Int
ern
at
i
onal
Journ
al of Ele
ctrical
an
d
Co
mput
er
En
gin
eeri
ng
(IJ
E
C
E)
Vo
l.
10
,
No.
3
,
June
2020
,
pp.
3152
~
3157
IS
S
N:
20
88
-
8708
,
DOI: 10
.11
591/
ijece
.
v
10
i
3
.
pp3152
-
31
57
3152
Journ
al
h
om
e
page
:
http:
//
ij
ece.i
aesc
or
e.c
om/i
nd
ex
.ph
p/IJ
ECE
Sim
ulation of LT
E
-
TDD i
n the
HAPS
cha
nn
el
Aws M
ajeed
Al_
A
w
ad
i
1
, M
ohammed
J
aw
ad
Al
_
Du
j
ai
li
2
1
Com
m
unic
at
i
on
and
Med
ia Com
m
ission Ba
ghda
d
,
Ir
aq
2
Depa
rt
m
ent
of Electronic and
C
om
m
unic
at
ion
,
Facul
t
y
of Engin
ee
ring
,
Univ
ersity
of
Kufa
,
Ir
aq
Art
ic
le
In
f
o
ABSTR
A
CT
Art
ic
le
history:
Re
cei
ved
J
un
20
, 2
019
Re
vised
Dec
6
,
2019
Accepte
d
Dec
14
, 20
19
LT
E
st
ands
for
Long
Te
rm
Evo
l
uti
on.
Thi
s
t
ec
hn
olog
y
enha
n
ce
s
t
he
data
ra
te
and
ca
p
ac
i
t
y
us
ing
a
new
rad
io
int
erf
a
ce
and
a
n
opti
m
iz
ed
cor
e
net
work
.
Thi
s
progre
ss
was
done
to
sati
sf
y
st
anda
rds
def
i
ned
for
th
e
fourt
h
gene
r
at
io
n
of
ce
ll
u
la
r
com
m
unic
at
ions
in
ITU.
LTE
has
two
t
y
p
es
of
tra
nsm
ission:
Freque
nc
y
Divi
sion
Duplex
(FD
D)
and
Ti
m
e
Division
Dupl
ex
(TDD).
Now
aday
s,
LTE
-
TDD
rap
idly
G
rows
and
ta
kes
pla
c
e
of
old
f
ix
ed
cellular
comm
unic
at
ions
,
li
k
e
W
iMAX
.
Another
upcoming
tech
nolog
y
in
the
comm
unic
ation
industr
y
is
High
Am
pli
tude
Plat
form
Sta
ti
o
ns
(HA
PS
).
Stud
y
ing
th
e
capabi
lit
y
of
HA
PS
as
a
base
stat
ion
for
LT
E
-
TD
D
is
the
m
ai
n
purpose
of
thi
s
pape
r.
Sim
ula
t
ions
have
done
using
HAPS
c
hanne
l
and
compare
d
to
S
ta
nford
Unive
rsit
y
Int
eri
m
(SU
I)
standa
rd
c
hanne
ls
fo
r
diffe
ren
t
sce
nar
i
os.
Result
s
were
compare
d
to
achie
ve
a
conc
lusi
on
on
HA
P
S
imp
le
m
ent
ation
for
LTE
-
TDD b
ase
d
on
B
ER an
d
data throughpu
t.
Ke
yw
or
d
s
:
HAPS
HA
R
Q
LTE
MIM
O
OFDM
Copyright
©
202
0
Instit
ut
e
o
f Ad
vanc
ed
Engi
n
ee
r
ing
and
S
cienc
e
.
Al
l
rights re
serv
ed
.
Corres
pond
in
g
Aut
h
or
:
Moh
am
m
ed
J
awad
Al
D
uj
ai
li
,
Dep
a
rt
m
ent o
f
Ele
ct
ro
nic
and
Com
m
un
ic
at
io
n
E
ngineeri
ng
,
Faculty
of E
ngineerin
g,
Un
i
ver
sit
y o
f Kufa,
Kufa,
Iraq
.
Em
a
il
:
Moh
am
m
ed.
chall
ab
@
uokufa
.edu.i
q
1.
INTROD
U
CTION
Nowa
days,
ci
ti
es
are
gro
wing
s
o
fast
a
nd
hence
bas
e
s
ta
ti
on
s
s
hould
be
re
pla
nted
to
m
ai
ntain
the
net
wor
k
c
ov
e
ra
ge.
Netw
ork
plan
ning
i
s
a
ti
m
e
-
con
sum
ing
pr
ocess
and
co
sts
a
lot
.
T
he
best
s
olu
ti
on
i
s
us
in
g
a b
ase
st
at
ion
that
ca
n
r
el
ocate
easi
ly
,
and
it
s
com
m
u
nicat
ion
ta
kes
a
m
ini
m
u
m
eff
ect
of
m
ulti
path
f
r
om
bu
il
di
ngs.
Be
c
ause
of
t
he
hig
h
al
ti
tud
e
o
f
HAPS,
it
ca
n
pro
vid
e
a
c
omm
un
ic
at
ion
li
nk
sim
i
la
r
to
sa
te
ll
it
es.
This
li
nk
is
res
ist
ant
to
horizo
ntal
m
ulti
path
and
m
akes
H
A
PS
a
go
od
al
te
rn
at
ive
for
fi
xe
d
la
nd
base
sta
ti
on
s.
Also
,
LTE
is
a
le
adin
g
te
ch
no
l
og
y
in
cel
lular
c
omm
un
i
cat
ion
wit
h
go
od
su
sta
ina
bili
ty
against
dest
ru
ct
iv
e
eff
ect
s
in
ra
dio
cha
nn
el
s
.
To
stud
y
the
pe
rfor
m
ance
of
LTE
on
HAP
S,
the
dif
fer
e
nt
sit
uation
sh
ould
be
si
m
ulate
d.
In
[
1]
MIM
O
com
m
un
ic
at
ion
wa
s
analy
zed
f
or
HAPS,
bu
t
thi
s
arti
cl
e
do
es
n
o
t
sim
ula
te
p
hysic
al
la
ye
r.
First
effor
t
f
or
im
p
lem
entat
ion
of
the
broa
dband
ph
ysi
cal
la
ye
r
fo
r
H
AP
S
w
as
do
ne
for
W
iM
AX
i
n
[2
-
4]
.
A
fter
presenti
ng
LTE
te
chnolo
gy
r
esearche
rs
sta
r
te
d
to
stu
dy
it
s
i
m
ple
m
enta
ti
on
on
HAPS
[5
-
7]
.
The
m
ai
n
si
m
ulati
on
in
a
rtic
le
s
stud
ie
s
F
DD
ty
pe
of
L
TE
ti
ll
tod
ay
.
Be
cause
of
th
e
good
perform
ance
of
TD
D
-
LTE
i
n
fixe
d
cel
lular
netw
orks,
this
paper
is
go
i
ng
to
sim
ulate
t
his
ty
pe
of
LT
E
for
HAPS.
I
n
the
nex
t
sect
io
n
L
TE
te
chnolo
gy
is
intro
duce
d.
The
th
ir
d
sect
ion
co
ns
ist
s
of
HA
P
S
intr
oduction
and
it
s
cha
nn
el
si
m
ulati
on
m
et
ho
ds.
I
n
the
four
t
h
sect
ion
a
br
ie
f
in
tro
du
ct
io
n
of
MIM
O
is
pre
sented
.
In
the
fi
fth
sec
ti
on
,
the
sim
ul
at
ion
scena
rio
is
exp
la
ine
d.
I
n
the
sixth
sect
ion
res
ults
with
com
plete
discuss
io
n
is pr
e
sente
d.
A
t t
he
en
d
c
oncl
us
io
n
is m
ade
on the
res
ults.
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
om
p
En
g
IS
S
N:
20
88
-
8708
Simulati
on
of
LTE
-
TDD i
n
th
e H
AP
S
c
hann
el
(A. M. Al
A
wad
i)
3153
2.
LTE
Lo
ng
Te
rm
E
vo
l
ution
(LTE
)
is
a
sta
nd
ar
d
f
or
wi
reless
broadba
nd
c
om
m
un
ic
at
ion
.
It
increases
the
capaci
ty
and
data
rate
by
us
ing
a
dif
f
eren
t
ra
dio
i
nt
erf
aces
t
og
et
he
r
with
c
or
e
ne
twork
im
pr
ov
e
m
ents.
In
the
rad
i
o
i
nterf
ace,
LTE
us
es
O
rth
ogonal
Fr
e
qu
e
ncy
Divi
sion
M
ulti
plexing
(OFDM)
.
This
is
a
m
et
ho
d
of
encodin
g
dig
it
al
data
on
m
ulti
ple
carrier
fr
e
qu
e
ncies
without
any
f
reque
nc
y
gu
a
rd
ba
nd.
Also
,
unli
ke
UMTS
that
us
es
fixe
d
5
MHz
ba
ndwidth,
LTE
ca
n
util
iz
e
a
va
riou
s
ra
ng
e
of
ba
ndwidt
h
f
ro
m
1.4
MHz
t
o
20
MH
z
.
This
al
lows
op
erators
to
fle
xib
il
it
y
to
cho
ose
the
desire
d
c
onfig
ur
at
io
n
to
dep
e
nd
on
the
ir
ded
ic
at
e
d
spe
ct
ru
m
and
data
tra
ff
i
c.
LTE
has
tw
o
data
tra
ns
m
i
ssion
te
c
hnologies,
F
reque
nc
y
Divi
sio
n
D
uple
x
(FDD
)
an
d
Tim
e
Divisio
n
Dupl
ex
(
TD
D).
Whi
le
LTE
-
F
DD
us
es
paire
d
f
re
qu
e
ncies
t
o
up
load
a
nd
download
data,
L
TE
-
TD
D
us
es
a
sin
gle
fr
e
qu
e
ncy,
al
te
rn
at
in
g
betw
een
uploa
ding
an
d
dow
nlo
a
ding
data
th
r
ough
ti
m
e.
The
rati
o
betwee
n
uploa
ds
an
d
do
wn
l
oads
o
n
an
L
TE
-
T
DD
netw
ork
can
be
ch
ang
e
d
dynam
i
cal
ly
,
dep
en
di
ng
on
wh
et
her
m
or
e
data
nee
ds
to
be
sent
or
recei
ve
d.
Be
ca
us
e
of
these
a
dvanta
ges,
now
adays
LTE
-
T
D
D
is
use
d
as
a
fixed
cel
lula
r
com
m
un
ic
at
i
on
s
olu
ti
on
in
m
any
countrie
s.
This
te
ch
nolog
y
is
the
be
st
ch
oice
f
or
rea
chin
g
a n
e
w
e
ra
of
c
om
m
un
ic
at
ion
[
8,
9]
.
3.
HAP
S
High
-
Alti
tud
e
Plat
fo
rm
Stat
io
n
or
High
-
Alti
tud
e
Pse
udo
-
Sa
te
ll
it
e
(H
APS)
is
def
ine
d
as
"a
sta
ti
on
on
an
ob
j
ect
at
an
al
ti
tud
e
of
20
to
50
km
and
at
a
sp
eci
fie
d,
no
m
inal,
fixe
d
point
relat
iv
e
to
the
Ea
rth"
.
This
platfo
rm
is
ab
le
to
deliver
hi
gh
-
sp
ee
d
c
onnect
ivit
y
to
use
rs,
ov
e
r
area
s
of
up
t
o
40
0
km
.
It
has
gaine
d
sign
ific
a
nt
inte
rest
beca
us
e
HAPS
will
be
able
to
delive
r
ba
ndwi
dth
a
nd
ca
pacit
y
sim
il
ar
to
a
bro
adb
a
nd
wireless
acce
s
s
netw
ork
(su
c
h
as
W
iM
AX)
wh
il
e
prov
i
din
g
a
co
ver
a
ge
area
sim
i
la
r
to
that
of
a
s
at
el
li
te
.
It
can
pr
ov
i
de
Line
of
sig
ht
com
m
un
ic
at
ion
sim
i
la
r
to
G
EO
sat
el
li
te
s,
wh
ic
h
gua
ran
t
eed
lo
w
la
te
nc
y
and
m
ini
m
u
m
path
loss
eff
ect
s.
Also
,
m
ulti
ple
HAPS
platf
orm
s
can
be
synchro
nized
to
f
or
m
a
cellular
netw
ork
and
ta
ke
ad
va
ntage
of
fr
e
qu
ency
re
us
e.
H
AP
S
is
m
or
e
c
os
t
ef
fici
ent
th
an
a
com
m
un
ic
at
ion
sat
el
li
te
and
is
m
or
e
flexible
t
han
a
fixe
d
la
nd
base
sta
ti
on
,
with
al
s
o
a
bet
te
r
co
ve
rag
e
.
C
on
si
der
i
ng
thes
e
be
nef
it
s,
H
A
PS
is
a f
utu
re
step
in
providi
ng f
ixe
d
cel
lular
co
m
m
un
ic
at
ion
[
10]
.
HAPS
c
hannel
act
s
li
ke
sat
ellit
e
chan
nel.
T
his
cha
nnel
ha
s
a
li
ne
of
si
ght,
so
t
ran
sm
itted
powe
r
at
te
nu
at
ed
pro
portio
nally
to
the
s
qu
a
re
of
the
distance
.
Mostl
y,
fa
di
ng
in
sat
el
li
te
an
d
HAPS
c
hannel
descr
i
bes
us
in
g
Ri
cean
distr
ibu
ti
on
f
unct
ion,
bu
t
in
de
ns
e
a
reas
li
ke
la
rg
e
ci
ti
es,
Ra
yl
ei
gh
distr
ibu
ti
on
functi
on is
pr
e
f
err
e
d.
The
re a
r
e f
our
m
ajo
r
cha
nn
el
s
m
od
el
ing
for HA
PS:
‐
Pr
act
ic
al
m
od
el
(
desc
ribe
d
i
n
[
11
-
14]
)
‐
Geo
m
et
rical
m
od
el
(
descr
i
bed in
[
1, 15,
16]
)
‐
Stat
ist
ic
al
m
od
el
(
desc
ribe
d
i
n
[
2, 3]
)
‐
Mult
ist
at
e
m
od
el
(
desc
ribe
d
i
n
[4,
17
-
19]
)
Most
of
t
he
pr
ese
nted
c
ha
nn
el
s
for
H
A
PS
are
m
od
el
le
d
with
the
narrow
band
assum
pti
on.
This
m
eans
these
m
od
el
s
do
no
t
co
ns
i
der
i
ncohe
ren
t
ba
ndwi
dth
f
or
bro
adb
a
nd
com
m
un
ic
at
io
n.
Be
c
ause
of
tim
e
-
var
ia
nt
na
ture
in
real
c
hannels,
broa
dban
d
c
ha
nn
el
s
m
od
el
us
i
ng
Tap
ped
Delay
Line
(T
DL
).
In
t
his
m
od
el
,
each
ta
p
ha
s
di
ff
e
ren
t
weig
hts
an
d
di
stribu
ti
ons.
A
r
ti
cl
es
[20
,
21
]
us
e
d
a
T
DL
m
od
el
for
sim
ulati
ng
HAPS c
ha
nn
el
. T
his p
a
pe
r us
es HAPS
m
od
el
s p
re
sente
d
in
these arti
cl
es.
4.
MI
MO
Mult
iple Inpu
t
and
M
ulti
ple Outp
ut (
MIM
O)
is a
m
et
ho
d
f
or
m
ulti
plyi
ng
the cap
aci
ty
o
f
a
rad
i
o
li
nk,
us
in
g
m
ulti
ple
transm
issi
on
s
and
rec
ei
ving
an
te
nnas
to
ex
plo
it
m
ulti
path
pr
opa
gatio
n.
Mult
ipath
pro
pa
gati
on
of
sig
nal
is
t
he
m
ai
n
lim
i
t
er
f
or
high
ra
te
data
tra
nsm
issi
on
.
Re
ce
ived
sig
nal
al
so
s
uffe
rs
fro
m
oth
er
ph
e
nom
ena like f
adi
ng, Co
-
Chan
nel Inte
r
fere
nce (
CC
I
),
a
nd Inter
-
Sym
bo
l In
te
r
fer
e
nce
(ISI).
Wh
e
n
the
r
e isn'
t
enou
gh
kn
ow
l
edg
e
on
the
c
hannel
in
t
he
transm
itter
or
channel
that
va
ries
ve
ry
fast
,
the
best
so
l
ut
ion
is
div
e
rsity
cod
in
g
pro
vid
e
d
by
MIM
O.
I
n
dive
rsity
m
e
tho
ds,
a
sing
le
stream
(u
nlike
m
ult
iple
strea
m
s
in
sp
at
ia
l
m
ul
ti
plexing)
i
s
transm
it
te
d,
bu
t
t
he
sig
nal
i
s
co
ded
us
i
ng
te
chn
i
qu
e
s
cal
le
d
s
pace
-
ti
m
e
cod
i
ng.
T
he
si
gn
al
is
e
m
itted
from
each
of
the
tra
ns
m
it
antenn
as
with
fu
ll
or
ne
ar
or
t
hogonal
co
ding.
Dive
r
sit
y
cod
ing
e
xploits
the
ind
e
pende
nt
fad
i
ng
in
t
he
m
ult
iple
antenn
a
li
nks
to
enh
a
nc
e
sig
na
l
div
ersit
y.
A
s
Pr
op
os
ed
by
Vah
id
Tarok
h,
Nam
bi
Sesh
a
dr
i
,
an
d
Ro
be
rt
Ca
ld
erb
a
nk,
th
ese
sp
ace
-
ti
m
e
cod
es
(S
TCs
)
ac
hieve
si
gn
i
ficant
er
ror
rate
im
pr
ov
em
ents
ov
e
r
si
ngle
-
anten
na
syst
e
m
s.
Their
ori
gin
al
s
chem
e
was
base
d
on
trel
li
s
c
od
es
but
the
si
m
pler
bloc
k
co
des
we
re
util
iz
ed
by
Siavas
h
Alam
ou
ti
to
dev
el
op
spa
ce
-
tim
e
blo
ck
-
cod
e
s
(S
TBC
s)
[22]
.
In
Alam
ou
ti
cod
i
ng,
tw
o
c
onsec
utive
sym
bo
ls
a
re
c
od
e
d
co
ded
i
n
2
by
2
or
t
hogonal
m
at
rix.
Each
c
olu
m
n
sent
f
ro
m
two
MIM
O
a
nten
na
s
at
each
burs
t
of
tra
ns
m
issio
n.
I
n
the
recei
ver,
sig
nals
gat
her
from
two
MIM
O
anten
nas
an
d
c
om
bin
e
to
const
ru
ct
sent
sym
bo
ls
.
In
t
his
pa
per,
Alam
ou
ti
cod
i
ng
is
im
pl
e
m
ented
us
in
g
2
by
2
MI
MO sc
hem
e
.
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
om
p
En
g,
V
ol.
10
, No
.
3
,
J
une
2020
:
3
4
5
2
-
3
1
5
7
3154
5.
SIMULATI
O
N
S
CEN
A
RI
O
The
stim
u
la
te
d
al
go
rithm
is
s
how
n
in
Figur
e
1.
At
the
first
ste
p
of
al
gorithm
ran
do
m
dat
a
pr
od
uced.
The
n
data
ar
e
div
ide
d
int
o
two
st
ream
s
for
each
MIM
O
ante
nna.
A
fter
that
each
data
e
ncode
d.
In
the
e
ncodin
g
ste
p,
fi
rst
of
a
ll
,
Cy
clic
Re
du
nda
ncy
Chec
k
(CRC
)
ad
de
d
to
data
stream
as
an
error
-
det
ect
ing
cod
e
,
an
d
the
n
Re
ed
So
lom
on
co
des
ad
de
d
to
data
as
Fo
r
ward
-
Er
ror
-
C
orrecti
on
(F
EC
)
cod
in
g.
At
le
ast
data
cod
e
d
us
i
ng
V
it
erb
i
an
d
the
n
m
od
ulate
d
wi
th
de
fine
d
m
od
ulati
on
orde
r.
The
n
i
n
OFD
M
m
od
ulator
blo
c
k,
data
tra
ns
f
or
m
ed
to
the
ti
m
e
do
m
ai
n
us
i
ng
I
nv
e
rse
Fast
F
ourier
Tra
nsfo
r
m
(I
FFT)
.
A
fte
r
that
guar
d
ba
nd
an
d
a
cy
cl
ic
pr
efix
(CP)
a
dd
e
d
to
m
od
ulate
d
dat
a.
At
this
le
vel,
Alta
m
on
te
blo
ck
queues
t
he
se
sy
m
bo
ls
accor
ding
to
STBC
and
TDD
sc
hem
e
f
or
tra
ns
m
issi
o
n.
Pr
e
pa
red
dat
a
transm
i
tt
ed
t
hro
ugh
tim
e
-
var
yi
ng
H
APS
channe
l
.
Param
et
ers
of
this
cha
nn
el
ca
n
be
def
i
ne
d
by
the
us
er
.
I
n
this
pa
per,
just
the
dow
nlin
k
s
i
m
ulate
d
with
TD
D
up
li
nk/d
ownlin
k
co
nfi
gurati
on
ty
pe
1.
It
m
eans
half
of
t
he
tim
e
date
sent
from
HA
P
S
towa
rd
a
rece
iver
on
the
gr
ound,
a
nd
at
the
rest
of
the
ti
m
e,
ACK/N
ACK
data
sent
bac
k
to
the
HAPS.
At
t
he
receiver
sid
e,
data
cat
ch
us
i
ng
Al
tam
on
te
receiv
er.
T
his
recei
ve
r
com
bin
es
th
e
sign
al
f
r
om
t
wo
c
onsec
utiv
e
receives
a
nd
al
so
trie
s
to
om
it
c
hannel
ef
fects
from
i
t.
In
the
OFDM
dem
od
ulato
r
gu
a
rd
band
a
nd
C
P
e
lim
inate
d.
Af
t
er
that
sign
al
br
i
ng
ba
ck
to
t
he
f
requ
ency
dom
ai
n
us
ing
Fast
Fou
r
ie
r
Tra
ns
f
or
m
(F
FT
).
In
t
he
decode
r,
blo
c
k
data
extract
us
i
ng
a
Vite
rb
i
decod
er
an
d
c
orrecte
d
us
i
ng
Re
e
d
So
lom
on
c
od
e
s.
A
fter
that
,
C
RC
cod
es
c
hec
ked
t
o
see
receive
d
da
ta
is
valid
or
no
t.
At
t
his
le
ve
l,
if
we
ha
ve
no
er
ror
in
dat
a
ACK
sent
ba
ck
to
the
tra
nsm
itter
to
sen
d
the
ne
xt
data.
If
the
da
ta
has
a
n
er
ror,
N
ACK
sent
bac
k
t
o
the
t
ran
sm
it
te
r
to
inf
or
m
it
for
s
end
i
ng
the
sa
m
e
data.
Af
te
r
re
sen
ding
data
HA
R
Q
com
bin
es
two
new
recei
ve
d
sign
al
s
with
the
old
er
one
to
in
creas
e
SN
R.
T
his
act
ion
rep
eat
e
d
f
or
4
ti
m
es
if
an
er
ror
do
es
no
t
el
im
inate
.
The
nu
m
ber
of
su
cce
ssive
pa
ck
of
receive
d
data
com
par
e
to
t
he
total
sent
da
ta
stream
fo
r
cal
culat
ing
t
hroug
hput.
At
la
st
the
receive
d
data
check
e
d wit
h r
aw dat
a at t
he
tran
sm
it
te
r
to calc
ulate
BER
.
Figure
1
.
Stim
ulate
d
al
gorith
m
6.
SIMULATI
O
N RESULT
First
of
al
l,
it
’
s
bette
r
t
o
fin
d
the
be
st
set
ti
ng
s
f
or
com
par
in
g
HA
P
S
c
hannel
with
S
UI
c
ha
nnel
s
.
Modula
ti
on
is
the
fi
rst
c
on
ce
r
n.
So,
sim
ulatio
n
w
as
done
usi
ng
5
MHz
as
the
l
ow
e
st
co
m
m
ercial
ban
dwidt
h,
cy
cl
e
pr
efix
e
qu
al
to
0.1
25
and
var
io
us
m
odulati
on
s.
Fig
ur
e
2
s
hows
th
e
resu
lt
of
thr
oughpu
t
again
st
SN
R
and Fig
ure
3
s
hows
a
BER
c
om
par
ison agai
ns
t S
NR.
Figure
2
.
Thr
ough
pu
t a
gain
st
SN
R
Figure
3
.
BER
com
par
ison ag
ai
ns
t S
NR
G
e
n
e
r
a
t
e
D
a
t
a
E
n
c
o
d
e
r
O
F
D
M
M
o
d
u
l
a
t
o
r
E
n
c
o
d
e
r
O
F
D
M
M
o
d
u
l
a
t
o
r
A
l
a
m
o
u
t
i
T
r
a
n
s
m
i
t
e
r
H
A
P
S
C
h
a
n
n
e
l
A
l
a
m
o
u
t
i
R
e
c
i
e
v
e
r
O
F
D
M
D
e
m
o
d
u
l
a
t
o
r
O
F
D
M
D
e
m
o
d
u
l
a
t
o
r
D
e
c
o
d
e
r
D
e
c
o
d
e
r
E
r
r
o
r
D
e
t
e
c
t
i
o
n
R
e
c
e
i
v
e
d
D
a
t
a
C
a
l
c
u
l
a
t
e
B
E
R
C
a
l
c
u
l
a
t
e
T
h
r
o
u
g
h
p
u
t
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
om
p
En
g
IS
S
N:
20
88
-
8708
Simulati
on
of
LTE
-
TDD i
n
th
e H
AP
S
c
hann
el
(A. M. Al
A
wad
i)
3155
As
it
can
be
seen
in
t
he
lo
w
S
NR
Q
PS
K
ha
s
bette
r
perform
ance
than
t
he
oth
e
r
m
odulati
on
s
.
In
th
e
SN
R
rate
higher
t
han
4
dB,
BPS
K
s
hows
a
bette
r
re
su
lt
.
Be
cause
of
the
hi
gh
al
ti
tud
e
of
H
AP
S
,
it
can
be
ex
pected th
at
the r
ecei
ved
sign
al
h
a
s a low
SN
R
. S
o QP
SK
m
od
ulati
on
selec
te
d
for
th
e rest o
f
sim
ul
at
ion
s
.
Nex
t
sim
ulati
on
w
as
do
ne
to
choose
t
he
bes
t
cy
cl
ic
pr
efix
.
Pr
ope
rtie
s
are
set
to
5
M
Hz
f
or
ba
ndwidt
h,
QP
S
K
as
m
od
ulat
io
n
and
t
he
CP
is
var
ia
ble.
Fig
ure
4
s
hows
t
he
resu
lt
of
th
r
oughput
agai
nst
SN
R
an
d
Fi
gur
e
5
sh
ows
a BER c
om
par
ison agai
ns
t S
NR.
Figure
4
.
Thr
ough
pu
t a
gain
st
SN
R
Figure
5
.
A
BE
R com
par
iso
n against
SN
R
As
ca
n be see
n, CP
of
0.2
5 ha
s the
best r
es
ul
t, but hi
gh
e
r
C
P m
eans w
e s
houl
d
lose
a
gr
e
at
p
ort
io
n
of
tim
e
fo
r
sen
ding
re
dunda
nt
da
ta
.
CP
of
0.12
5
has
the sam
e
ou
tc
om
e
as
a
0.25,
s
o
we
c
hoose
it
as
the
m
a
i
n
CP
for
the r
est
of the
sim
ulatio
ns
.
Nex
t, sim
ulatio
n was do
ne
to
ev
al
uate b
a
ndwidth
e
ff
ect
s. I
n
this sim
ulati
on
C
P
equ
al
to
0.1
25
,
QP
S
K
as
m
o
du
la
ti
on
an
d
ba
ndwidt
h
is
var
ia
ble.
Fig
ur
e
6
sh
ows
the
resu
lt
of
th
r
oughput
against
SN
R a
nd Fig
ure
7
s
hows
a BER c
om
par
ison
a
gain
st SNR.
Figure
6
.
T
hr
ough
pu
t a
gain
st
SN
R
Figure
7
.
BER
com
par
ison ag
ai
ns
t S
NR
As
it
can
be
predict
ed,
a
lowe
r
band
width
is
m
or
e
resist
ant
to
the
no
ise
of
the
channel.
S
o,
5
MHz
of
band
width
is
t
he
be
st
opti
on
for
se
nd
i
ng
dat
a
in
the
HAPS
channel.
In
t
he
nex
t
ste
p,
by
reachi
ng
t
o
th
e
best
op
ti
on,
the
si
m
ula
t
ion
was
done
to
com
par
e
the
HAPS
channel
with
S
UI
cha
nnel
s.
S
UI
cha
nnel
s
add
verb
com
plete
ly
in
[23
,
24
]
.
SUI
-
1
&
S
UI
-
2
defi
nes
m
os
tl
y
flat
te
rr
ai
n
wit
h
l
igh
t
tree
den
s
it
ie
s.
It
c
orres
po
nd
s
to
m
ini
m
u
m
p
at
h
loss cond
it
io
ns com
fo
rt f
or subur
ban
are
as.
SU
I
-
2
has
m
or
e d
el
ay
sp
rea
d
than
S
UI
-
1. S
UI
-
3 &
SU
I
-
4
de
fines
hill
y
env
iro
nme
nt
bu
t
ra
re
ve
getat
ion
, o
r
high
ve
getat
ion
but
flat
te
rr
ai
n.
I
nterm
ediat
e
path
loss
conditi
on
is
ty
pical
of
this
ca
te
gory
an
d
ca
n
m
od
el
urban
area
f
or
sm
al
l
to
m
ediu
m
ci
ties.
A
gain
S
UI
-
4
has
m
or
e
delay
spread
t
han
SUI
-
3.
S
U
I
-
5
&
S
UI
-
6
de
fines
hi
ll
y
te
rr
ai
n
with
m
od
erate
-
to
-
heav
y
tree
de
nsi
ti
es,
wh
ic
h
res
ults
in
the
m
axi
m
um
path
loss.
It
us
ed
t
o
m
od
el
m
et
ro
po
li
s
with
la
rg
e
bu
il
di
ng
s
.
S
UI
-
6
has
m
or
e
delay
sp
r
ead
t
ha
n
S
UI
-
5.
F
or
these
sim
ulati
o
n
pr
op
e
rtie
s
se
t
to
the
ba
ndwi
dth
of
5
M
Hz,
the
cy
cl
ic
pr
efi
x
of
0.
125
a
nd
m
odulati
on
of
QPS
K.
Fi
gure
8
shows
t
he
re
su
lt
for
th
rou
ghput
against
S
NR
and
Fig
ur
e
9
s
ho
ws
a BER
co
m
pari
so
n agai
ns
t S
NR
[25
]
.
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
om
p
En
g,
V
ol.
10
, No
.
3
,
J
une
2020
:
3
4
5
2
-
3
1
5
7
3156
Figure
8
.
Thr
ough
pu
t a
gain
st
SN
R
Figure
9
.
BER
com
par
ison ag
ai
ns
t S
NR
As
it
can
be
s
een,
the
H
APS
channel
has
a
si
m
i
la
r
beh
a
vior
as
the
SUI
-
3
&
SUI
-
4
a
nd
act
away
bette
r
tha
n
SUI
-
5
&
S
UI
-
6.
I
t
m
eans
us
i
ng
HAPS
for
T
D
D
-
L
TE
in
the
urba
n
area
is
a
good
decisi
on
.
As
it
m
entioned
SUI
-
3
ha
s
lo
w
s
pr
ea
ding
delay
&
SUI
-
4
has
a
m
ediu
m
sp
read
i
ng
delay
.
HAPS
has
a
bette
r
perform
ance
than
SUI
-
4
un
de
r
SN
R
rate
eq
ual
to
0
dB,
a
nd
al
so
ha
s
a
be
tt
er
per
f
orm
ance
than
S
UI
-
4
unde
r
SN
R
rate
e
qua
l
to
-
3
dB.
It
s
hows
t
hat
in
the
en
vir
onm
ent
with
hi
gh
no
ise
,
or
c
on
diti
on
s
that
can
r
ecei
ve
enou
gh
po
we
r
in
the
recei
ver
po
i
nt,
H
APS
has
bette
r
pe
rfor
m
ance
f
or
th
e
urba
n
area
.
I
n
the
s
uburb
an
area
s
fixe
d
la
nd
net
work
act
bette
r
than
the
H
APS
de
pe
nds
on
the
resu
lt
for
S
UI
-
1
&
SUI
-
2.
At
the
ne
xt
se
ct
ion,
a m
or
e com
pr
e
hensi
ve
c
oncl
usi
on takes
f
or
t
his c
om
pa
rison.
7.
CONCL
US
I
O
N
In
t
he
m
et
ro
poli
s an
d hil
ly
terrain wit
h
m
od
e
rate
-
to
-
hea
vy t
ree d
e
ns
it
ie
s, HA
PS
is t
he
best
ch
oice
for
a
cel
lular
netw
ork.
In
t
he
ur
ba
n
area
with
m
edium
bu
il
ding
den
sit
ie
s,
if
t
he
fixe
d
la
nd
ne
twork
ca
n’
t
prov
i
de
enou
gh
po
wer,
H
AP
S
ha
s
a
be
tt
er
pe
rfor
m
ance
i
n
the
lo
w
SN
R
rate.
J
us
t
in
the
flat
a
rea
with
l
ow
de
ns
i
ti
es
of
trees
and
bu
il
ding
li
ke
suburba
n,
the
fi
xe
d
la
nd
netw
ork
pe
rfor
m
s
bette
r
than
H
APS.
H
ow
e
ver,
in
these
reg
i
on
s
,
HAPS
has
oth
e
r
s
upe
rior
it
y.
HAPS
ha
s
c
overa
ge
of
ab
out
40
0
km,
w
hile
m
axi
m
um
co
ver
a
ge
of
la
nd
base
sta
ti
on
is
lim
it
ed
to
70
km
in
the
i
deal
conditi
on.
So,
it
can
be
c
on
cl
ud
e
d
t
hat
HAP
S
can
s
uccessi
vely
su
pe
rse
de
t
he
f
ixed
la
nd cell
ul
ar n
et
work f
or TDD
-
LTE tec
hnology.
REFERE
NCE
S
[1]
E.
T
.
Mich
ailidi
s
and
A.
G.
Kan
at
as,
"S
ta
ti
st
ical
sim
ula
ti
on
m
o
del
ing
of
3
-
D
HA
P
-
MI
MO
cha
nnel
s,"
W
ireless
Pe
rs
onal
Comm
unic
ati
ons
,
vol
.
65,
no
.
4
,
pp
.
83
3
-
841,
2012
.
[2]
I.
-
R.
Pa
lma
-
L
a
z
gar
e,
J.
-
A.
Delg
ado
-
Penin,
F
.
Pere
z
-
Font
an,
"A
n
adva
nc
e
in
wi
rel
ess
broa
db
an
d
comm
unic
at
io
ns
bas
ed
on
a
W
i
MA
X
-
H
AP
S
arc
hitect
ur
e,
"
in
26th
int
ernati
on
al
communic
ations
satel
li
te
sy
stems
conf
ere
nc
e
(
ICSSC
)
,
2008.
[3]
E.
T
.
Mich
ai
l
idis
and
A.
G.
Kan
at
as,
"
Thre
e
-
d
imensional
HA
P
-
MIM
O
cha
nnel
s:
Modeli
ng
and
ana
l
y
sis
of
spac
e
-
ti
m
e
cor
r
elati
on
,
"
IEE
E
Tr
ansac
t
ions o
n
V
ehicula
r Tec
hnology
,
vo
l.
59
,
no
.
5
,
pp
.
2
232
-
2242,
2010
.
[4]
D.
R.
Putro,
"
P
erf
orm
anc
e
ev
al
ua
ti
on
of
bro
adba
nd
W
iMAX
s
erv
ic
es
ove
r
high
al
t
it
ude
pla
tforms
(HA
P
s)
Com
m
unic
at
ion
Channe
l,
"
IC
WMC'08,
The
F
ourth
Inte
rnational
Confer
enc
e
on
Wire
le
ss
a
nd
Mobile
Comm
unic
ati
ons
,
IE
EE
,
pp
.
55
-
5
9,
2008
[5]
S.
Gratsia
and
M.
Erna
wan,
"LT
E
upli
nk
c
el
lular
ca
p
acit
y
ana
l
y
s
is
in
a
high
al
ti
tud
e
pla
tforms
(HA
P
s
)
comm
unic
at
ion,"
in
11th
Inte
r
nati
onal
Conf
ere
nce
on
Te
leco
mm
unic
ati
on
Sy
stems
Serv
ic
es
and
Appl
i
cat
ion
s
(
TSSA
)
,
IEE
E
,
p
p.
1
-
5
,
2017
.
[6]
M.
R.
K.
Aziz,
"Channe
l
est
imation
for
LT
E
downlink
in
high
a
lt
i
tude
platform
s
(HA
Ps
)
sy
stems
,
"
in
Inte
rnationa
l
Confe
renc
e
of
In
formation
and
C
omm
unic
ati
on
T
ec
hnolog
y
(
ICoICT)
,
IEE
E
,
pp.
1
82
-
186,
2013
.
[7]
M.
Aziz,
"A
stu
d
y
of
HA
PS
-
LT
E
downlink
cha
n
nel
p
erf
orm
anc
e
sim
ula
ti
on
dep
lo
y
ed
for
h
igh
spe
ed
user
v
ehi
c
le,"
in
8th
Inte
rnat
io
nal
Conf
ere
nce
on
Telecomm
unic
ati
on
Syste
ms
S
erv
ices and Appl
ic
ati
ons
,
IE
EE
,
pp.
1
-
5
,
2014
.
[8]
S.
Sesia,
M.
B
a
ker
,
and
I
.
Toufik,
"
LTE
-
th
e
U
MTS
long
t
erm
evo
lut
ion
:
from
the
or
y
to
pr
ac
t
ice,
"
John
W
ile
y
&
Sons
,
2011.
[9]
B.
Furht
and
S.
A.
Ahs
on,
"
Lon
g
Te
rm
Evol
ut
io
n:
3GP
P L
TE radio
and
c
el
lu
la
r
t
ec
hnolo
g
y
,
"
Crc
Press
,
2016.
[10]
X.
Cao,
P.
Ya
ng,
M.
Al
ze
n
ad
,
X.
Xi
,
D.
W
u,
and
H.
Yan
i
ko
m
ero
glu,
"A
i
rborne
comm
unic
a
ti
on
n
et
work
s:
A surve
y
,
"
IE
EE J
ournal
on
Se
lecte
d
Areas in Co
mm
unic
ati
ons
,
v
ol.
36
,
no
.
9
,
pp
.
1907
-
1926,
201
8.
[11]
L.
Han
zo,
Y.
(J
os)
Akhtm
an,
L
.
W
ang
,
and
M.
Jiang
,
"
MIM
O
-
OF
DM
for
LT
E,
W
i
-
Fi
and
W
iMAX
,
"
ed:
John
W
il
e
y
and
S
ons
Lt
d
.
,
Uni
te
d
Kin
gdom
,
2011.
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
om
p
En
g
IS
S
N:
20
88
-
8708
Simulati
on
of
LTE
-
TDD i
n
th
e H
AP
S
c
hann
el
(A. M. Al
A
wad
i)
3157
[12]
H.
S.
W
ang
an
d
N.
Moay
eri,
"F
ini
te
-
sta
te
Mar
kov
cha
nnel
-
a
useful
m
odel
for
rad
io
comm
uni
ca
t
ion
cha
nne
ls,
"
IEE
E
Tr
ansacti
o
ns on
Ve
h
ic
u
lar
Technol
ogy
,
vo
l. 44,
no.
1,
pp.
16
3
-
171,
1995
.
[13]
J.
L.
Cu
eva
s
-
Ruí
z
and
J.
A
.
De
lg
ado
-
Penín,
"Ch
a
nnel
m
odel
ing
a
nd
sim
ula
ti
on
in
HA
PS
sy
st
ems
,
"
In
Proce
ed
ing
s
of
th
e
F
if
th
Euro
pean
Wire
le
ss
C
onfe
renc
e
, Spai
n,
pp
.
24
-
27
,
20
04.
[14]
A.
Im
ran
and
R.
Ta
fa
zol
l
i,
"P
erf
orm
anc
e
&
c
apaci
t
y
of
m
obil
e
b
roa
dband
W
iMAX
(802.
16
e)
depl
o
y
ed
vi
a
hig
h
al
titude
platform
,
"
in
Wireless
Co
nfe
renc
e,
IEEE
,
pp.
319
-
323
,
20
09.
[15]
J.
L.
Cueva
s
-
Ru
iz
and
J.
A.
Del
gado
-
Penín,
"A
stat
isti
ca
l
sw
it
ch
ed
broa
db
and
c
hanne
l
m
ode
l
fo
r
HA
PS
li
nks,"
in
WCNC
,
pp.
2
90
-
294,
2004
.
[16]
I.
R.
Pa
lma
-
Lázg
are
,
J.
A.
De
lga
d
o
-
Penín,
"W
iMAX
haps
-
base
d
downlink
per
for
m
anc
e
emplo
y
in
g
geometrica
l
an
d
stat
isti
ca
l
propa
g
at
ion
-
cha
nne
l ch
ara
c
te
rist
ic
s,"
URSI
Rad
io
S
ci
en
ce
Bul
l
et
in
,
vo
l.
2010,
no
.
333
,
p
p.
50
-
66
,
2010
.
[17]
A.
Gh
osh,
D.
R
.
W
olt
er,
J.
G.
Andrews
,
and
R
.
Chen,
"Broadb
and
wire
le
ss
ac
c
ess
with
W
i
Max/802.
16:
cur
r
ent
per
form
anc
e
ben
chmarks a
nd
fu
t
ure
pot
ent
i
al,"
I
EE
E
Comm
unications
Magazine
,
vol
.
43
,
no
.
2
,
p
p.
129
-
136
,
200
5.
[18]
C.
Ho
y
m
ann,
"A
naly
s
is
and
per
form
anc
e
e
val
u
ation
of
the
OF
D
M
-
base
d
m
et
ropoli
ta
n
ar
ea
ne
twork
IEE
E
802.
16
,
"
Computer
Net
wo
rks
,
vol
.
49
,
no
.
3,
pp
.
341
-
363
,
2005.
[19]
R.
Miura
and
M.
Oodo,
"W
ir
el
ess
comm
unic
at
ions
s
y
stem
using
strat
ospheri
c
pla
tforms
:
R
and
D
progra
m
on
te
l
ec
om
and
bro
adc
ast
ing
s
y
s
tem
using
high
a
ltitude
platform
st
at
ions,"
Journal
of
the
Comm
unic
ati
on
R
ese
arc
h
Labor
atory
,
vol. 48, pp. 33
-
48,
2
001.
[20]
Z.
L
ia
n,
L.
Ji
an
g,
and
C.
He
,
"
A
3
-
D
wideba
nd
m
odel
base
d
o
n
d
y
nami
c
evolution
of
sca
t
te
r
e
rs
fo
r
HA
P
-
MI
MO
cha
nne
l,
"
I
EEE
Comm
unic
ati
ons Let
t
ers
,
vol
.
21
,
no
.
3
,
pp
.
684
-
6
87,
2017
.
[21]
Y.
Ding,
Y.
Xi
a
o,
J.
Xie
,
and
T.
Zha
ng,
"A
ti
m
e
-
var
y
ing
tr
ansit
i
on
cha
nne
l
m
odel
for
ai
r
-
ground
comm
unic
at
ion
,
"
IEE
E
/A
IA
A
36
th
Digit
al
A
vi
oni
cs
Syste
ms
Conf
ere
nce (
DASC)
,
pp.
1
-
8
,
2017
.
[22]
S.
Alamouti,
P.
Poon,
and
V.
T
a
rokh,
"Low
complexi
t
y
m
ax
imum
li
kel
ihood
d
etec
t
ing
of
con
ca
t
ena
t
ed
spac
e
cod
es
for
wire
l
ess a
ppl
ic
a
ti
ons
,
"
ed:
Go
ogle
Pa
te
nts,
20
02.
[23]
R.
Jain
,
"Cha
nne
l
m
odel
s: A
tut
or
ia
l
,
"
in
WiMAX f
orum
AA
TG
,
pp.
1
-
6,
2007.
[24
]
Cueva
s
-
Ruíz,
Jo
sé
Lui
s,
and
Jo
sé
A.
Delg
ado
-
Penín.
"Chann
el
m
odel
ing
and
sim
ula
ti
on
in
HA
PS
sy
stems
,
"
European
Wire
l
ess 2004
,
pp
.
24
-
27
,
2004
.
[2
5
]
Jac
kson,
T
hom
a
s,
George
Ea
p
en
,
and
Fa
Dai
,
"L
ine
ar
iz
ed
of
fset
QP
SK
m
odula
ti
on
uti
lizing
a
si
gm
a
-
del
ta
b
ase
d
fre
quency
m
odu
l
at
or,
"
U.S.
Pa
te
n
t
No.
6,
975
,
687.
13
Dec
.
2005
.
BIOGR
AP
H
I
ES
OF
A
UTH
ORS
Aw
s
Maj
ee
d
G
halib
Alaw
adi
,
As
socia
te
Dire
ctor
of
the
comm
u
nic
a
ti
on
Regulat
or
y
Authori
t
y
,
Hea
d
of
al
l
oc
ation
fre
quency
aff
ai
rs
Depa
r
tment
Com
m
unic
at
io
n
and
Media
Comm
ission,
Baghda
d
,
Ira
q
.
E
m
ai
l:
A.m
ajee
d
@cm
c.
iq
Mohamme
d
Ja
w
ad
Al_Du
jail
i
AL_k
haz
ra
ji
awa
rd
ed
B.
S.
degr
e
e
in
comm
unic
at
ion
engi
ne
eri
ng
fro
m
Univer
sit
y
of
Al
-
Furat
Al
-
Aw
sat
T
ec
hni
cal,
T
ec
hni
ca
l
Coll
ege
of
Eng
ineeri
ng
,
Naja
f
,
Ira
q
in
2008
and
M.S.
degr
e
e
in
comm
unic
at
ion
s
y
st
em
engi
ne
eri
ng
from
Ferdows
i
unive
rsit
y
,
Ira
n
,
in
2018
.
Curr
e
ntly
,
h
e
is
a
m
e
m
ber
staff
at
th
e
Depa
rtmen
t
of
El
e
ct
roni
c
and
Com
m
unic
at
ion,
Facul
t
y
of
Enginee
ring
,
Univer
s
ity
of
Kufa
,
Ira
q
.
His
rese
ar
ch
in
te
rest
inc
lud
es
the
dev
el
opm
ent
of
W
ire
le
ss
comm
unic
at
ions
and
signal
proc
es
sing
as
well
as
image,
spee
c
h
proc
essing
and
r
ada
r,
5G.
Email:
Moham
m
ed.
challa
b@uokufa
.
ed
u.
iq
Evaluation Warning : The document was created with Spire.PDF for Python.