
,
p
p
.
1
1
5
3
7
-
1
1
5
4
1
,
2
0
1
6
.
[1
6
]
A
.
F.
Mo
rab
i
t
o
,
et
al
.
,
“Mas
k
-
co
n
s
t
rai
n
ed
p
o
w
er
s
y
n
t
h
e
s
i
s
o
f
max
i
mal
l
y
s
p
ars
e
l
i
n
ear
a
rray
s
t
h
r
o
u
g
h
a
co
mp
re
s
s
i
v
e
-
s
en
s
i
n
g
-
d
ri
v
en
s
t
rat
e
g
y
,
”
Jo
u
r
n
a
l
o
f
E
l
e
ct
r
o
m
a
g
n
et
i
c
W
a
ve
s
a
n
d
A
p
p
l
i
c
a
t
i
o
n
s
,
v
o
l
.
2
9
,
n
o
.
1
0
,
p
p
.
1
3
8
4
-
1
3
9
6
,
2
0
1
5
.
[1
7
]
A
.
F.
Mo
rab
i
t
o
,
et
al
.
,
“Is
o
p
h
o
ri
c
array
an
t
en
n
as
w
i
t
h
a
l
o
w
n
u
m
b
er
o
f
co
n
t
r
o
l
p
o
i
n
t
s
:
a
's
i
ze
t
ap
er
ed
'
s
o
l
u
t
i
o
n
,
”
P
r
o
g
r
es
s
i
n
E
l
ec
t
r
o
m
a
g
n
et
i
cs
R
e
s
ea
r
ch
Le
t
t
e
r
s
, v
o
l
.
3
6
,
p
p
.
1
2
1
-
1
3
1
,
2
0
1
3
.
[1
8
]
A
.
Fran
ci
s
,
et
al
.
,
“D
es
i
g
n
an
d
A
n
al
y
s
i
s
o
f
a
Bro
ad
ca
s
t
N
et
w
o
r
k
U
s
i
n
g
L
o
g
i
ca
l
Seg
men
t
at
i
o
n
,
”
TE
LKO
M
NI
KA
Tel
eco
m
m
u
n
i
ca
t
i
o
n
Co
m
p
u
t
i
n
g
E
l
ect
r
o
n
i
c
s
a
n
d
Co
n
t
r
o
l
,
v
o
l
.
1
6
,
n
o
.
2
,
p
p
.
8
0
3
-
8
1
0
,
A
p
r
2
0
1
8
.
[1
9
]
U
.
St
an
l
ey
,
et
al
.
,
“
E
x
p
er
i
men
t
al
A
n
al
y
s
i
s
o
f
Cab
l
e
D
i
s
t
an
ce
E
ffect
o
n
Si
g
n
a
l
A
t
t
e
n
u
a
t
i
o
n
i
n
Si
n
g
l
e
an
d
M
u
l
t
i
m
o
d
e
Fi
b
er
O
p
t
i
c
s
,
”
In
t
er
n
a
t
i
o
n
a
l
J
o
u
r
n
a
l
o
f
E
l
ec
t
r
i
ca
l
a
n
d
Co
m
p
u
t
e
r
E
n
g
i
n
ee
r
i
n
g
,
v
o
l
.
8
,
n
o
.
3
,
p
p
.
1
5
7
7
-
1
5
8
2
,
J
u
n
2
0
1
8
.
[2
0
]
F.
I.
A
n
y
as
i
an
d
U
zai
r
u
e
S.
I.
,
“D
et
ermi
n
a
t
i
o
n
o
f
G
SM
Si
g
n
al
St
re
n
g
t
h
i
n
So
me
Sel
ec
t
ed
L
o
cat
i
o
n
i
n
E
K
P
O
MA
:
IO
SR,
”
Jo
u
r
n
a
l
o
f
E
l
ect
r
o
n
i
c
s
a
n
d
C
o
m
m
u
n
i
c
a
t
i
o
n
E
n
g
i
n
eer
i
n
g
(IO
S
R
-
J
E
CE
)
,
v
o
l
.
9
,
n
o
.
2
,
p
p
.
0
8
-
1
5
,
Mar
-
A
p
r
2
0
1
4
.
[2
1
]
A
.
Pas
s
i
a
n
d
D
.
Bat
ra,
“F
u
t
u
re
o
f
i
n
t
ern
e
t
o
f
t
h
i
n
g
s
(Io
T
)
i
n
5
G
w
i
rel
e
s
s
n
e
t
w
o
rk
s
,
”
In
t
er
n
a
t
i
o
n
a
l
J
o
u
r
n
a
l
o
f
E
n
g
i
n
eer
i
n
g
&
Tech
n
o
l
o
g
y
,
v
o
l
.
7
,
p
p
.
2
4
5
-
2
4
8
,
2
0
1
8
.
[2
2
]
A
.
G
u
p
t
a
an
d
R.
K
.
J
h
a,
“A
Su
rv
ey
o
f
5
G
N
et
w
o
r
k
:
A
rc
h
i
t
ect
u
re
an
d
E
merg
i
n
g
T
ech
n
o
l
o
g
i
e
s
,
”
IE
E
E
A
cce
s
s
,
v
o
l
3
,
p
p
.
1
2
0
6
-
1
2
3
2
,
2
0
1
5
.
[2
3
]
T.
Pad
map
ri
y
a
an
d
V
.
Sami
n
a
d
an
,
“Imp
r
o
v
i
n
g
T
h
ro
u
g
h
p
u
t
fo
r
D
o
w
n
l
i
n
k
Mu
l
t
i
u
s
er
MIM
O
-
L
T
E
A
d
v
a
n
ced
N
et
w
o
r
k
s
u
s
i
n
g
SIN
R
ap
p
r
o
x
i
mat
i
o
n
an
d
H
i
erarch
i
cal
CSI
feed
b
ack
,
”
In
t
er
n
a
t
i
o
n
a
l
Jo
u
r
n
a
l
o
f
M
o
b
i
l
e
D
e
s
i
g
n
Net
wo
r
k
a
n
d
In
n
o
v
a
t
i
o
n
-
In
d
e
r
s
c
i
en
ce
P
u
b
l
i
s
h
er
,
v
o
l
.
6
,
p
p
.
1
4
-
2
3
,
May
2
0
1
5
.
[2
4
]
T
.
S.
Rap
p
ap
o
rt
,
“W
i
rel
e
s
s
Co
mm
u
n
i
cat
i
o
n
s
Pr
i
n
c
i
p
l
es
an
d
Pract
i
ce,
”
Pren
t
i
ce
H
a
l
l
,
2
0
0
2
.
[2
5
]
A
.
G
o
l
d
s
m
i
t
h
s
,
“
W
i
re
l
es
s
Co
mmu
n
i
ca
t
i
o
n
s
,
”
Camb
ri
d
g
e
U
n
i
v
ers
i
t
y
Pres
s
,
v
o
l
.
7
,
p
p
.
4
3
-
6
7
,
2
0
0
5
.
Evaluation Warning : The document was created with Spire.PDF for Python.