Content uploaded by Kazuki Maruta
Author content
All content in this area was uploaded by Kazuki Maruta on Mar 01, 2019
Content may be subject to copyright.
Optimized Adaptive Modulation with Considering
Mobile Relay on FSS-OFDM System
Shun Kojima†,Ken Teduka, Kazuki Maruta and Chang-Jun Ahn
Graduate School of Engineering, Chiba University
1–33, Yayoi-cho, Inage-ku, Chiba-shi, Chiba 263–8522, Japan
Email:pkuronek055@chiba-u.jp†
Abstract—This paper proposes an adaptive modulation on
frequency symbol spreading (FSS) based orthogonal frequency
division multiplexing (OFDM) relay system with considering the
mobility of relay node. Wireless communications often suffer
from large propagation loss due to shadowing and multipath
fading. As a solution to this problem, Relay communications,
which forward data messages from source to destination via
intermediate station(s), have been focused on. It can obtain
the space diversity and expand the area coverage. However,
here arises an another issue on relaying how to deploy relay
nodes in an efficient manner. In this paper, we present a whole
system design of FSS-OFDM employing an adaptive modulation
considering mobility of relay node. Exploiting a good match
of an adaptive modulation and FSS, throughput performance
can be significantly improved. Computer simulation verifies its
effectiveness and reveals the proposed system is the most valuable
means which realizes the flexible relay node deployment.
I. INT ROD UC TI ON
With the development of mobile communication systems,
the use of the Internet on smart phones and tablets is in-
creasing. For this reason, the realization of high-date-rate,
high-capacity and high-quality is required. Wireless com-
munications often suffer from its propagation characteristics
such as shadowing and multipath fading. Relaying, forwarding
data messages from source to destination via intermediate
station(s), has been widely investigated as one of the attracting
solutions to overcome the above drawback [1] [2]. It offers a
stable transmission to the destination terminals particularly in
a bad channel condition or in a place far-off from the serving
base station (BS). Thus it is effective in coverage expansion.
Relay communications can be classified into two schemes:
Amplify-and-Forward (AF) and Decode-and-Forward (DF).
AF is the relaying scheme that relay node forwards the
incoming signal only by amplifying in the analog domain,
so that forwarding delay is relatively low while additive noise
may also be enhanced [3]. In DF, relay node demodulates
and decodes the source message and transmits the re-encoded
message [4] [5]. Although DF incurs oppressed computation
complexity and delay, it can provide better communication
quality than AF. Another issue on relaying is practical de-
ployment of relay nodes (RNs). Additional deployment of
fixed RNs increases extra cost, especially in power supply
installation [6]. A possible solution is to install the relaying
function on user equipment or moving vehicles. In this case,
the system should be designed with considering their mobility
unlike the fixed relay node deployment scenario.
Adaptive modulation is capable of tracking to mobility
by adequately setting modulation and coding scheme (MCS)
according to the feedback channel state information (CSI) [7]
[8]. It is especially effective even in orthogonal frequency
division multiplexing (OFDM) where MCS can be optimally
determined per frequency component, i.e. subcarrier or sub-
channel. Meanwhile, it requires feedback information (FBI)
including MCS level for all frequency components. To reduce
such overhead, frequency symbol spreading (FSS) has been
proposed [9]– [11]. With FSS-OFDM, the transmission sym-
bols of each subcarrier are spread to all frequency components
by orthogonal spreading code. Power density of each symbol
originally mapped to the corresponding subcarrier is unified
even though they go through a frequency selective fading
channel. It means that FSS virtually provides additive white
Gaussian noise (AWGN) channel and the detected symbols
have the same signal-to-noise power ratio (SNR) at the re-
ceiver. Therefore, the same MCS level can be assigned for
all transmission subcarriers and it can improve bit error rate
(BER) performance [12]. Based on the above background,
the combination of adaptive modulation and FSS-OFDM is
expected to be effective in enhancing the transmission capacity
of mobile relay system.
This paper presents an improved method for throughput
performance of DF mobile relay system incorporated with
adaptive modulation and FSS-OFDM. The rest of the paper is
organized as follows. Section II describes the channel model.
Section III presents the proposed mobile relay system. Section
IV shows the computer simulation results. Finally, the paper
is concluded in Section V.
II. CH AN NE L MOD EL
This paper assumes multipath fading channel which is
expressed as,
h(τ, t) =
J−1
X
j=0
hj(t)δ(τ−τj),(1)
where hjis the complex channel coefficient and δindicates
the Dirac’s delta function. Jand τldenote the number of
discrete paths and the delay time, respectively. Here assumes
normalized path gain, i.e. PJ−1
j=0 Eh2
j= 1 where E[·]
stands for the expectation (ensemble average) operation. We
The 20th International Symposium on Wireless Personal Multimedia Communications (WPMC2017)
978-1-5386-2768-6/17/$31.00 ©2017 IEEE 41
Tx
+GIP/SIFFT
Input
Data Mod Mux
FSS
Pilot S/P
FSS
N
SF
Channel Estimation
Output
Data
P/S
MMSEC
Detection
MMSEC
Rx
-GI S/P FFT
N
SF
Channel Estimation
+GIP/SIFFT
Mod Mux
FSS
Pilot
S/P
FSS
N
SF
P/S
MMSEC
Detection
MMSEC
Rx
-GI S/P FFT
N
SF
Tx
(a) Source
(b) Relay
(c) Destination
Feedback
Information
Feedback
Information
Fig. 1. FSS-OFDM based mobile relay system with adaptive modulation.
can then obtain the frequency response H(f, t)via Fourier
transform of the impulse response.
H(f, t) = Z∞
0
h(τ, t)e−j2π fτ dτ
=
J−1
X
j=0
hj(τ, t)e−j2π fτ ,(2)
where fdenotes the carrier frequency. In a mobile commu-
nication environment, the frequency response is generally not
flat. J > 1provides frequency selective fading channel where
|H(f, t)|fluctuates in the transmission bandwidth. In this case,
FSS can artificially mitigate the effect of frequency selective
fading and improve the BER and throughput performance [12].
Path loss characteristics in this paper follows the Friis
transmission equation. Let Pt,Gt,Gr,λand ddenote
the transmission power, antenna gains of transmitter/receiver
antennas, wavelength and the propagation distance, reception
power Pris given by
Pr=λ
4πd 2
GrGtPt,(3)
From Eq. (3), it is obvious that the reception power mitigates
according to square of the distance d.
III. PROP OS ED FSS-OFDM BASED MOBILE RELAY
SYS TE M
In mobile relay communications, optimal MCS should be
determined with reduced FBI. We present a system design
based on FSS-OFDM to improve the BER and throughput
performances.
A. Source Node
Fig. 1(a) shows the block diagram of the source node
for FSS-OFDM based mobile relay system. The transmission
signal produced by FSS-OFDM is expressed as
s(t) =
Nd+Np−1
X
i=0
g(t−iT)·r2P
Nc
·"Nc−1
X
m=0
u(m, i)·exp{−j2π(t−iT )m/Ts}#,
(4)
where Nd,Npand Ncdenote the number of data symbols,
pilot symbols and subcarriers, respectively. Tindicates the
OFDM symbol duration including guard interval (GI) and Tsis
the effective symbol duration without GI. In OFDM, the inter-
symbol interference (ISI) caused by the multipath fading can
be avoided by inserting GI. When its length is Tg,T=Ts+Tg
is satisfied. Subcarrier spacing is 1/Ts.Pindicates the average
transmitting power. g(t)is the transmission pulse which is
given by
g(t) = 1−Tg≤t≤Ts
0otherwise.(5)
Defining OFDM signal of the i-th symbol at the m-th
subcarrier as u(m, i), it is expressed as,
u(m, i) =
NSF −1
X
k=0
ck(mmod NSF )
·d({m/NSF }NSF +k, i),
(6)
The 20th International Symposium on Wireless Personal Multimedia Communications (WPMC2017)
978-1-5386-2768-6/17/$31.00 ©2017 IEEE 42
Fig. 2. A concept of FSS.
where d(m, i)is the i-th modulated symbol at the m-th
subcarrier satisfying E[|d(m, i)|] = 1.ck(n)is the Hadamard
code to perform FSS and satisfies
NSF −1
X
n=0
ck(n)c∗
ω(n)=NSF for k=ω
0for k6=ω(7)
where |ck(n)|= 1 and (·)∗represent the complex conjugate.
A concept of FSS in terms of the power spectra is shown
in Fig. 2. Each modulation symbol d(m, i)is copied NSF
times and multiplied by orthogonal spreading code. These
spread symbols are combined to the same FSS block. On
that account, all data symbols are spread and superposed
over the transmission bandwidth. Although the power of each
input data symbol is divided by NSF subcarriers, receiver can
extract it by despreading with frequency diversity. It enables
for the spread data to obtain a frequency diversity without
additional increase of transmission power for each subcarrier.
It results in unified reception power, i.e. SNR, for each data
symbol.
On the source node as illustrated in Fig. 1(a), the data
stream is mapped to the Ncmodulation symbols by use of
FBI from the relay node. Then the pilot and data signals
are serially concatenated and serial-to-parallel (S/P) converted
to be allocated to subcarriers. Each symbol of subcarrier is
spread by the Hadamard code with length of NSF =Ncand
is superposed. Here assumes that symbol superposition with
the equal power ratio. The transmission signal in time domain
is obtained via inverse fast Fourier transform (IFFT) and GI
insertion. It goes through a broadband propagation channel.
B. Relay Node
Fig.1(b) shows the relay node structure which employs the
DF method. DF based relay node detects the received signals
rsr(m, i)as following manner,
ˆu(m, i) = rsr(m, i)·ωsr (m, i),(8)
where ω(m, i)is the FSS combining weight which will be
explained later in the next section. The detected signal ˆ
u(m, i)
is re-modulated and transmitted to the destination node.
C. Destination Node
Fig. 1(c) depicts the destination node structure. The received
signals are performed S/P conversion and GI removal. Nc
parallel sequences are then applied fast Fourier transform
(FFT). The converted signals in frequency domain is going
to be despread by spreading code. However, the orthogonality
among spread symbols is disordered due to the multipath
fading. To restore it, we use a minimum mean square error
combining (MMSEC) based frequency equalization combin-
ing.
The received signal is expressed as
r(t) = Z∞
−∞
h(τ, t)s(t−τ)dτ +n(t),(9)
where n(t)is the AWGN with power spectral density of N0.
Resolving r(t)into Ncsubcarriers via FFT, its frequency
domain expression ˜r(m, i)is give by
˜r(m, i) = 1
TsZiT +Ts
iT
r(t) exp{−j2π(t−iT )m/Ts}dt
=r2P
Nc
Nc−1
X
e=0
u(e, i)·1
TsZTs
0
exp{j2π(e−m)
·t/Ts} · Z∞
−∞
h(τ, t +iT )g(t−τ)
·exp(−2πeτ /Ts)dτ }dt + ˆn(m, i)
=r2P
Nc
H(m, i)u(m, i) + ˆn(m, i),
(10)
where ˆn(m, i)is AWGN component having a variance of
2N0/Ts. After separating pilot signals, frequency domain
equalization is applied and the demodulated data symbol is
expressed as
˜
d(m, i) =
NSF −1
X
k=0
ω(m, i)·˜r({m/NSF } · NSF +k, i)
·c∗
mmod NSF (k),
(11)
where ω(m, i)is the FSS combining weight. In this paper, we
use MMSEC that is expressed as
ω(m, i) = q2P
Nc·˜
H(m, i)
q2P
Nc·˜
H(m, i)
2
+ 2˜σ2
,(12)
where ˜
H(m, i)is the channel estimate of the i-th OFDM
symbol at the m-th subcarrier, ˜σ2is the noise variance per
subcarrier, which is assumed to be identically known for all
subcarrier.
As explained above, each subcarrier holds the modulated
symbols with equal power ratio. It indicates that all despread
The 20th International Symposium on Wireless Personal Multimedia Communications (WPMC2017)
978-1-5386-2768-6/17/$31.00 ©2017 IEEE 43
symbols exhibit the same SNR even though they experience
the frequency selective fading channel. Therefore, we can
apply the same MCS level for each subcarrier block; it is
enough to provide just one piece of FBI and MCS level
information (MLI) to control the MCS level, compared to
the conventional scheme. The proposed method can reduce
the overhead quantities of FBI and MLI transmission. From
these reasons, FSS can improve BER as well as throughput
performance.
D. Adaptive Modulation
In the proposed system, the adaptive modulation is per-
formed by the source nodes using link quality fed back by
the destination node. First, the destination node estimates the
reception power as a link quality observing CSI from the relay
node. Estimated results of reception power level is fed back
to the adaptive modulation decision block of the source node.
The source node then chooses the MCS level referring to the
predetermined threshold information.
IV. COM PU TE R SIM UL ATIO N
A. Simulation Parameter
Simulation condition assumes the 15 path Rayleigh fading
with exponential decay and their interval is Tpath = 50 ns.
This provides severe frequency selectivity. We also evaluated
two cases of mobility; the Doppler frequencies of 10 and
200 Hz, respectively. The path loss model is Friis free space
propagation. One OFDM symbol consists of 64 subcarriers
and the packet is structured by Np= 2 pilot and Nd= 20
data symbols. Adaptive modulation is ideally performed to
achieve the largest throughput. Here it simply employs two
MCS settings: QPSK and 16QAM. Table 1 summarizes the
detailed simulation parameters.
The simulation topology of the mobile relay system is
shown in Fig. 3. Suppose the positions of the source node and
the destination node are fixed, the position of the relay node
is determined at random within the circular region around the
TABLE I
SIM ULATI ON PAR AM ETE RS
Transmission scheme FSS-OFDM
Data Modulation QPSK,16QAM
FFT size 64
Number of subcarriers 64
Number of symbols 22 symbols (Np=2,Nd=20)
Guard interval 16
Doppler frequency 10 Hz, 200 Hz
Bandwidth 20 MHz
Fading 15 path Rayleigh fading
Path loss model Friis free space propagation
Transmission Power 2 W/MHz
Noise Power -90 dBm
Spreading code length 64
Relay
Destination
Source
100 m 100 m
10 m
Fig. 3. Simulation topology.
source node. The minimum distance between the relay and the
destination node is 10 m, and maximum is 200 m.
B. Simulation Results
Fig. 4 shows the BER performances of fixed QPSK and
16QAM, the conventional OFDM with adaptive modulation,
and the proposed FSS-OFDM with adaptive modulation at
Doppler frequencies of 10 and 200 Hz, respectively. Here,
the received power is defined as the average power from the
relay node to the destination node and it depends on the
position of the relay node. It can be confirmed that the adaptive
modulation is effective in improving BER performance. The
proposed system achieves the best BER performance since
it obtains the frequency diversity gain due to the FSS. On
the other hand, as Doppler frequency rises, BER performance
of both systems deteriorated and showed the error floor. For
this reason, the fluctuation of the channel states is too fast
to track for the receiver, hence it worsen the channel estima-
tion accuracy and distorted orthogonality between subcarriers.
Even in such high mobility situation, the proposed FSS-OFDM
relay system showed improved BER performance than the
conventional adaptive modulation system.
Fig. 5 shows the throughput performances of conventional
OFDM with adaptive modulation, fixed QPSK, 16QAM,
and the proposed FSS-OFDM with adaptive modulation, at
Doppler frequency of 10 and 200 Hz. As shown, the pro-
posed system significantly improves throughput performance
in comparison with the case where the modulation level was
fixed as well as the conventional adaptive modulation system.
As a result, throughput can be improved maximally by 180%
compared to the conventional OFDM adaptive modulation
system for Doppler frequency of 10 Hz at reception power of
-50 dBm. Furthermore, it should be noted that improvement by
100% (reception power is -50 dBm here) can also be attained
even in a high mobility situation such as with 200 Hz Doppler
frequency. We can conclude that the proposed FSS-OFDM
mobile relay system can be the most valuable solution which
can realize the flexible relay node deployment.
V. CO NC LU SI ON
This paper proposed an adaptive modulation on FSS-OFDM
relay system where considered mobility of relay node. FSS
The 20th International Symposium on Wireless Personal Multimedia Communications (WPMC2017)
978-1-5386-2768-6/17/$31.00 ©2017 IEEE 44
10
-5
10
-4
10
-3
10
-2
10
-1
10
0
-70 -65 -60 -55 -50 -45 -40 -35 -30
16QAM: 10Hz
QPSK: 10Hz
conventional adapt mod: 10Hz
proposed adapt mod: 10Hz
16QAM: 200Hz
QPSK: 200Hz
conventional adapt mod: 200Hz
proposed adapt mod: 200Hz
BER
Average Received Power [dBm]
Fig. 4. BER performance.
0
5
10
15
20
25
30
-70 -65 -60 -55 -50 -45 -40 -35 -30
16QAM: 10Hz
QPSK: 10Hz
conventional adapt mod: 10Hz
proposed adapt mod: 10Hz
16QAM: 200Hz
QPSK: 200Hz
conventional adapt mod: 200Hz
proposed adapt mod: 200Hz
Throughput [Mbps]
Average Received Power [dB]
Fig. 5. Throughput performance.
yields the same SNR for transmission symbols by frequency
spreading code. It allows to apply the same MCS level for all
symbols which can reduce the controlling overhead. Our pro-
posal exploiting good match of adaptive modulation and FSS-
OFDM can improve throughput performance of the mobile
relay system. Computer simulation clarified that the proposed
system improved throughput performance by 100% even in
a high mobility environment where Doppler frequency is
200Hz, in comparison with conventional OFDM with adaptive
modulation. The proposed adaptive FSS-OFDM mobile relay
system can be the most valuable solution which realizes the
flexible relay node deployment.
ACK NOW LE DG EM EN T
This work is supported by the Grant of Scientific Research
No.17K06415 from the Japan Society for the Promotion of
Science (JSPS).
REF ER EN CE S
[1] R. Pabst, B. H. Walke, D. C. Schultz, D. C. Herhold, H. Yanikomeroglu,
S. Mukherjee, H. Viswanathan, M. Lett, W. Zirwas, M. Dohler, H.
Aghvami, D. D. Falconer, and G. P. Fettweis, “Relay-based deployment
concepts for wireless and mobile broadband radio,” IEEE Commun. Mag.,
vol. 42, no. 9, pp. 80-89, Sep. 2004.
[2] B. Rankov and A. Wittneben,“Spectral efficient protocols for half-duplex
fading relay channels,” IEEE J. Select. Areas Commun., vol. 25, no. 2,
pp. 379-389, Feb. 2007.
[3] R. U. Nabar, H. Bolcskei, and F. W. Kneubuhler, “Fading relay channels:
performance limits and space-time signal design,” IEEE J. Sel. Areas
Commun., vol. 22, no. 6, pp. 1099–1109, Aug. 2004.
[4] Y. Ida, C. J. Ahn, T. Kamio, H. Fujisaka and K. Haeiwa, “Variable
splitting transmission in multi-relay cooperative communications for SC-
FDMA,” Proc. of 15th International Conference on Advanced Commu-
nication Technology (ICACT 2013), pp. 623–627, Jan. 2013.
[5] Y. Ida, C. J. Ahn, T. Matsumoto and S. Matsufuji, “Removing Deep
Faded Subcarrier Channel for Cooperative Multiuser Diversity OFDMA
Based on Low Granularity Block,” IEICE Transactions on Fundamentals
of Electronics, Communications and Computer Sciences, vol. E97-A, No.
12, pp. 2586–2594, Dec. 2014.
[6] R. Fantini, D. Sabella and M. Caretti, “Energy Efficiency in LTE-
Advanced Networks with Relay Nodes,” Proc. of IEEE 73rd Vehicular
Technology Conference (VTC 2011), Yokohama, pp. 1-5, May 2011.
[7] Y. Marye and H. Zhao, “Nearest Neighbor Relay Selection with Adaptive
Modulation for Improved Throughput and Scalability of Cooperative
Wireless Networks,” Proc. of 5th International Conference on Intelligent
and Advanced Systems (ICIAS 2014), Kuala Lumpur, pp. 1–5, Jun. 2014.
[8] C. J. Ahn and I. Sasase, “The effects of modulation combination target
BER Doppler frequency and adaptive interval on the performance of
adaptive OFDM in broadband mobile channel,” IEEE Trans. Consumer
Electronics, vol. 48, no. 1, pp. 167–174, Feb. 2002.
[9] C. J. Ahn, H. Harada and Y. Kamio, “Superimposed Frequency Symbol
Based Adaptive Downlink OFDM with Frequency Spreading and Equal-
ization,” IEICE Trans. Communications., vol. E89-B, no. 2, pp. 500–508,
Feb. 2006.
[10] F. Ukaji, et al, “Performance enhancement with applying FSS and ASB
method for MU-MIMO,” Proc. of IEEE International Conference on
Industrial Technology (ICIT 2016), pp. 860–865, Mar. 2016.
[11] C. J. Ahn, D. Har, T. Omori, and K. Hashimoto, “Frequency symbol
spreading based adaptive subcarrier block selection for OFDMA system”
Elsevier Digital Signal Processing, vol. 22, no. 3, pp. 518–525, May 2012.
[12] C. J. Ahn, “Reinforced multiuser diversity using frequency symbol
spreading and adaptive subcarrier block selection for OFDMA,” IEICE
Technical Report, vol. 109, no. 266, CS2009-44, pp. 13–18, Nov. 2009.
The 20th International Symposium on Wireless Personal Multimedia Communications (WPMC2017)
978-1-5386-2768-6/17/$31.00 ©2017 IEEE 45
