T
E
L
KO
M
N
I
KA
T
e
lec
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.
90
~
98
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.
13674
90
Jou
r
n
al
h
omepage
:
ht
tp:
//
jour
nal.
uad
.
ac
.
id/
index
.
php/T
E
L
K
OM
N
I
K
A
In
t
e
g
r
at
in
g
m
il
li
m
e
t
e
r
w
av
e
w
ith
h
y
b
r
i
d
p
r
e
c
od
in
g
m
u
ltiu
s
e
r
m
assi
ve
M
IM
O fo
r
5G c
om
m
u
n
ic
a
t
io
n
M
oh
am
m
e
d
Kh
u
d
h
u
r
Hu
s
s
e
in
,
Nas
s
e
r
N.
Kh
a
m
is
s
Co
l
l
eg
e
o
f
I
n
fo
rma
t
i
o
n
E
n
g
i
n
eer
i
n
g
,
A
l
-
N
ah
r
ai
n
U
n
i
v
er
s
i
t
y
,
Bag
h
d
a
d
,
Iraq
Ar
t
icle
I
n
f
o
AB
S
T
RA
CT
A
r
ti
c
le
h
is
tor
y
:
R
e
c
e
ived
J
ul
22
,
2019
R
e
vis
e
d
Aug
21
,
2019
Ac
c
e
pted
S
e
p
15
,
2019
N
o
w
ad
a
y
s
,
t
h
ere
h
a
s
b
een
g
ro
w
i
n
g
i
n
t
eres
t
i
n
t
h
e
Mas
s
i
v
e
MIMO
a
s
a
res
u
l
t
o
f
i
mp
r
o
v
i
n
g
t
h
r
o
u
g
h
p
u
t
b
y
l
e
v
erag
i
n
g
s
p
a
t
i
a
l
freed
o
m
an
d
array
g
ai
n
.
It
i
s
eq
u
i
p
p
ed
w
i
t
h
mi
l
l
i
met
er
w
av
e
(mm
W
a
v
e)
b
a
n
d
s
t
o
res
o
l
v
e
t
h
e
h
i
g
h
p
at
h
-
l
o
s
s
.
It
i
s
k
n
o
w
n
fro
m
t
h
e
l
i
t
erat
u
re
t
h
a
t
i
t
era
t
ed
a
l
g
o
ri
t
h
ms
are
u
s
u
al
l
y
u
s
e
d
t
o
at
t
a
i
n
t
h
e
h
y
b
r
i
d
p
rec
o
d
er
s
t
o
acco
mp
l
i
s
h
a
s
p
ec
i
fi
c
o
p
t
i
m
i
zat
i
o
n
o
b
j
ect
i
v
e.
T
h
u
s
,
t
h
e
c
o
mp
l
ex
i
t
y
rema
i
n
s
h
i
g
h
b
ecau
s
e
each
i
t
erat
i
o
n
may
i
n
c
l
u
d
e
s
i
n
g
u
l
ar
v
a
l
u
e
d
eco
m
p
o
s
i
t
i
o
n
,
t
h
e
ma
t
ri
x
i
n
v
e
rs
i
o
n
,
an
d
s
o
o
n
t
h
a
t
mo
t
i
v
a
t
es
u
s
t
o
s
p
l
i
t
t
h
e
h
y
b
r
i
d
p
reco
d
i
n
g
an
d
co
m
b
i
n
i
n
g
p
r
o
b
l
em
i
n
t
o
sub
-
p
r
o
b
l
ems
.
T
h
e
p
ro
p
o
s
ed
s
o
l
u
t
i
o
n
i
s
a
m
u
l
t
i
-
u
s
er
Mas
s
i
v
e
MIM
O
h
y
b
ri
d
b
eamfo
rm
i
n
g
b
as
ed
o
n
a
co
n
v
ex
o
p
t
i
mi
za
t
i
o
n
p
r
o
b
l
e
m
t
h
at
i
s
ap
p
l
i
e
d
an
d
s
o
l
v
e
d
fo
r
es
t
i
mat
i
n
g
t
h
e
d
i
g
i
t
al
p
r
e
co
d
i
n
g
t
o
el
i
m
i
n
a
t
e
i
n
t
er
-
u
s
er
i
n
t
erfere
n
ce
w
h
i
l
e
u
s
i
n
g
co
d
e
b
o
o
k
s
t
o
s
e
l
ect
an
a
l
o
g
b
eamfo
rmer
s
.
It
i
s
ap
p
are
n
t
i
n
t
h
e
maj
o
ri
t
y
o
f
cas
e
s
;
t
h
e
p
ro
p
o
s
ed
b
eamfo
rm
i
n
g
p
erfo
r
man
ce
i
s
h
i
g
h
er
t
h
a
n
o
n
l
y
-
a
n
al
o
g
b
eamfo
rm
i
n
g
,
s
i
n
g
l
e
-
u
s
er
(n
o
i
n
t
erfere
n
ce
),
t
h
e
Z
F
p
r
eco
d
i
n
g
,
t
h
e
MMSE
p
reco
d
i
n
g
,
an
d
t
h
e
K
a
l
man
p
rec
o
d
i
n
g
w
h
ere
t
h
e
fu
l
l
d
i
g
i
t
al
s
o
l
u
t
i
o
n
i
s
a
co
n
s
i
d
erab
l
e
as
t
h
e
b
en
c
h
mark
p
o
i
n
t
w
i
t
h
d
i
fferen
t
s
cen
ar
i
o
s
d
u
e
t
o
t
h
e
r
ed
u
ct
i
o
n
o
f
u
s
er
i
n
t
erfere
n
ce.
T
h
u
s
,
t
h
ere
i
s
n
o
c
o
n
s
i
d
erat
i
o
n
f
o
r
co
mp
l
i
ca
t
ed
o
p
era
t
i
o
n
s
s
u
c
h
as
S
V
D
o
r
i
n
v
ers
i
o
n
mat
r
i
c
es
as
w
el
l
a
s
n
o
n
e
ed
fo
r
d
at
a
e
s
t
i
mat
i
o
n
.
O
u
r
p
ro
p
o
s
ed
s
o
l
u
t
i
o
n
can
s
erv
e
a
l
arg
e
n
u
m
b
er
o
f
u
s
er
s
s
i
mu
l
t
an
e
o
u
s
l
y
d
u
e
t
o
mo
re
d
i
rec
t
i
v
e
g
ai
n
b
y
u
s
i
n
g
n
u
mero
u
s
a
n
t
e
n
n
a
s
at
BS.
Bas
ed
o
n
i
t
s
l
es
s
co
m
p
l
e
x
i
t
y
a
n
d
k
ee
p
i
n
g
p
erf
o
rma
n
ce
,
o
u
r
s
o
l
u
t
i
o
n
can
b
e
reco
mmen
d
ed
.
K
e
y
w
o
r
d
s
:
C
onve
x
opti
mi
z
a
ti
on
Hybr
id
pr
e
c
oding
M
il
li
mete
r
wa
ve
(
mm
W
a
ve
)
M
ult
i
-
us
e
r
mas
s
ive
M
I
M
O
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
:
M
oha
mm
e
d
Khudhur
Hus
s
e
in
,
C
oll
e
ge
of
I
n
f
or
mation
E
nginee
r
ing
,
Al
-
Na
hr
a
in
Unive
r
s
it
y,
B
a
ghda
d,
I
r
a
q
.
E
mail:
mohammed
.
khudhur
@gmail
.
c
om
1.
I
NT
RODU
C
T
I
ON
I
n
r
e
c
e
nt
ye
a
r
s
,
r
e
s
e
a
r
c
he
r
s
ha
ve
s
hown
a
r
e
ne
we
d
int
e
r
e
s
t
in
mi
ll
im
e
ter
wa
ve
(
mm
W
a
ve
)
ba
nds
f
or
f
utur
e
c
e
ll
ular
s
ys
tems
[
1
-
3]
.
T
he
s
hor
tage
of
s
ub
-
6
GH
z
s
pe
c
tr
um
r
e
s
our
c
e
s
mea
n
s
that
c
onve
nti
ona
l
c
e
ll
ular
s
ys
tems
s
uf
f
e
r
f
r
om
a
hos
t
o
f
pit
f
a
ll
s
s
uc
h
a
s
the
r
a
pid
gr
owth
in
mobi
le
inf
o
r
mation
tr
a
f
f
ic,
low
late
nc
y,
e
nor
mous
c
onne
c
ti
vit
y,
a
nd
low
e
ne
r
gy
c
ons
umpt
ion
in
2020
a
nd
be
yond.
T
he
r
a
nge
be
twe
e
n
3
a
nd
30
0
GH
z
e
na
bles
the
a
c
c
e
s
s
to
wir
e
les
s
tr
a
ns
mi
s
s
ion
s
with
lar
ge
unde
r
us
e
d
ba
ndwidths
a
nd
make
s
it
e
a
s
ier
to
a
pply
c
ompac
t
lar
ge
a
ntenna
a
r
r
a
ys
due
to
it
s
s
hor
t
wa
ve
length
[
4
-
6]
,
a
s
de
tailed
in
F
igur
e
1
.
M
a
s
s
ive
M
ult
ipl
e
-
I
nput
M
ult
ipl
e
-
Output
(
M
a
s
s
ive
M
I
M
O)
i
s
a
r
e
li
a
ble
tec
hnique,
whic
h
im
pr
ove
s
thr
oughput
by
leve
r
a
ging
s
pa
ti
a
l
f
r
e
e
dom
a
nd
a
r
r
a
y
ga
in
[
7]
.
T
o
e
na
ble
mul
ti
-
gigabit
da
t
a
r
a
tes
,
the
int
e
gr
a
ti
on
of
mm
W
a
ve
ba
nds
with
mul
ti
-
us
e
r
M
a
s
s
ive
M
I
M
O
(
M
U
-
M
a
s
s
ive
M
I
M
O)
s
ys
tems
a
r
e
a
vit
a
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
I
ntegr
ati
ng
mill
ime
ter
w
av
e
w
it
h
hy
br
id
pr
e
c
oding
multi
us
e
r
mas
s
ive
…
(
M
ohamm
e
d
K
hudhur
Hus
s
e
in
)
91
a
s
pe
c
t
[
8]
.
M
or
e
ove
r
,
the
lar
ge
-
s
c
a
le
a
ntenna
a
r
r
a
ys
f
or
e
a
c
h
B
a
s
e
S
tation
(
B
S
)
a
nd
M
obil
e
S
tation
(
M
S
s
)
,
a
s
we
ll
a
s
the
pr
e
c
oding
(
be
a
mf
or
m
ing)
,
whic
h
c
ont
r
ibut
e
to
the
e
li
mi
na
ti
on
of
us
e
r
’
s
int
e
r
f
e
r
e
nc
e
a
nd
a
c
hieve
va
r
ious
be
ne
f
it
s
s
uc
h
a
s
c
a
nc
e
li
ng
out
nois
e
a
nd
f
a
s
t
f
a
di
ng
thr
ough
highl
y
di
r
e
c
ti
ona
l
be
a
m
f
or
mi
n
g
[
9
,
10]
.
On
the
other
ha
nd
,
s
mall
-
c
e
ll
s
s
uc
h
a
s
mi
c
r
o
-
c
e
ll
s
,
f
e
mt
o
-
c
e
ll
s
,
a
nd
p
ico
-
c
e
ll
s
that
c
a
n
c
ombi
ne
m
m
W
a
ve
a
nd
M
a
s
s
ive
M
I
M
O
to
a
void
s
ignal
a
tt
e
nua
ti
on
a
n
d
a
c
hieve
3D
be
a
mf
or
mi
ng
[
11
]
.
F
igur
e
1.
M
m
W
a
ve
s
pe
c
tr
um
[
6]
F
or
mm
W
a
ve
M
a
s
s
ive
M
I
M
O
s
ys
tems
,
ba
s
e
d
on
a
li
ter
a
tur
e
pe
r
s
pe
c
ti
ve
,
a
f
ull
y
digi
tal
pr
e
c
oding
s
olut
ion
whe
r
e
e
a
c
h
a
ntenna
li
nks
to
a
de
dica
ted
R
F
c
ha
in
that
is
known
a
s
a
n
im
p
r
a
c
ti
c
a
l
s
olut
ion
f
or
h
igh
f
r
e
que
nc
y
be
c
a
us
e
of
high
c
os
ts
a
nd
high
pow
e
r
c
ons
umpt
ion
[
12
,
13]
.
Al
though
a
n
a
na
log
p
r
e
c
oding
s
olut
ion
is
les
s
c
ompl
ica
ted
with
a
pha
s
e
s
hif
t
that
c
ontr
ols
s
ignal
pha
s
e
s
,
the
c
a
pa
c
it
y
c
a
nnot
be
c
on
s
ider
a
bly
im
pr
ove
d.
As
a
r
e
s
ult
,
the
a
na
log
p
r
e
c
ode
r
s
pe
r
f
or
m
les
s
than
the
f
ul
ly
digi
t
a
l
pr
e
c
ode
r
s
[
14
]
.
I
n
th
is
c
ontext,
a
hyb
r
id
be
a
mf
or
mi
ng
s
olut
ion
e
x
ploi
ts
a
na
log
be
a
mf
or
mer
s
in
the
R
F
domain
a
nd
digi
tal
pr
e
c
ode
r
s
in
the
ba
s
e
ba
nd,
whic
h
a
r
e
c
o
ns
ider
e
d
a
s
a
pr
omi
s
ing
s
olut
ion
to
thes
e
c
ha
ll
e
nge
s
a
nd
e
na
ble
us
to
take
the
a
dva
ntage
s
of
both
the
s
olut
i
ons
[
1
5]
.
M
os
t
o
f
the
c
u
r
r
e
nt
r
e
s
e
a
r
c
h
ha
s
tende
d
to
f
oc
us
on
s
ingl
e
-
us
e
r
M
I
M
O
(
S
U
-
M
I
M
O)
s
c
he
me
s
in
t
he
li
ter
a
tur
e
,
a
lt
hough
ther
e
a
r
e
f
e
w
s
tudi
e
s
on
t
he
hybr
id
be
a
mf
or
mi
ng
f
or
mul
ti
-
us
e
r
M
a
s
s
ive
M
I
M
O
(
M
U
-
M
a
s
s
ive
M
I
M
O)
s
c
he
mes
that
c
a
n
im
pr
ov
e
s
ys
tem
c
a
pa
c
it
y
a
nd
s
pe
c
tr
a
l
e
f
f
icie
nc
y
[
11
,
16
-
22]
.
T
he
hybr
id
p
r
e
c
oding
a
lgo
r
it
hms
r
e
quir
e
a
pe
r
f
e
c
t
c
ha
nne
l
s
tate
inf
or
mation
(
C
S
I
)
thr
ough
thei
r
de
s
ign,
a
lt
hough
it
is
dif
f
icult
in
mm
W
a
ve
M
I
M
O
s
ys
tems
due
to
a
c
ha
nne
l
matr
ix
mea
s
ur
e
d
ba
s
e
d
on
the
s
e
lec
ti
on
of
a
na
log
be
a
mf
or
mer
s
a
t
the
ba
s
e
ba
nd
[
23]
.
M
a
ny
a
ppli
c
a
ti
ons
ne
e
d
the
s
pe
c
tr
um
that
r
a
nge
s
be
twe
e
n
3
a
nd
300GH
z
.
T
he
a
uthor
s
in
[
24
]
us
e
d
C
ognit
ive
R
a
dio
Ne
twor
ks
to
a
void
the
lac
k
of
S
p
e
c
tr
um.
I
n
[
25]
,
the
a
uthor
s
pr
opos
e
d
a
mathe
matica
l
m
ode
l
that
r
e
quir
e
s
the
high
c
a
r
r
ier
f
r
e
que
nc
y
f
o
r
us
ing
a
W
ir
e
les
s
Vide
o
to
moni
tor
T
r
a
ns
por
t
I
n
f
r
a
s
tr
uc
tu
r
e
.
I
n
[
26
]
,
the
a
utho
r
s
de
s
igned
a
c
ir
c
uit
to
e
nha
nc
e
ga
in
a
nd
r
e
duc
e
d
powe
r
c
ons
umpt
ion
us
e
d
in
dif
f
e
r
e
nt
wir
e
les
s
s
y
s
tems
.
T
he
a
uthor
s
in
[
27]
o
f
f
e
r
e
d
a
B
la
c
k
S
pots
W
a
r
ning
Applica
ti
on
that
r
e
duc
e
s
c
r
a
s
he
s
a
t
blac
k
s
pots
.
F
or
mm
W
a
ve
M
U
-
M
a
s
s
ive
M
I
M
O
s
ys
tem
s
,
t
he
r
e
a
r
e
s
e
ve
r
a
l
p
r
e
vious
s
tudi
e
s
.
T
he
a
uthor
s
in
[
28
,
29
]
s
ugge
s
ted
the
a
na
log
pr
e
c
oding
s
olut
ions
with
low
-
c
os
t
pha
s
e
s
hif
ter
s
a
s
a
n
a
lt
e
r
na
ti
ve
to
the
f
ull
-
digi
tal
p
r
e
c
oding
s
olut
ion.
How
e
ve
r
,
it
ha
s
li
mi
ted
a
bil
it
y
to
ha
ndle
int
e
r
-
us
e
r
int
e
r
f
e
r
e
nc
e
.
Author
s
in
[
30]
pr
opos
e
d
hybr
id
be
a
mf
or
mi
ng
ba
s
e
d
on
a
Ka
lm
a
n
c
r
it
e
r
ion
to
e
li
mi
na
te
int
e
r
-
us
e
r
int
e
r
f
e
r
e
nc
e
.
I
n
[
31
]
,
the
a
utho
r
s
u
s
e
d
a
z
e
r
o
-
f
or
c
ing
(
Z
F
)
p
r
e
c
oding
s
olut
ion
with
the
pr
opos
e
d
c
ha
nne
l
e
s
ti
mation
a
lgor
it
hm.
Author
s
in
[
32
,
33]
de
ve
loped
hybr
i
d
be
a
mf
or
m
ing
ba
s
e
d
on
a
mi
nim
um
mea
n
s
qua
r
e
e
r
r
o
r
(
M
M
S
E
)
.
I
n
[
34]
,
the
a
uthor
s
de
ve
loped
a
f
e
e
dba
c
k
mec
ha
nis
m
that
would
a
ll
ow
th
e
B
S
to
pr
oduc
e
a
s
ophis
ti
c
a
ted
R
F
pr
e
c
oding
s
tr
uc
tur
e
.
I
n
[
2]
,
the
a
uthor
s
I
nve
s
ti
ga
ted
the
hybr
id
pr
e
c
oding
a
nd
c
ombi
ning
ba
s
e
d
on
the
pe
r
f
e
c
t
knowle
dge
o
f
the
C
S
I
,
while
a
s
ingul
a
r
va
lue
de
c
ompos
it
ion
(
S
VD
)
is
us
e
d
to
a
c
hieve
the
a
na
log
c
ombi
ne
r
o
f
e
a
c
h
us
e
r
while
the
F
r
obe
nius
matr
ix
of
the
mat
r
ix
is
m
ini
mi
z
e
d
to
c
ompl
e
te
the
a
na
log
a
nd
digi
tal
pr
e
c
oding
thr
ough
the
a
lt
e
r
n
a
ti
ng
opti
mi
z
a
ti
on
a
ppr
oa
c
h.
I
n
[
35]
,
the
a
uthor
s
pr
opos
e
d
the
us
e
o
f
the
a
lt
e
r
na
ti
ve
M
M
S
E
-
ba
s
e
d
a
ge
ne
r
a
li
z
e
d
E
igen
-
de
c
ompos
it
ion
(
GE
VD
)
to
a
c
hiev
e
a
na
log
be
a
mf
or
mi
ng,
while
a
Ka
r
us
h
-
Kuhn
-
T
uc
ke
r
(
KK
T
)
is
us
e
d
to
opti
mi
z
e
d
igi
tal
pr
e
c
oding
.
I
n
[
36]
,
the
a
uthor
s
us
e
d
metr
ics
via
the
opti
mi
z
a
ti
on
a
ppr
oa
c
h
to
a
c
hieve
a
na
log
a
nd
digi
tal
p
r
e
c
oding.
F
inally,
the
a
uthor
s
in
[
37]
s
ugge
s
ted
a
n
it
e
r
a
ti
ve
a
lgor
it
hm
us
ing
the
KK
T
-
ba
s
e
d
pe
na
lt
y
dua
l
de
c
ompos
it
ion
tec
hnique.
I
n
c
ompar
is
on,
the
two
a
ppli
c
a
ti
ons
,
the
s
pe
c
if
ica
ti
ons
,
a
nd
ne
e
ds
of
e
a
c
h
a
ppli
c
a
ti
on
f
o
r
thei
r
r
e
gular
a
c
ti
vit
y
s
hould
be
de
ter
mi
ne
d
a
nd
then
c
ompar
e
d
[
38]
.
Va
r
ious
mea
s
ur
e
s
a
nd
tec
hniques
mus
t
be
a
dopted
to
mi
nim
ize
r
e
a
s
ons
a
nd
im
pa
c
ts
to
im
pr
o
ve
c
omm
unica
ti
on
[
39]
.
I
t
is
known
f
r
om
the
li
te
r
a
tur
e
that
it
e
r
a
ted
a
lgor
it
hms
a
r
e
us
ua
ll
y
us
e
d
to
a
tt
a
in
th
e
hybr
id
pr
e
c
ode
r
s
to
a
c
c
ompl
is
h
a
s
pe
c
if
ic
opti
mi
z
a
ti
on
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
:
90
-
98
92
objec
ti
ve
.
T
hus
,
th
e
c
ompl
e
xit
y
r
e
mains
high
be
c
a
us
e
e
a
c
h
it
e
r
a
ti
on
may
include
s
ingul
a
r
va
lue
de
c
ompos
it
ion,
the
matr
ix
inver
s
ion,
a
nd
s
o
on
.
T
he
a
bove
r
e
a
s
ons
mot
ivate
us
to
s
pli
t
the
hybr
id
p
r
e
c
oding
a
nd
c
ombi
ning
p
r
oblem
int
o
s
ub
-
pr
oblems
.
T
he
p
r
opos
e
d
s
olut
ion
invol
ve
s
tw
o
pha
s
e
s
:
f
ir
s
tl
y
,
the
a
na
log
be
a
mf
o
r
mi
ng
a
nd
c
o
mbi
ning
matr
ice
s
a
r
e
de
s
igned
to
obtain
the
maximum
e
ne
r
gy
pr
in
c
ipl
e
f
or
s
ingl
e
-
us
e
r
s
ys
tems
.
S
e
c
ondly,
a
c
onve
x
opti
mi
z
a
ti
on
p
r
oblem
is
a
ppli
e
d
a
nd
s
olved
to
e
s
ti
mate
the
d
igi
tal
pr
e
c
oding,
whic
h
is
us
e
d
to
e
li
mi
na
te
int
e
r
-
us
e
r
int
e
r
f
e
r
e
nc
e
.
T
he
nove
lt
y
of
th
is
wor
k
c
a
n
be
s
umm
a
r
ize
d
a
s
:
-
W
e
f
or
m
the
c
ha
nne
l
matr
ix
ba
s
e
d
on
a
c
oll
e
c
ti
on
of
a
r
r
a
y
r
e
s
pons
e
ve
c
tor
s
with
low
f
e
e
dba
c
k
r
a
te
while
us
ing
c
ode
books
to
s
e
lec
t
a
na
log
be
a
mf
or
mer
s
.
Af
ter
that
,
a
c
onve
x
opt
im
iza
ti
on
pr
ob
lem
is
a
ppl
ied
a
nd
s
olved
to
e
s
ti
mate
the
digi
tal
pr
e
c
oding.
T
hus
,
the
r
e
is
no
c
ons
ider
a
ti
on
f
or
c
ompl
ica
ted
ope
r
a
ti
ons
s
uc
h
a
s
S
VD
or
inve
r
s
ion
matr
ice
s
while
ke
e
ping
pe
r
f
or
m
a
nc
e
.
-
T
he
F
r
obe
nius
nor
m
of
the
matr
ix
include
s
on
l
y
the
a
na
log
pr
e
c
oding,
a
na
log
c
ombi
ning,
a
nd
c
ha
nne
l
matr
ix
while
ther
e
is
no
ne
e
d
f
or
da
ta
e
s
ti
mation.
-
Unde
r
the
s
a
me
c
ondit
ions
,
our
a
na
lyt
ica
l
a
nd
s
i
mul
a
ti
on
f
indi
ngs
s
how
that
pr
opos
e
d
pr
e
c
oding
a
c
hieve
s
be
tt
e
r
s
pe
c
tr
a
l
e
f
f
icie
nc
y
than
other
e
xis
ti
n
g
hybr
ids
s
uc
h
a
s
the
Z
F
pr
e
c
oding
[
27]
,
the
M
M
S
E
pr
e
c
oding
[
21,
28]
,
a
nd
the
Ka
lm
a
n
p
r
e
c
oding
[
26]
.
2.
RE
S
E
AR
CH
M
E
T
HO
D
No
tations
:
I
n
thi
s
s
tudy,
A
a
nd
a
a
r
e
a
matr
i
x
a
nd
a
ve
c
tor
,
while
|
|
,
‖
‖
,
,
a
r
e
it
s
de
ter
mi
na
nt,
F
r
obe
nius
nor
m,
tr
a
ns
pos
e
,
a
nd
He
r
mi
ti
a
n,
r
e
s
pe
c
ti
ve
ly.
I
a
nd
E
[
.
]
de
note
the
identit
y
matr
ix
a
nd
the
e
xpe
c
tation.
I
s
us
e
d
to
indi
c
a
te
(
N×
M
)
c
o
mpl
e
x
matr
ix
.
2.
1
.
S
ys
t
e
m
m
od
e
l
F
or
the
mul
ti
-
us
e
r
mm
W
a
ve
mas
s
ive
M
I
M
O
s
ys
tem,
we
c
ons
ider
the
s
ingl
e
B
S
a
nd
K
us
e
r
s
a
s
il
lus
tr
a
ted
in
F
igu
r
e
2,
whe
r
e
the
B
S
with
N
BS
a
nt
e
nna
s
a
nd
R
F
c
ha
ins
that
tr
a
ns
mi
ts
N
S
da
ta
s
tr
e
a
ms
to
K
us
e
r
s
,
e
a
c
h
with
N
MS
a
ntenna
.
F
igur
e
2.
P
r
opos
e
d
hyb
r
id
p
r
e
c
oding
s
ys
tem
At
the
B
S
s
ide,
the
digi
tal
p
r
e
c
ode
r
mat
r
ix
∈
digi
tally
pr
oc
e
s
s
e
s
NS
da
ta
s
tr
e
a
ms
f
oll
owe
d
by
the
a
na
log
be
a
mf
o
r
mi
ng
mat
r
ix
∈
that
e
xploi
ts
pha
s
e
s
hif
ter
s
to
mi
nim
ize
e
ne
r
gy
c
ons
umpt
ion
a
nd
c
os
ts
,
s
o
that
th
e
B
S
tr
a
ns
mi
ts
the
f
inal
hybr
id
pr
e
c
ode
d
s
ignal
to
K
us
e
r
s
,
that
is
whe
r
e
∈
1
de
notes
the
t
r
a
ns
mi
tt
e
d
ve
c
tor
a
nd
∈
1
de
notes
the
input
ba
s
e
ba
nd
ve
c
tor
.
Ne
xt,
the
r
e
c
e
ived
s
ignal
a
t
us
e
r
k
be
c
omes
;
=
R
F
B
B
1
F
F
S
=
(
1)
=
RF
BB
+
.
(
2)
wh
e
r
e
∈
ℂ
1
is
the
r
e
c
e
ived
ve
c
tor
,
∈
ℂ
is
the
c
ha
nne
l
matr
ix,
a
nd
∈
ℂ
1
is
the
Ga
us
s
ian
nois
e
ve
c
tor
s
a
ti
s
f
ying
2
[]
S
H
k
k
N
E
n
n
I
=
.
At
the
r
e
c
e
i
ve
r
s
ide,
a
n
a
na
log
c
ombi
ne
r
mat
r
ix
RF
∈
ℂ
×
c
ombi
ne
s
the
r
e
c
e
ived
s
ignal
to
e
s
ti
ma
te
the
pr
oc
e
s
s
e
d
da
ta,
given
by;
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
I
ntegr
ati
ng
mill
ime
ter
w
av
e
w
it
h
hy
br
id
pr
e
c
oding
multi
us
e
r
mas
s
ive
…
(
M
ohamm
e
d
K
hudhur
Hus
s
e
in
)
93
̂
=
(
RF
)
RF
BB
+
(
RF
)
.
(
3)
T
he
pa
th
los
s
in
the
mm
W
a
ve
ba
nd
is
r
e
a
li
s
ti
c
.
T
he
mm
-
W
a
ve
M
I
M
O
c
ha
nne
l
model
be
twe
e
n
the
B
S
a
nd
the
K
us
e
r
s
that
ha
s
li
mi
ted
s
c
a
tt
e
r
s
with
N
ra
y
s
c
a
tt
e
r
s
in
c
ontr
a
s
t
to
the
low
-
f
r
e
que
nc
y
c
ha
nne
l,
that
is
;
=
√
+
∑
ℓ
(
,
ℓ
)
(
(
,
ℓ
)
)
ℓ
=
1
(
4)
whe
re
,
MS
l
a
nd
,
BS
l
a
r
e
Angle
s
o
f
Ar
r
ival
(
AoA
)
a
nd
An
gles
of
De
pa
r
tu
r
e
(
AoD
)
r
e
s
pe
c
ti
ve
ly,
a
nd
u
l
i
s
c
ompl
e
x
pa
th
ga
ins
.
T
he
a
r
r
a
y
r
e
s
pons
e
ve
c
tor
f
o
r
li
ne
a
r
a
r
r
a
ys
,
whic
h
take
s
the
f
or
m
o
f
whe
r
e
de
notes
the
c
a
r
r
ier
wa
ve
length
,
a
nd
/2
d
=
that
ind
ica
tes
the
int
e
r
-
a
ntenna
s
pa
c
ing.
(
)
=
1
√
[
1
,
.
.
.
.
.
.
.
.
.
.
,
(
−
1
)
2
(
)
]
(
5)
(
)
=
1
√
[
1
,
.
.
.
.
.
.
.
.
.
.
,
(
−
1
)
2
(
)
]
(
6)
2.
2
.
P
r
ob
lem
f
or
m
u
lat
io
n
I
n
the
be
ginni
ng,
a
n
e
f
f
e
c
ti
ve
f
e
e
dba
c
k
c
ha
nne
l
is
a
f
e
a
s
ibl
e
s
olut
ion
to
tac
kle
huge
tr
a
ini
ng
s
ignal
ove
r
he
a
d.
T
he
n,
the
B
S
c
a
lcula
tes
the
d
igi
tal
p
r
e
c
ode
r
that
c
a
n
e
li
mi
na
te
int
e
r
-
us
e
r
int
e
r
f
e
r
e
nc
e
.
F
inally
,
the
pr
oblem
of
int
e
r
e
s
t
is
to
maximi
z
e
the
a
c
hieva
ble
r
a
te
of
the
s
ys
tem
a
f
ter
c
a
lcula
ti
ng
the
a
na
log
be
a
mf
or
mi
ng,
the
e
f
f
e
c
ti
ve
c
ha
nne
l
a
nd
the
d
igi
tal
pr
e
c
oding,
whic
h
take
s
the
f
or
m
of
;
=
(
RF
)
RF
(
7)
=
∑
2
(
1
+
|
BB
|
2
∑
|
BBu
|
2
+
2
=
1
≠
)
=
1
(
8)
2.
3.
P
r
op
os
e
d
h
yb
r
id
b
e
am
f
or
m
i
n
g
T
he
qua
nti
z
e
d
R
F
pr
e
c
oding
ve
c
tor
s
or
the
a
r
r
a
y
r
e
s
pons
e
ve
c
tor
s
pr
oduc
e
c
ode
wor
ds
(
c
olum
ns
)
a
t
the
AoD
.
T
he
pr
opos
e
d
hybr
id
pr
e
c
oding
e
xploi
ts
c
od
e
books
f
or
s
e
lec
ti
ng
the
a
na
log
be
a
mf
or
mi
ng/combini
ng
ve
c
tor
s
.
On
the
other
h
a
nd,
‖
‖
2
indi
c
a
tes
tr
a
ns
mi
tt
e
d
powe
r
c
ons
tr
a
int
,
that
is
;
,
‖
̂
−
‖
2
su
bjec
t
to
1
{
,
.
.
.
.
.
.
.
,
}
R
F
m
F
f
f
(
9)
2
R
F
B
B
K
F
FF
P
whe
r
e
K
P
de
notes
the
tr
a
ns
mi
tt
e
d
powe
r
.
F
or
e
a
c
h
dir
e
c
ti
on,
the
B
S
pe
r
f
or
ms
tr
a
ini
ng
pa
c
ke
ts
f
oll
owing
by
c
a
lcula
ti
ng
the
r
e
c
e
ived
s
ignal
s
tr
e
ngth
indi
c
a
tor
(
R
S
S
I
)
.
T
he
n,
e
a
c
h
us
e
r
e
s
ti
m
a
tes
the
e
f
f
e
c
ti
ve
c
ha
nne
l
in
a
ll
dir
e
c
ti
on
to
f
e
e
d
the
B
S
.
F
igur
e
3
s
hows
the
t
r
a
ns
mi
s
s
ion
pr
otocol.
T
he
c
onve
x
opti
mi
z
a
ti
on
p
r
oblem
is
a
ppli
e
d
a
nd
s
olved
f
or
a
ppr
oxim
a
ti
ng
pr
oblem
in
(
9)
by
us
ing
opti
mi
z
a
ti
on
tool
s
.
BB
,
2
m
i
n
R
F
B
B
S
N
e
f
f
F
FF
IH
F
E
−
s
ubjec
t
to
1
{
,
.
.
.
.
.
.
.
,
}
R
F
m
F
f
f
(
10)
2
R
F
B
B
K
F
FF
P
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
:
90
-
98
94
F
igur
e
3.
T
r
a
ns
mi
s
s
ion
pr
otocol
be
twe
e
n
tr
a
ns
mi
tt
e
r
a
nd
the
two
us
e
r
B
a
s
e
d
on
the
a
bov
e
e
xpr
e
s
s
ion,
ther
e
is
no
ne
e
d
f
or
e
s
ti
mation
da
ta.
At
the
f
i
r
s
t
pha
s
e
,
the
a
na
log
pr
e
c
oding
a
nd
c
ombi
ning
matr
ice
s
a
r
e
c
a
lcula
ted
to
c
onve
xit
y
the
non
-
c
onve
x
e
xpr
e
s
s
ion.
T
hus
,
th
e
digi
tal
pr
e
c
oding
is
opti
mi
z
e
d.
I
n
thi
s
s
e
c
ti
on,
we
c
om
pa
r
e
the
pr
opos
e
d
s
olut
ion
with
the
p
r
e
vious
w
or
ks
a
nd
the
f
ull
y
digi
tal
pr
e
c
ode
r
(
opti
mal
c
a
s
e
)
.
T
he
s
of
twa
r
e
pa
c
ka
ge
is
M
a
tl
a
b
f
or
s
im
ulation
a
nd
e
v
a
luation.
T
a
ble
1
s
hows
the
s
im
ulation
pa
r
a
mete
r
s
.
T
a
ble
1.
T
he
s
im
ulation
pa
r
a
mete
r
s
P
a
r
a
me
te
r
s
V
a
lu
e
s
N
umbe
r
of
U
P
A
T
X
a
nt
e
nna
s
N
umbe
r
of
U
P
A
R
X
a
nt
e
nna
s
T
he
numbe
r
of
us
e
r
s
T
he
a
z
im
ut
h A
oA
s
/AoDs
T
he
e
le
va
ti
on A
oA
s
/AoDs
T
he
numbe
r
of
c
ha
nne
l
pa
th
s
64,81,256
4,16
4,8
[0
;
2
π
]
[
-
π/
2
;
π/
2]
1,10
N
umbe
r
of
i
te
r
a
ti
ons
1000
3.
RE
S
UL
T
S
A
ND
AN
AL
YSI
S
W
it
h
the
incr
e
a
s
e
of
S
NR
,
ther
e
is
no
doubt
that
t
he
pr
opor
ti
ona
l
logar
it
hmi
c
r
e
lations
hip
incr
e
a
s
e
s
the
s
pe
c
tr
a
l
e
f
f
icie
nc
y.
F
igur
e
4
de
mons
tr
a
tes
that
only
-
a
na
log
be
a
mf
or
mi
ng
is
not
a
de
qua
te
due
to
the
ove
r
a
ll
r
e
s
tr
iction
of
one
R
F
c
ha
in.
M
or
e
ove
r
,
pha
s
e
s
hif
te
r
s
c
a
n
only
be
d
igi
tally
c
ontr
ol
led
with
qua
nti
z
e
d
pha
s
e
s
,
whic
h
r
e
duc
e
s
the
pos
s
ibi
li
ti
e
s
f
or
a
dva
nc
e
d
pr
oc
e
s
s
ing
a
nd
lea
ds
to
poo
r
pe
r
f
or
manc
e
.
T
he
r
e
f
or
e
,
only
one
da
ta
s
tr
e
a
m
c
a
n
be
ha
ndled,
a
nd
a
s
ignal
be
a
m
c
a
n
be
ge
ne
r
a
ted.
W
hil
e
ther
e
a
r
e
ma
ny
R
F
c
ha
ins
us
e
d
by
the
digi
tal
pr
e
c
oding.
C
ons
e
que
ntl
y,
the
r
e
a
r
e
s
e
ve
r
a
l
da
ta
s
tr
e
a
ms
to
ha
ndle,
a
nd
mul
ti
ple
be
a
ms
a
r
e
c
r
e
a
ted
f
r
om
a
s
ingl
e
a
r
r
a
y
s
im
ult
a
ne
ous
ly.
As
a
r
e
s
ult
,
our
pr
opos
e
d
s
olut
ion
e
xploi
t
s
a
na
log
be
a
mf
or
mer
s
in
the
R
F
domain
a
nd
digi
t
a
l
pr
e
c
o
de
r
s
in
the
ba
s
e
ba
nd
that
lea
ds
to
incr
e
a
s
e
da
ta
r
a
tes
a
nd
s
pe
c
tr
a
l
e
f
f
icie
nc
y
with
di
mi
nis
hing
the
numbe
r
of
R
F
c
ha
ins
a
nd
pr
oc
e
s
s
ing
mul
ti
ple
da
ta
s
tr
e
a
ms
.
T
he
pr
opos
e
d
s
olut
ion
include
s
s
ys
tem
a
r
c
hit
e
c
tur
e
with
the
number
of
R
F
c
ha
ins
a
t
the
B
S
a
nd
a
n
R
F
c
ha
in
pe
r
us
e
r
unde
r
s
im
ulation
c
onf
igur
a
ti
on
in
the
mul
ti
pa
th
c
ondit
ion.
On
the
o
ther
ha
nd
,
th
e
other
s
pr
e
c
oding
pe
r
f
or
ms
be
tt
e
r
than
the
Z
F
pr
e
c
odin
g
be
c
a
us
e
they
do
no
t
a
mp
li
f
y
the
nois
e
c
ompa
r
e
d
with
the
Z
F
p
r
e
c
oding.
T
he
f
indi
ngs
obtaine
d
by
the
pr
opos
e
d
s
olut
ion
c
los
e
to
the
s
ingl
e
-
us
e
r
one
that
m
e
a
ns
our
pr
opos
e
d
c
a
n
e
li
mi
na
te
int
e
r
-
us
e
r
int
e
r
f
e
r
e
nc
e
,
a
s
we
ll
a
s
the
number
of
a
ntenna
s
a
t
B
S
a
nd
M
S
s
t
ha
t
give
a
c
ha
nc
e
to
r
e
duc
e
int
e
r
f
e
r
e
nc
e
.
I
t
is
e
xpli
c
it
that
the
pr
opos
e
d
a
lgor
it
hm
pe
r
f
or
manc
e
is
higher
tha
n
hybr
id
pr
e
c
oding
f
or
K
a
lm
a
n,
a
nd
M
M
S
E
by
ne
a
r
ly
0.
3b/s
/Hz
,
a
nd
0.
7b/s
/
Hz
,
r
e
s
pe
c
ti
ve
ly,
a
t
S
NR
=
20dB
.
T
he
r
e
a
s
on
f
or
thi
s
is
that
our
pr
opos
e
d
s
olut
ion
s
pli
t
the
hybr
id
pr
e
c
oding
a
nd
c
ombi
n
in
g
pr
o
blem
int
o
s
ub
-
pr
oblems
.
As
a
r
e
s
ult
,
that
lea
ds
to
e
na
bli
ng
be
tt
e
r
a
djus
tm
e
nt
of
the
pr
e
c
oding
ba
s
e
ba
nd
ma
tr
ix,
a
nd
other
s
matr
ice
s
.
F
igur
e
5
indi
c
a
tes
the
s
im
ulation
c
onf
igu
r
a
ti
on
w
i
th
the
number
of
a
ntenna
s
NB
S
=
81
,
NM
S
=
4,
the
number
of
us
e
r
s
=
4,
a
nd
the
number
o
f
c
ha
nn
e
l
pa
ths
=
10
while
F
igu
r
e
6
r
e
f
e
r
s
NB
S
=
256,
N
M
S
=
16,
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
I
ntegr
ati
ng
mill
ime
ter
w
av
e
w
it
h
hy
br
id
pr
e
c
oding
multi
us
e
r
mas
s
ive
…
(
M
ohamm
e
d
K
hudhur
Hus
s
e
in
)
95
the
number
of
us
e
r
s
=
4
,
a
nd
the
number
of
c
ha
nne
l
pa
ths
=
10
.
W
it
h
the
incr
e
a
s
e
in
the
nu
mber
o
f
a
ntenna
s
a
t
B
S
a
nd
M
S
.
As
a
r
e
s
ult
,
that
lea
ds
to
incr
e
a
s
e
t
he
s
pe
c
tr
a
l
e
f
f
icie
nc
y,
a
nd
da
ta
r
a
te
owning
of
mo
r
e
r
e
us
e
c
ha
nne
ls
ba
ndwidths
.
I
n
pr
a
c
ti
c
e
,
mo
r
e
a
ntenna
s
a
r
e
r
e
quir
e
d
a
t
the
B
S
than
a
t
the
M
S
.
T
he
f
indi
n
gs
ve
r
if
y
that
the
p
r
opos
e
d
s
olut
ions
a
r
e
be
ing
im
p
r
ove
d
whe
r
e
the
number
o
f
the
a
ntenna
is
mor
e
.
Ac
c
or
ding
to
t
he
two
f
igu
r
e
s
be
low,
the
p
r
opos
e
d
a
lgor
it
hm
pe
r
f
or
manc
e
is
higher
than
hybr
id
pr
e
c
oding
f
or
Ka
l
man,
Z
F
,
a
nd
M
M
S
E
.
T
he
r
e
a
s
on
f
o
r
thi
s
is
that
the
numbe
r
of
c
ha
nne
l
f
e
e
dba
c
k
bit
s
a
nd
a
ntenna
s
a
r
e
dir
e
c
tl
y
r
e
late
d.
F
igur
e
s
7
a
nd
8
indi
c
a
tes
the
s
im
ulation
c
onf
ig
ur
a
ti
on
w
it
h
the
number
of
c
ha
nne
l
pa
ths
=
10,
the
number
of
us
e
r
s
=
8,
a
nd
other
di
f
f
e
r
e
nt
pa
r
a
mete
r
s
.
W
it
h
the
incr
e
a
s
e
the
number
o
f
us
e
r
s
the
Z
F
pr
e
c
oding
is
not
s
uf
f
icie
nt
owing
to
it
s
mul
ti
-
pa
th
f
a
il
ur
e
,
while
the
pr
opos
e
d
a
lgor
it
hm
p
r
ovides
the
be
s
t
s
pe
c
tr
a
l
e
f
f
icie
nc
y
with
the
Ka
lm
a
n
a
nd
the
M
M
S
E
pr
e
c
oding
.
T
he
r
e
a
s
on
f
or
thi
s
the
be
tt
e
r
a
djus
tm
e
nt
of
the
pr
e
c
oding
ba
s
e
ba
nd
matr
ix,
a
nd
othe
r
s
matr
ice
s
by
pr
opos
e
d
s
olut
ion.
F
igur
e
9
indi
c
a
tes
the
s
im
ulation
c
onf
igur
a
ti
on
with
the
number
of
us
e
r
s
=
8
a
nd
the
num
be
r
o
f
a
ntenna
s
NB
S
=
64,
NM
S
=
4.
I
t
s
hows
that
the
d
if
f
e
r
e
nt
hybr
id
a
lgor
it
hms
ha
ve
a
s
im
il
a
r
r
e
s
ult
f
o
r
a
s
ingl
e
pa
th
s
c
e
na
r
io.
T
he
incr
e
a
s
ing
number
of
pa
ths
mea
ns
that
their
s
pe
c
tr
a
l
e
f
f
icie
nc
ies
a
r
e
s
hif
ted
a
w
a
y
f
r
om
the
f
ull
y
digi
tal
c
ur
ve
,
whe
r
e
a
s
the
Z
F
pr
e
c
oding
c
ons
ider
s
wor
s
e
c
a
s
e
due
to
it
s
lac
k
in
e
xploi
ti
ng
mul
ti
pa
th
c
ha
nne
l
ga
ins
.
F
igur
e
4.
V
a
r
ious
a
lgor
it
hms
in
a
64×
4
mm
W
a
ve
M
I
M
O
s
ys
tem
with
the
number
o
f
us
e
r
s
=
4
a
nd
the
number
o
f
c
ha
nne
l
pa
ths
=
10
F
igur
e
5.
D
if
f
e
r
e
nt
a
lgor
it
hms
in
a
n
81×
4
the
mm
W
a
ve
M
I
M
O
s
ys
tem
with
the
number
of
us
e
r
s
=
4
a
nd
the
number
o
f
the
c
ha
nne
l
pa
ths
=
10
F
igur
e
6
.
D
if
f
e
r
e
nt
a
lgor
it
hms
in
a
256×
16
the
mm
W
a
ve
M
I
M
O
s
ys
tem
with
the
number
of
us
e
r
s
=
4
a
nd
the
number
o
f
the
c
ha
nne
l
pa
ths
=
10
F
igur
e
7.
D
if
f
e
r
e
nt
a
lgor
it
hms
in
a
64×
4
the
mm
W
a
ve
M
I
M
O
s
ys
tem
with
the
number
of
us
e
r
s
=
8
a
nd
the
number
o
f
the
c
ha
nne
l
pa
ths
=
10
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
:
90
-
98
96
F
igur
e
8.
D
if
f
e
r
e
nt
a
lgor
it
hms
in
a
256×
16
mm
W
a
ve
M
I
M
O
s
ys
tem
with
the
number
of
us
e
r
s
=
8
a
nd
the
number
o
f
the
c
ha
nne
l
pa
ths
=
10
F
igur
e
9
.
D
if
f
e
r
e
nt
a
lgor
it
hms
in
a
256
×
16
mm
W
a
ve
M
I
M
O
s
ys
tem
with
n
the
number
of
us
e
r
s
=
8
a
nd
the
number
o
f
c
ha
nne
l
pa
ths
=
1
I
t
is
a
ppa
r
e
nt
in
the
major
it
y
of
c
a
s
e
s
;
the
pr
o
pos
e
d
be
a
mf
or
mi
ng
pe
r
f
or
manc
e
is
higher
than
only
-
a
na
log
be
a
mf
or
mi
ng,
s
ingl
e
-
us
e
r
(
no
int
e
r
f
e
r
e
nc
e
)
,
the
Z
F
pr
e
c
oding,
the
M
M
S
E
pr
e
c
odi
ng,
a
nd
the
Ka
lm
a
n
pr
e
c
oding
whe
r
e
the
f
ull
digi
tal
s
olut
ion
is
a
c
ons
i
de
r
a
ble
a
s
the
be
nc
hmar
k
point
with
dif
f
e
r
e
nt
s
c
e
na
r
ios
.
Hybr
id
pr
e
c
oding
ha
s
a
higher
c
ove
r
a
ge
ga
in
than
a
na
log
be
a
mf
o
r
mi
ng
,
e
s
pe
c
ially
f
or
mas
s
ive
number
s
of
B
S
a
ntenna
s
.
Ou
r
pr
opos
e
d
s
olut
ion
c
a
n
s
e
r
ve
a
lar
ge
number
of
us
e
r
s
s
im
ult
a
ne
ous
ly
due
to
mor
e
dir
e
c
ti
ve
ga
in
by
us
ing
numer
ous
a
nten
na
s
a
t
B
S
.
B
a
s
e
d
on
it
s
les
s
c
ompl
e
xit
y
a
nd
ke
e
ping
the
pe
r
f
or
manc
e
,
our
s
olut
ion
c
a
n
be
r
e
c
omm
e
nde
d.
4.
CONC
L
USI
ON
I
n
thi
s
wo
r
k,
we
ha
ve
pr
opos
e
d
a
hybr
id
be
a
mf
o
r
mi
ng
s
c
he
me
ba
s
e
d
on
the
c
onve
x
opti
mi
z
a
ti
on
pr
oblem
f
o
r
M
U
-
M
a
s
s
ive
M
I
M
O
s
ys
tem
s
.
T
he
a
na
log
be
a
mf
or
mi
ng
a
nd
c
ombi
ning
a
r
e
de
s
igned
to
obtain
the
maximum
e
ne
r
gy
pr
inciple
f
or
s
ingl
e
-
us
e
s
ys
te
ms
.
Af
ter
that
,
the
c
onve
x
opti
m
iza
ti
on
pr
oblem
is
us
e
d
to
e
s
ti
mate
the
digi
tal
pr
e
c
oding
to
e
li
mi
na
te
int
e
r
-
u
s
e
r
int
e
r
f
e
r
e
nc
e
.
Unde
r
the
s
a
me
c
ondit
ions
,
our
a
na
lyt
ica
l
a
nd
s
im
ulation
f
indi
ngs
s
how
that
p
r
opos
e
d
pr
e
c
o
ding
a
c
hieve
s
be
tt
e
r
s
pe
c
tr
a
l
e
f
f
icie
nc
y
than
other
e
xis
ti
ng
hybr
ids
.
I
n
the
f
utur
e
,
our
wor
k
will
be
e
xtende
d
to
joi
n
t
hybr
id
pr
e
c
oding
with
us
e
r
-
be
a
m
s
c
he
du
li
ng
to
lowe
r
c
ompl
e
xit
y
mo
r
e
.
RE
F
E
RE
NC
E
S
[1
]
R.
I.
Bo
b
y
,
K
.
A
b
d
u
l
l
ah
,
A
.
Z
.
J
u
s
o
h
,
N
.
Par
v
een
,
an
d
A
.
L
.
A
s
n
aw
i
,
“A
w
i
re
l
es
s
p
reco
d
i
n
g
t
ec
h
n
i
q
u
e
fo
r
mi
l
l
i
me
t
re
-
w
a
v
e
MIMO
s
y
s
t
em
b
as
e
d
o
n
SIC
-
MMSE
,
”
TE
LKO
M
NIK
A
Tel
eco
m
m
u
n
i
ca
t
i
o
n
Co
m
p
u
t
i
n
g
E
l
ec
t
r
o
n
i
c
s
a
n
d
C
o
n
t
r
o
l
,
v
o
l
.
1
7
,
n
o
.
6
,
D
ec
2
0
1
9
,
d
o
i
:
1
0
.
1
2
9
2
8
/
T
E
L
K
O
MN
I
K
A
.
v
1
7
i
6
.
1
2
8
0
2
.
[2
]
K
.
D
u
an
,
H
.
D
u
,
an
d
Z
.
W
u
,
“H
y
b
r
i
d
A
l
t
ern
a
t
i
n
g
Pr
eco
d
i
n
g
an
d
Co
m
b
i
n
i
n
g
D
e
s
i
g
n
f
o
r
mmW
av
e
Mu
l
t
i
-
U
s
er
MIMO
Sy
s
t
ems
,
”
2
0
1
8
IE
E
E
/
CIC
In
t
er
n
a
t
i
o
n
a
l
Co
n
f
e
r
en
ce
o
n
C
o
m
m
u
n
i
c
a
t
i
o
n
s
i
n
Ch
i
n
a
(ICCC)
,
Bei
j
i
n
g
,
Ch
i
n
a
,
p
p
.
2
1
7
-
2
2
1
,
2
0
1
8
,
d
o
i
:
1
0
.
1
1
0
9
/
ICCCh
i
n
a.
2
0
1
8
.
8
6
4
1
1
6
3
.
[3
]
F.
L
i
u
,
X
.
K
an
,
X
.
Bai
,
R.
D
u
,
an
d
Y
.
Z
h
an
g
,
“T
w
o
-
St
a
g
e
H
y
b
ri
d
Preco
d
i
n
g
A
l
g
o
ri
t
h
m
Ba
s
ed
o
n
Sw
i
t
c
h
N
et
w
o
rk
fo
r
Mi
l
l
i
me
t
er
W
av
e
Mi
m
o
Sy
s
t
ems
,
”
P
r
o
g
r
es
s
i
n
E
l
e
ct
r
o
m
a
g
n
et
i
cs
R
e
s
ea
r
ch
M
,
v
o
l
.
7
7
,
p
p
.
1
0
3
–
1
1
3
,
J
an
u
ary
2
0
1
9
,
d
o
i
:
1
0
.
2
5
2
8
/
PI
E
RM1
8
1
0
2
8
0
1
.
[4
]
Y
.
L
u
,
C.
Ch
en
g
,
J
.
Y
an
g
,
an
d
G
.
G
u
i
,
“Imp
ro
v
ed
h
y
b
r
i
d
p
reco
d
i
n
g
s
c
h
eme
fo
r
mmW
av
e
l
ar
g
e
-
s
ca
l
e
MIMO
s
y
s
t
ems
,
”
I
E
E
E
A
cces
s
,
v
o
l
.
7
,
p
p
.
1
2
0
2
7
–
1
2
0
3
4
,
2
0
1
9
,
doi
:
1
0
.
1
1
0
9
/
A
CC
E
SS.
2
0
1
9
.
2
8
9
2
1
3
6
.
[5
]
H
.
Y
u
an
,
J
.
A
n
,
N
.
Y
an
g
,
K
.
Y
an
g
,
an
d
T
.
Q
.
D
u
o
n
g
,
“L
o
w
co
mp
l
e
x
i
t
y
h
y
b
r
i
d
p
rec
o
d
i
n
g
fo
r
mu
l
t
i
u
s
er
mi
l
l
i
m
et
er
w
av
e
s
y
s
t
ems
o
v
er
freq
u
en
c
y
s
e
l
ect
i
v
e
ch
a
n
n
e
l
s
,
”
IE
E
E
Tr
a
n
s
a
ct
i
o
n
s
o
n
V
e
h
i
c
u
l
a
r
Tech
n
o
l
o
g
y
,
v
o
l
.
6
8
,
n
o
.
1
,
p
p
.
9
8
3
-
9
8
7
,
J
an
2
0
1
9
,
d
o
i
:
1
0
.
1
1
0
9
/
T
V
T
.
2
0
1
8
.
2
8
8
0
7
8
7
.
[6
]
D
eep
i
k
a
D
.
Pai
,
“A
Su
rv
ey
o
n
Mi
l
l
i
met
er
W
a
v
e
Mo
b
i
l
e
Co
mmu
n
i
ca
t
i
o
n
s
fo
r
5
G
Cel
l
u
l
ar
N
et
w
o
r
k
s
,
”
IJIR
E
E
I
CE
In
t
e
r
n
a
t
i
o
n
a
l
Jo
u
r
n
a
l
o
f
I
n
n
o
va
t
i
ve
R
es
e
a
r
c
h
i
n
E
l
ec
t
r
i
ca
l
,
E
l
ec
t
r
o
n
i
cs
,
In
s
t
r
u
m
en
t
a
t
i
o
n
a
n
d
C
o
n
t
r
o
l
E
n
g
i
n
eer
i
n
g
,
v
o
l
.
5
,
n
o
.
6
,
p
p
.
2
7
8
–
2
8
4
,
J
u
n
e
2
0
1
7
.
[7
]
Y
.
G
u
o
,
L
.
L
i
,
X
.
W
en
,
W
.
Ch
en
,
an
d
Z
.
H
an
,
“Su
b
-
arra
y
-
b
a
s
ed
h
y
b
r
i
d
p
reco
d
i
n
g
d
e
s
i
g
n
fo
r
d
o
w
n
l
i
n
k
mi
l
l
i
me
t
er
-
w
a
v
e
mu
l
t
i
-
u
s
er
mas
s
i
v
e
MIM
O
s
y
s
t
e
ms
,
”
2
0
1
7
9
th
In
t
e
r
n
a
t
i
o
n
a
l
Co
n
f
er
e
n
ce
o
n
W
i
r
el
es
s
Co
m
m
u
n
i
c
a
t
i
o
n
s
a
n
d
S
i
g
n
a
l
P
r
o
ce
s
s
i
n
g
(W
C
S
P
)
,
N
a
n
j
i
n
g
,
p
p
.
1
–
4
,
2
0
1
7
,
d
o
i
:
1
0
.
1
1
0
9
/
W
CSP.
2
0
1
7
.
8
1
7
1
1
5
9
.
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
I
ntegr
ati
ng
mill
ime
ter
w
av
e
w
it
h
hy
br
id
pr
e
c
oding
multi
us
e
r
mas
s
ive
…
(
M
ohamm
e
d
K
hudhur
Hus
s
e
in
)
97
[8
]
S.
Bl
an
d
i
n
o
,
G
.
Man
g
rav
i
t
i
,
C.
D
e
s
s
e
t
,
A
.
Bo
u
r
d
o
u
x
,
P.
W
amb
acq
,
an
d
S.
Po
l
l
i
n
,
“Mu
l
t
i
-
U
s
er
H
y
b
ri
d
MIM
O
at
6
0
G
H
z
U
s
i
n
g
1
6
-
A
n
t
en
n
a
T
ra
n
s
m
i
t
t
ers
,
”
I
E
E
E
Tr
a
n
s
a
ct
i
o
n
s
o
n
C
i
r
c
u
i
t
s
a
n
d
S
y
s
t
e
m
s
I:
R
eg
u
l
a
r
P
a
p
er
s
,
v
o
l
.
6
6
,
n
o
.
2
,
p
p
.
8
4
8
-
8
5
8
,
Feb
.
2
0
1
9
,
d
o
i
:
1
0
.
1
1
0
9
/
T
CSI.
2
0
1
8
.
2
8
6
6
9
3
3
.
[9
]
J
.
D
u
,
W
.
X
u
,
H
.
Sh
e
n
,
X
.
D
o
n
g
,
an
d
C.
Z
h
ao
,
“H
y
b
r
i
d
p
rec
o
d
i
n
g
arch
i
t
ec
t
u
re
f
o
r
ma
s
s
i
v
e
m
u
l
t
i
u
s
er
MIM
O
w
i
t
h
d
i
s
s
i
p
at
i
o
n
:
Su
b
-
co
n
n
ec
t
ed
o
r
fu
l
l
y
co
n
n
ect
e
d
s
t
r
u
ct
u
res
?,
”
IE
E
E
Tr
a
n
s
a
c
t
i
o
n
s
o
n
W
i
r
e
l
es
s
Co
m
m
u
n
i
ca
t
i
o
n
s
,
v
o
l
.
1
7
,
n
o
.
8
,
p
p
.
5
4
6
5
-
5
4
7
9
,
A
u
g
2
0
1
8
,
d
o
i
:
1
0
.
1
1
0
9
/
T
W
C.
2
0
1
8
.
2
8
4
4
2
0
7
.
[1
0
]
J
.
J
i
n
,
Y
.
R.
Z
h
en
g
,
W
.
Ch
en
,
an
d
C.
X
i
ao
,
“H
y
b
ri
d
Pr
eco
d
i
n
g
fo
r
Mi
l
l
i
met
er
W
a
v
e
MIMO
Sy
s
t
em
s
w
i
t
h
Fi
n
i
t
e
-
A
l
p
h
a
b
et
In
p
u
t
s
,
”
G
L
O
B
E
CO
M
2
0
1
7
-
2
0
1
7
IE
E
E
G
l
o
b
a
l
C
o
m
m
u
n
i
c
a
t
i
o
n
s
Co
n
f
e
r
en
ce
,
S
i
n
g
ap
o
re,
p
p
.
1
–
6
,
201
7
,
d
o
i
:
1
0
.
1
1
0
9
/
G
L
O
CO
M.
2
0
1
7
.
8
2
5
4
8
5
.
[1
1
]
S.
A
l
i
,
S.
A
l
q
u
h
a
i
f
,
an
d
I.
A
h
mad
,
“A
n
o
p
t
i
mi
ze
d
h
y
b
ri
d
b
eamfo
rmi
n
g
fo
r
mi
l
l
i
me
t
er
w
av
e
mu
s
as
s
i
v
e
MIMO
s
y
s
t
em,
”
3
C
Tecn
o
l
o
g
i
a
G
l
o
s
a
s
d
e
In
n
o
va
c
i
ó
n
A
p
l
i
ca
d
a
s
a
l
a
p
y
m
e
,
v
o
l
.
7
,
n
o
.
4
,
p
p
.
9
3
–
1
0
8
,
J
an
u
ary
201
9
,
doi
:
1
0
.
1
7
9
9
3
/
3
c
t
ecn
o
.
2
0
1
9
.
s
p
ec
i
al
i
s
s
u
e.
0
9
[1
2
]
A
.
F.
Mo
l
i
s
ch
et
al
.
,
“H
y
b
ri
d
Beamfo
rmi
n
g
fo
r
Mas
s
i
v
e
MIMO
:
A
Su
rv
ey
,
”
i
n
IE
E
E
Co
m
m
u
n
i
c
a
t
i
o
n
s
M
a
g
a
z
i
n
e
,
v
o
l
.
5
5
,
n
o
.
9
,
p
p
.
1
3
4
-
1
4
1
,
Sep
t
2
0
1
7
,
d
o
i
:
1
0
.
1
1
0
9
/
MC
O
M.
2
0
1
7
.
1
6
0
0
4
0
0
.
[1
3
]
Z
.
Z
h
o
u
,
N
.
G
e,
Z
.
W
an
g
,
an
d
S.
Ch
e
n
,
“H
ar
d
w
are
-
ef
fi
ci
e
n
t
h
y
b
ri
d
p
rec
o
d
i
n
g
fo
r
m
i
l
l
i
me
t
er
w
a
v
e
s
y
s
t
ems
w
i
t
h
mu
l
t
i
-
fee
d
refl
ect
arr
ay
s
,
”
I
E
E
E
A
cces
s
,
v
o
l
.
6
,
p
p
.
6
7
9
5
-
6
8
0
6
,
2
0
1
8
,
d
o
i
:
1
0
.
1
1
0
9
/
A
CC
E
SS.
2
0
1
8
.
2
7
9
3
2
2
3
.
[1
4
]
O
.
A
l
l
u
h
ai
b
i
,
Q
.
Z
.
A
h
me
d
,
C.
Pan
,
a
n
d
H
.
Z
h
u
,
“
H
y
b
ri
d
D
i
g
i
t
a
l
-
to
-
A
n
al
o
g
Beam
fo
rm
i
n
g
A
p
p
r
o
ach
e
s
t
o
Ma
x
i
mi
s
e
t
h
e
Cap
ac
i
t
y
o
f
mm
-
W
av
e
S
y
s
t
ems
,
”
2
0
1
7
I
E
E
E
8
5
th
V
eh
i
cu
l
a
r
Tech
n
o
l
o
g
y
Co
n
f
e
r
en
ce
(V
TC
S
p
r
i
n
g
)
,
Sy
d
n
ey
,
N
SW
,
p
p
.
1
-
5,
2
0
1
7
,
d
o
i
:
1
0
.
1
1
0
9
/
V
T
CSp
r
i
n
g
.
2
0
1
7
.
8
1
0
8
3
8
5
.
[1
5
]
T
.
E
.
Bo
g
al
e
a
n
d
L
.
B.
L
e,
“Beamfo
rmi
n
g
fo
r
m
u
l
t
i
u
s
er
ma
s
s
i
v
e
MIM
O
s
y
s
t
ems
:
D
i
g
i
t
a
l
v
er
s
u
s
h
y
b
ri
d
an
al
o
g
-
d
i
g
i
t
al
,
”
2
0
1
4
IE
E
E
G
l
o
b
a
l
Co
m
m
u
n
i
ca
t
i
o
n
s
C
o
n
f
er
e
n
ce
,
A
u
s
t
i
n
,
T
X
,
p
p
.
4
0
6
6
–
4
0
7
1
,
2
0
1
4
,
doi
:
1
0
.
1
1
0
9
/
G
L
O
CO
M.
2
0
1
4
.
7
0
3
7
4
4
4
.
[1
6
]
M.
K
i
m,
J
.
L
ee,
an
d
J
.
L
ee,
“H
y
b
ri
d
b
eam
f
o
r
mi
n
g
f
o
r
mu
l
t
i
-
u
s
er
t
ra
n
s
m
i
s
s
i
o
n
i
n
mi
l
l
i
me
t
er
w
av
e
co
mmu
n
i
ca
t
i
o
n
s
,
”
2
0
1
7
In
t
er
n
a
t
i
o
n
a
l
Co
n
f
e
r
e
n
ce
o
n
I
n
f
o
r
m
a
t
i
o
n
a
n
d
Co
m
m
u
n
i
ca
t
i
o
n
Tec
h
n
o
l
o
g
y
Co
n
ver
g
e
n
ce
(ICTC)
,
J
ej
u
,
p
p
.
1
2
6
0
–
1
2
6
2
,
2
0
1
7
,
d
o
i
:
1
0
.
1
1
0
9
/
ICT
C.
2
0
1
7
.
8
1
9
0
9
1
5
.
[1
7
]
A
.
A
l
k
h
at
ee
b
,
G
.
L
eu
s
,
an
d
R.
W
.
H
eat
h
,
“L
i
mi
t
ed
Feed
b
ac
k
H
y
b
r
i
d
Preco
d
i
n
g
fo
r
Mu
l
t
i
-
U
s
e
r
Mi
l
l
i
met
er
W
av
e
Sy
s
t
em
s
,
”
IE
E
E
Tr
a
n
s
a
ct
i
o
n
s
o
n
W
i
r
e
l
es
s
Co
m
m
u
n
i
ca
t
i
o
n
s
,
v
o
l
.
1
4
,
n
o
.
1
1
,
p
p
.
6
4
8
1
-
6
4
9
4
,
N
o
v
2
0
1
5
,
doi
:
1
0
.
1
1
0
9
/
T
W
C.
2
0
1
5
.
2
4
5
5
9
8
0
.
[1
8
]
W
.
N
i
a
n
d
X
.
D
o
n
g
,
“H
y
b
r
i
d
B
l
o
c
k
D
i
ag
o
n
a
l
i
za
t
i
o
n
f
o
r
Mas
s
i
v
e
Mu
l
t
i
u
s
er
MIMO
S
y
s
t
ems
,
”
IE
E
E
T
r
a
n
s
a
ct
i
o
n
s
o
n
Co
m
m
u
n
i
ca
t
i
o
n
s
,
v
o
l
.
6
4
,
n
o
.
1
,
p
p
.
2
0
1
-
2
1
1
,
J
an
2
0
1
6
,
doi
:
1
0
.
1
1
0
9
/
T
CO
MM.
2
0
1
5
.
2
5
0
2
9
5
4
.
[1
9
]
F.
So
h
rab
i
an
d
W
.
Y
u
,
“H
y
b
ri
d
D
i
g
i
t
a
l
an
d
A
n
al
o
g
Beamfo
rmi
n
g
D
es
i
g
n
fo
r
L
arg
e
-
Sca
l
e
A
n
t
e
n
n
a
A
rr
ay
s
,
”
i
n
IE
E
E
Jo
u
r
n
a
l
o
f
S
e
l
ect
ed
To
p
i
c
s
i
n
S
i
g
n
a
l
P
r
o
ces
s
i
n
g
,
v
o
l
.
1
0
,
n
o
.
3
,
p
p
.
5
0
1
-
5
1
3
,
A
p
r
i
l
2
0
1
6
,
doi
:
1
0
.
1
1
0
9
/
J
S
T
SP.
2
0
1
6
.
2
5
2
0
9
1
2
.
[2
0
]
D
.
Y
i
n
g
,
F.
W
.
V
o
o
k
,
T
.
A
.
T
h
o
ma
s
,
an
d
D
.
J
.
L
o
v
e,
“H
y
b
ri
d
s
t
r
u
ct
u
re
i
n
mas
s
i
v
e
MIMO
:
A
c
h
i
e
v
i
n
g
l
arg
e
s
u
m
rat
e
w
i
t
h
few
er
RF
ch
ai
n
s
,
”
2
0
1
5
IE
E
E
In
t
e
r
n
a
t
i
o
n
a
l
Co
n
f
e
r
en
ce
o
n
Co
m
m
u
n
i
c
a
t
i
o
n
s
(ICC)
,
L
o
n
d
o
n
,
p
p
.
2
3
4
4
–
2
3
4
9
,
2
0
1
5
,
d
o
i
:
1
0
.
1
1
0
9
/
ICC.
2
0
1
5
.
7
2
4
8
6
7
5
[2
1
]
L
.
L
i
an
g
,
W
.
X
u
,
an
d
X
.
D
o
n
g
,
“L
o
w
-
Co
mp
l
ex
i
t
y
H
y
b
r
i
d
Preco
d
i
n
g
i
n
Mas
s
i
v
e
Mu
l
t
i
u
s
er
MIMO
Sy
s
t
ems
,
”
IE
E
E
W
i
r
el
e
s
s
Co
m
m
u
n
i
ca
t
i
o
n
s
Let
t
er
s
,
v
o
l
.
3
,
n
o
.
6
,
p
p
.
6
5
3
-
6
5
6
,
D
ec
2
0
1
4
,
d
o
i
:
1
0
.
1
1
0
9
/
L
W
C.
2
0
1
4
.
2
3
6
3
8
3
1
.
[2
2
]
J
.
Z
h
an
g
,
M.
H
aard
t
,
I.
So
l
o
v
ey
c
h
i
k
,
an
d
A
.
W
i
es
e
l
,
“A
ch
an
n
el
mat
c
h
i
n
g
b
a
s
ed
h
y
b
r
i
d
an
al
o
g
-
d
i
g
i
t
al
s
t
rat
e
g
y
fo
r
mas
s
i
v
e
m
u
l
t
i
-
u
s
er
MIM
O
d
o
w
n
l
i
n
k
s
y
s
t
ems
,
”
2
0
1
6
IE
E
E
S
e
n
s
o
r
A
r
r
a
y
a
n
d
M
u
l
t
i
c
h
a
n
n
e
l
S
i
g
n
a
l
P
r
o
ce
s
s
i
n
g
W
o
r
ks
h
o
p
(S
A
M
)
,
Ri
o
d
e
J
an
er
i
o
,
p
p
.
1
-
5
,
2
0
1
6
,
d
o
i
:
1
0
.
1
1
0
9
/
S
A
M.
2
0
1
6
.
7
5
6
9
6
9
3
.
[2
3
]
J
.
J
i
n
,
C.
X
i
a
o
,
W
.
Ch
e
n
,
an
d
Y
.
W
u
,
“Ch
an
n
el
-
S
t
at
i
s
t
i
c
s
-
Bas
e
d
H
y
b
r
i
d
Prec
o
d
i
n
g
fo
r
M
i
l
l
i
me
t
er
-
W
a
v
e
MI
MO
Sy
s
t
em
s
w
i
t
h
D
y
n
ami
c
S
u
b
arra
y
s
,
”
i
n
IE
E
E
T
r
a
n
s
a
ct
i
o
n
s
o
n
C
o
m
m
u
n
i
c
a
t
i
o
n
s
,
v
o
l
.
6
7
,
n
o
.
6
,
p
p
.
3
9
9
1
-
4
0
0
3
,
J
u
n
e
2
0
1
9
,
d
o
i
:
1
0
.
1
1
0
9
/
T
C
O
MM.
2
0
1
9
.
2
8
9
9
6
2
8
.
[2
4
]
A
.
A
.
A
s
t
an
e
h
an
d
S.
G
h
ei
s
ar
i
,
“Rev
i
e
w
an
d
Co
m
p
a
ri
s
o
n
o
f
R
o
u
t
i
n
g
Me
t
ri
c
s
i
n
C
o
g
n
i
t
i
v
e
Rad
i
o
N
et
w
o
r
k
s
,
”
E
m
e
r
g
i
n
g
S
ci
e
n
ce
Jo
u
r
n
a
l
,
v
o
l
.
2
,
n
o
.
4
,
p
p
.
1
9
1
–
2
0
1
,
2
0
1
8
,
d
o
i
:
1
0
.
2
8
9
9
1
/
es
j
-
2
0
1
8
-
0
1
1
4
3
.
[2
5
]
K
.
D
.
K
al
i
s
t
at
o
v
,
“W
i
rel
e
s
s
V
i
d
e
o
Mo
n
i
t
o
ri
n
g
o
f
t
h
e
Meg
aci
t
i
e
s
T
ran
s
p
o
rt
In
fra
s
t
r
u
ct
u
re,
”
in
Ci
vi
l
E
n
g
i
n
ee
r
i
n
g
Jo
u
r
n
a
l
v
o
l
.
5
,
n
o
.
5
,
p
p
.
1
0
3
3
–
1
0
4
0
,
May
2
0
1
9
,
d
o
i
:
1
0
.
2
8
9
9
1
/
cej
-
2
0
1
9
-
0
3
0
9
1
3
0
9
.
[2
6
]
A
.
O
mi
d
i
,
R.
K
aram
i
,
P.
S.
E
mad
i
,
an
d
H
.
Mo
ra
d
i
,
“
D
es
i
g
n
o
f
t
h
e
L
o
w
N
o
i
s
e
A
m
p
l
i
f
i
er
C
i
rcu
i
t
i
n
Ba
n
d
L
fo
r
Imp
ro
v
e
t
h
e
G
ai
n
an
d
C
i
rcu
i
t
S
t
ab
i
l
i
t
y
,
”
E
m
er
g
i
n
g
S
c
i
en
ce
Jo
u
r
n
a
l
,
v
o
l
.
1
,
n
o
.
4
,
p
p
.
1
9
2
–
2
0
0
,
D
ecem
b
er
2
0
1
7
,
doi
:
1
0
.
2
8
9
9
1
/
i
j
s
e
-
0
1
1
2
2
.
[2
7
]
M.
Y
azd
an
i
an
d
A
.
A
.
Ra
s
s
af
i
,
“E
v
a
l
u
a
t
i
o
n
o
f
D
r
i
v
er
s
’
A
ffect
a
b
i
l
i
t
y
a
n
d
Sat
i
s
fac
t
i
o
n
w
i
t
h
B
l
a
ck
S
p
o
t
s
W
arn
i
n
g
A
p
p
l
i
cat
i
o
n
,
”
in
Ci
v
i
l
E
n
g
i
n
eer
i
n
g
Jo
u
r
n
a
l
,
v
o
l
.
5
,
n
o
.
3
,
p
p
.
5
7
6
,
March
2
0
1
9
,
d
o
i
:
1
0
.
2
8
9
9
1
/
cej
-
2
0
1
9
-
0
3
0
9
1
2
6
9
.
[2
8
]
V
.
V
en
k
a
t
es
w
ara
n
an
d
A
.
-
J
.
v
an
d
er
V
ee
n
,
“A
n
a
l
o
g
Beamfo
rmi
n
g
i
n
MIMO
Co
mmu
n
i
ca
t
i
o
n
s
w
i
t
h
Ph
as
e
Sh
i
f
t
N
et
w
o
r
k
s
an
d
O
n
l
i
n
e
Ch
an
n
el
E
s
t
i
mat
i
o
n
,
”
in
IE
E
E
Tr
a
n
s
a
ct
i
o
n
s
o
n
S
i
g
n
a
l
P
r
o
ce
s
s
i
n
g
,
v
o
l
.
5
8
,
n
o
.
8
,
p
p
.
4
1
3
1
-
4
1
4
3
,
A
u
g
2
0
1
0
,
d
o
i
:
1
0
.
1
1
0
9
/
T
SP.
2
0
1
0
.
2
0
4
8
3
2
1
.
[2
9
]
F.
G
h
o
l
am,
J
.
V
i
a,
an
d
I.
San
t
amari
a,
“Beamf
o
rmi
n
g
D
es
i
g
n
f
o
r
Si
m
p
l
i
fi
e
d
A
n
a
l
o
g
A
n
t
e
n
n
a
Co
m
b
i
n
i
n
g
A
rch
i
t
ec
t
u
re
s
,
”
IE
E
E
Tr
a
n
s
a
ct
i
o
n
s
o
n
V
eh
i
cu
l
a
r
Tech
n
o
l
o
g
y
,
v
o
l
.
6
0
,
n
o
.
5
,
p
p
.
2
3
7
3
-
2
3
7
8
,
J
u
n
2
0
1
1
,
doi
:
1
0
.
1
1
0
9
/
T
V
T
.
2
0
1
1
.
2
1
4
2
2
0
.
[3
0
]
A
.
V
i
z
zi
e
l
l
o
,
P.
Sav
azzi
,
a
n
d
K
.
R.
C
h
o
w
d
h
u
r
y
,
“A
k
al
man
b
a
s
ed
h
y
b
ri
d
p
reco
d
i
n
g
f
o
r
m
u
l
t
i
-
u
s
er
m
i
l
l
i
me
t
er
w
av
e
MIMO
s
y
s
t
em
s
,
”
IE
E
E
A
cce
s
s
,
v
o
l
.
6
,
p
p
.
5
5
7
1
2
-
5
5
7
2
2
,
2
0
1
8
,
d
o
i
:
1
0
.
1
1
0
9
/
A
CCE
SS.
2
0
1
8
.
2
8
7
2
7
3
8
.
[3
1
]
L
.
Z
h
ao
,
D
.
W
.
K
.
N
g
,
an
d
J
.
Y
u
an
,
“Mu
l
t
i
-
U
s
er
Preco
d
i
n
g
an
d
Ch
a
n
n
e
l
E
s
t
i
ma
t
i
o
n
fo
r
H
y
b
r
i
d
Mi
l
l
i
me
t
er
W
av
e
Sy
s
t
em
s
,
”
IE
E
E
Jo
u
r
n
a
l
o
n
S
el
ect
e
d
A
r
e
a
s
i
n
Co
m
m
u
n
i
c
a
t
i
o
n
s
,
v
o
l
.
3
5
,
n
o
.
7
,
p
p
.
1
5
7
6
-
1
5
9
0
,
J
u
l
y
2
0
1
7
,
doi
:
1
0
.
1
1
0
9
/
J
S
A
C.
2
0
1
7
.
2
6
9
9
3
7
8
.
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
:
90
-
98
98
[3
2
]
D
.
H
.
N
.
N
g
u
y
en
,
L
.
B
.
L
e,
an
d
T
.
L
e
-
N
g
o
c,
“H
y
b
ri
d
MMSE
p
reco
d
i
n
g
fo
r
mmW
av
e
mu
l
t
i
u
s
er
MIMO
s
y
s
t
e
ms
,
”
2
0
1
6
IE
E
E
In
t
e
r
n
a
t
i
o
n
a
l
Co
n
f
e
r
en
ce
o
n
Co
m
m
u
n
i
ca
t
i
o
n
s
(ICC)
,
K
u
al
a
L
u
mp
u
r,
p
p
.
1
–
6,
2
0
1
6
,
doi
:
1
0
.
1
1
0
9
/
ICC.
2
0
1
6
.
7
5
1
0
8
4
4
.
[3
3
]
H
.
M.
E
l
mag
zo
u
b
an
d
Q
.
Ch
en
,
“O
n
t
h
e
Mu
l
t
i
-
U
s
er
MI
MO
H
y
b
r
i
d
Preco
d
i
n
g
D
es
i
g
n
i
n
Mi
l
l
i
me
t
er
W
av
e
Cel
l
u
l
ar
Sy
s
t
em
s
,
”
2
0
1
8
In
t
e
r
n
a
t
i
o
n
a
l
Co
n
f
e
r
en
ce
o
n
M
i
c
r
o
w
a
ve
a
n
d
M
i
l
l
i
m
e
t
er
W
a
ve
Tec
h
n
o
l
o
g
y
(IC
M
M
T)
,
Ch
e
n
g
d
u
,
p
p
.
1
–
3
,
2
0
1
8
,
d
o
i
:
1
0
.
1
1
0
9
/
ICMMT
.
2
0
1
8
.
8
5
6
3
6
3
9
.
[3
4
]
M.
R.
Cas
t
el
l
a
n
o
s
et
al
.
,
“Ch
an
n
e
l
-
Reco
n
s
t
ru
c
t
i
o
n
-
B
as
ed
H
y
b
ri
d
Preco
d
i
n
g
fo
r
Mi
l
l
i
me
t
er
-
W
a
v
e
Mu
l
t
i
-
U
s
e
r
MIMO
Sy
s
t
em
s
,
”
in
IE
E
E
Jo
u
r
n
a
l
o
f
S
el
ec
t
ed
To
p
i
c
s
i
n
S
i
g
n
a
l
P
r
o
ce
s
s
i
n
g
,
v
o
l
.
1
2
,
n
o
.
2
,
p
p
.
3
8
3
-
3
9
8
,
May
2
0
1
8
,
doi
:
1
0
.
1
1
0
9
/
J
S
T
SP.
2
0
1
8
.
2
8
1
9
1
3
5
.
[3
5
]
J
.
Co
n
g
,
T
.
L
i
n
,
an
d
Y
.
Z
h
u
,
“
H
y
b
ri
d
MMSE
b
ea
mfo
rmi
n
g
fo
r
m
u
l
t
i
u
s
er
m
i
l
l
i
me
t
er
-
w
a
v
e
co
mm
u
n
i
ca
t
i
o
n
s
y
s
t
ems
,
”
IE
E
E
Co
m
m
u
n
i
ca
t
i
o
n
s
Let
t
e
r
s
,
v
o
l
.
2
2
,
n
o
.
1
1
,
p
p
.
2
3
9
0
-
2
3
9
3
,
N
o
v
2
0
1
8
,
doi
:
1
0
.
1
1
0
9
/
L
CO
MM.
2
0
1
8
.
2
8
6
9
3
2
9
.
[3
6
]
P.
Y
.
T
s
a
i
an
d
X
.
S.
H
u
an
g
,
“D
es
i
g
n
o
f
It
era
t
i
v
e
H
y
b
r
i
d
Beamfo
rm
i
n
g
fo
r
M
u
l
t
i
-
U
s
er
mm
W
av
e
Ma
s
s
i
v
e
MI
MO
Sy
s
t
em
s
,
”
2
0
1
8
IE
E
E
I
n
t
e
r
n
a
t
i
o
n
a
l
W
o
r
ks
h
o
p
o
n
S
i
g
n
a
l
P
r
o
ces
s
i
n
g
S
y
s
t
e
m
s
(S
i
P
S
)
,
Cap
e
T
o
w
n
,
p
p
.
1
2
3
–
1
2
8
,
2
0
1
8
,
d
o
i
:
1
0
.
1
1
0
9
/
S
i
PS.
2
0
1
8
.
8
5
9
8
3
0
3
.
[3
7
]
Q
.
Sh
i
an
d
M.
H
o
n
g
,
“S
p
ect
ra
l
E
ff
i
ci
e
n
cy
O
p
t
i
mi
zat
i
o
n
Fo
r
Mi
l
l
i
met
er
W
av
e
M
u
l
t
i
u
s
er
MIMO
S
y
s
t
ems
,
”
IE
E
E
Jo
u
r
n
a
l
o
f
S
e
l
ect
e
d
To
p
i
c
s
i
n
S
i
g
n
a
l
P
r
o
ce
s
s
i
n
g
,
v
o
l
.
1
2
,
n
o
.
3
,
p
p
.
4
5
5
-
4
6
8
,
J
u
n
e
2
0
1
8
,
doi
:
1
0
.
1
1
0
9
/
J
S
T
SP.
2
0
1
8
.
2
8
2
4
2
4
6
.
[3
8
]
E
.
Meh
ri
n
ej
a
d
K
h
o
t
b
eh
s
ara
an
d
H
.
Safari
,
“A
Sy
s
t
em
at
i
c
Re
v
i
e
w
o
f
A
ffec
t
i
v
e
Fact
o
rs
o
n
L
o
cat
i
n
g
Sp
ec
i
al
i
zed
H
o
s
p
i
t
a
l
s
,
”
Ci
v
i
l
E
n
g
i
n
ee
r
i
n
g
J
o
u
r
n
a
l
,
v
o
l
.
4
,
n
o
.
9
,
p
p
.
2
2
1
0
-
2
2
1
7
,
Sep
t
emb
er
2
0
1
8
,
d
o
i
:
1
0
.
2
8
9
9
1
/
cej
-
0
3
0
9
1
1
5
1
.
[3
9
]
Y
.
G
ami
l
an
d
I.
A
b
d
u
l
Ra
h
man
,
“I
d
en
t
i
f
i
cat
i
o
n
o
f
Ca
u
s
e
s
an
d
E
ffe
c
t
s
o
f
Po
o
r
Co
mm
u
n
i
cat
i
o
n
i
n
Co
n
s
t
ru
c
t
i
o
n
In
d
u
s
t
r
y
:
A
T
h
eo
re
t
i
ca
l
Rev
i
e
w
,
”
E
m
e
r
g
i
n
g
S
c
i
en
ce
J
o
u
r
n
a
l
,
v
o
l
.
1
,
n
o
.
4
,
2
0
1
7
,
d
o
i
:
1
0
.
2
8
9
9
1
/
i
j
s
e
-
0
1
1
2
1
.
Evaluation Warning : The document was created with Spire.PDF for Python.