Orthogonal frequency division multiplexing with offset quadrature amplitude modulation (OFDMOQAM) technique has drawn significant interests in recent years. However, most of the existing OFDM peaktoaverage ratio (PAPR) reduction schemes cannot be used in the OFDMOQAM system directly. In this paper, a modified scheme called overlapped segmental clipping (OSclipping) is proposed to deal with the high PAPR problem specifically in the OFDMOQAM system. For the proposed OSclipping scheme, the input signals are divided into a number of overlapped segments and then the clipping operation is processed on each segment. Simulation results show that the modified scheme used in the OFDMOQAM system can provide better performance than conventional clipping scheme directly used in the OFDMOQAM system, and even outperforms conventional clipping scheme applied in the OFDM system.
1. Introduction
O
rthogonal frequency division multiplexing (OFDM) is one of the most popular multicarrier modulation scheme which has been widely used in many broadband communication systems
[1]
. However, the OFDM system uses rectangular pulse shaping on each subcarrier, which leads to relatively serious outofband radiations. Besides, the insertion of the cyclic prefix (CP) will reduce the data transmission rate of the OFDM system. In order to overcome the shortcomings of the OFDM systems, the OFDM with offset quadrature amplitude modulation (OFDMOQAM) technique has drawn significant interests in recent years
[2

3]
. Compared with the OFDM system, the OFDMOQAM systems provides lower side lobes due to the use of pulse shaping filters
[4]
. Moreover, the OFDMOQAM systems provides higher spectral efficiency and lower narrowband interference since it operates without CP
[5]
. Based on the above advantages, the OFDMOQAM technology has been considered for the physical layer of wireless communication systems recently.
However, as a multicarrier modulation technology, the OFDMOQAM system inevitably produces high peaktoaverage power ratio (PAPR), which degrades the efficiency of the high power amplifier (HPA)
[6]
. In the OFDM systems, much research has been done to deal with the PAPR problem, which can be classified into three categories: signal distortion algorithm, probability algorithm and block coding algorithm. The signal distortion algorithm limits the OFDM signal below a given threshold by nonlinear distortion and includes: signal clipping
[7]
and companding transformation
[8]
. The probability algorithm improves the PAPR performance by significantly reducing the probability that large peak power signal appears and optimizing multicarrier signal PAPR statistical properties. It includes selective mapping (SLM)
[9]
, partial transmit sequence (PTS)
[10
,
11]
, and tone reservation (TR)
[12]
. The coding algorithm
[13]
limits the signals that can be used for transmission, only the signals whose peak value is lower than the maximum peak threshold are allowed to be sent. Among these technologies, signal clipping is an appealing distortion algorithm due to its simplicity and low computation complexity, although the degraded BER and spectrum dispersion may deteriorate the system performance. To overcome the drawbacks of clipping scheme, various techniques have been proposed. The clipping and filtering (CF)
[14]
and repeated clipping and filtering (RCF)
[15]
proposed by Armstrong are of the most classical algorithms. Other modified clipping based schemes can refer to
[16

18]
.
Due to the similarity between the OFDMOQAM and the OFDM system, most research in recent years has been focused on employing the conventional PAPR reduction schemes for the OFDM systems in the OFDMOQAM systems. Whereas, because of the filtered OQAMOFDM signals are overlapped with adjacent data blocks, it is not very effective to employ these algorithms into the OQAMOFDM systems directly.
In recent years, several schemes for the PAPR reduction specifically in the OQAMOFDM systems have been proposed. The clippingbased method of the OFDMOQAM system was discussed in
[19]
, simulation results show that the algorithm designed for the OFDM system can be fully applied to the OFDMOQAM system with similar performance in the synchronized case. An overlapped selective mapping (OSLM) scheme was proposed in
[20]
, for this modified algorithm, the current data block and its adjacent data blocks are jointly considered to choose the optimal phase rotation sequence. Besides, a sliding window tone reservation (SWTR) technique was proposed in
[21]
, which uses the peak reduction tones of several consecutive data blocks to cancel the peak power of the OFDMOQAM signal inside a window. In
[22]
, the authors proposed an improved partial transmit sequence (PTS) scheme by employing multiblock joint optimization (MBJO) for the PAPR reduction of OFDMOQAM signals. The MBJObased scheme exploits the overlapping structure of the OFDMOQAM signal and jointly optimizes multiple data blocks. In
[23]
, a novel segmental PTS (SPTS) scheme is proposed, the key idea of the SPTS scheme is to divide the overlapped OQAMOFDM signals into a number of segments, and then some disjoint subblocks are divided and multiplied with different phase rotation factors in each segment. Aim at the overlapped structure of the OFDMOQAM signals, all of the above schemes provide effective solution for the high PAPR problem, and the main idea can be summarized as: deal with the multiple adjacent data blocks simultaneously instead of optimizing each data block independently.
In this paper, we proposed a modified scheme called overlapped segmental clipping (OSclipping) for the OFDMOQAM system. In this new OSclipping scheme, the input filtered signals are divided into several segments and then the clipping scheme is operated on each segment. For a better PAPR reduction performance, the adjacent segments are overlapped with each other. The simulation results show that, compared to the conventional clipping scheme directly used in the OFDMOQAM system, the modified scheme can provides a better PAPR reduction performance while maintain the same BER performance. Furthermore, the OSclipping scheme used in the OFDMOQAM system even shows better performance than the OFDM system with the conventional clipping scheme.
The rest of the paper is organized as follows. In Section 2, OFDMOQAM system model is illustrated, followed by the modified PAPR reduction schemes being proposed in Section 3. In Section 4, simulation results are given to compare the PAPR reduction performances of the proposed scheme and the conventional algorithm. Finally, the concluding remarks are given by Section 5.
2. Signal Model
In a baseband OFDMOQAM system with
N
subcarriers, real valued symbols modulated by offset QAM are transmitted on each subcarrier, and then the transmitted signal can be written as
[2]
:
where
M
represents the number of input data block,
x^{n}_{m}
denotes the baseband modulated signal on the
n
th subcarrier of the
m
th data block,
g^{n}_{m}
(
t
) represents the pulse shape of index (
m
,
n
) in the synthesis basis which is derived by the timefrequency translated version of the prototype filter function
h(t)
in the following way:
where
and
T
represents the symbol period. In order to maintain the orthogonality among the synthesis and analysis basis, the modified inner product can be written as:
ℜ{·} is the real part operator, and (·)* denotes the conjugate operation. That means only the real part of the correlation function is taken into consideration.
The OFDMOQAM system model is show in
Fig. 1
, to understand the OFDMOQAM signal structure better, we rewrite the complex input symbols
x^{n}_{m}
as:
The OFDMOQAM system model
where
a^{n}_{m}
and
b^{n}_{m}
are the real and imaginary parts of the mth symbol on the nth subcarrier respectively. For two successive subcarriers, the time offset
τ
_{0}
=
T
/2 is introduced onto the real part for the first one and onto the imaginary part for the second one, so the number of the input data block
M
has to be even. Then, the time domain staggered symbols are passed through a bank of transmission filters and are modulated with
N
subcarrier modulators. Thus, the OFDMOQAM modulated signal
s(t)
in (1) can be written as
Here the length of the prototype filter
h(t)
is
LT
. Apparently, the impulse response of prototype filters in the OFDMOQAM systems has longer time duration than
T
, and the real and imaginary parts of input symbols are time staggered, which leads to the signals of adjacent data blocks overlap with each other.
The definition of PAPR in the OFDM system is the ratio of maximum to the average power of the transmitted symbol. Let
s(t)
represents the transmitting complex symbols of duration
T
, then the PAPR of this continuoustime baseband signal can be expressed as:
where E[·] represents the expectation operation. Since the OQAMOFDM signals are overlapped with adjacent data blocks due to the bank of filters and the time offset between the real and imaginary parts, the conventional definition of PAPR for OFDM systems no longer match perfectly to the OQAMOFDM systems. However, it was declared in
[17]
that as both systems transmit the equivalent of one complex symbol at the same rate
T
, we can keep the same measurement for the OFDM/OQAM to provide a fair comparison with the OFDM system.
The PAPR is a random variable, so it is possible to characterize the PAPR distribution (probability that PAPR exceeds a given threshold
γ
) using complementary cumulative distribution function (CCDF) defined as:
3. The Proposed Clipping Scheme
 3.1 Traditional Clipping Scheme
The fundamental principle of the traditional clipping algorithm is to detect the amplitude of the input signals before the time domain OFDM signal is supplied to the power amplifier. If any part of the signal exceeds the preset threshold, a nonlinear processing should be implied to limit the amplitude within the preset threshold; otherwise let the signal through without interference
[6]
. The schematic diagram of clipping scheme is shown in
Fig. 2
.
The schematic diagram of traditional Clipping scheme
The clipped time domain signal can be written as:
where
x(t)
represents the time domain signal,
A
is the clipping threshold, and
ϕ
(
t
) represents the phase of the input signal
x(t)
.
The most important parameter for the performance of clipping scheme is Clipping Ratio (
CR
)
[6]
, which is defined as
where
σ
is the root mean square of the signal power.
x_{0}(t)
, …,
x_{N1}(t)
is the input signal,
N
the number of subcarriers. With
A
denotes clipping threshold, a bigger
CR
corresponds a higher threshold and a worse effect of PAPR reduction performance. And vice versa.
In the OFDM systems, the signal is processed by the conventional clipping scheme in each data block independently. However, the signals in the OFDMOQAM system are overlapped with multiple adjacent data blocks, and the actual PAPR in a certain time interval depends on several consecutive data blocks
[21]
. If we conduct the conventional clipping scheme in the OQAMOFDM systems directly, the peak signal in each data block may be decreased. However, the true PAPR in a certain time interval might not be reduced effectively, since the current data block is affected by multiple consecutive data blocks. Therefore, directly employing the conventional clipping scheme in the OFDMOQAM systems is not very effective.
 3.2 The Proposed Clipping Scheme
In the OFDMOQAM system, since the adjacent filtered signals are overlapped with each other, we conduct the clipping scheme on the signal which is composed of multiple overlapped data blocks, instead of on each data block independently. For our OSclipping scheme, we firstly divide the overlapped data blocks into several segments and then reduce the PAPR of each segment.
The structure diagram of the modified scheme is showed in
Fig. 3
. The time domain OFDMOQAM signal on the
m
th data block
s_{m}
(
t
) can be expressed as a superposition of
N
subcarriers:
The overlapped segmental scheme for the OFDMOQAM signals
As the length of the filtered data block is
LT
(which is the same with the length of the prototype filter) and the overlapped part between two adjacent data blocks is
T/2
, we can obtain the overall length of the
M
input data block:
As can be seen from
Fig.3
, the overlapped signal is divided into
K
segments and the length of each segment is
JT
, where
J
is an integer no less than
1
. The overlapped part of two adjacent segments is expressed as
OT
. The value of
O
ranges from
0
to
J1
. Notice that
O
cannot be equal to
J
, otherwise the OSclipping scheme wouldn't make sense. Theoretically, the larger the length of each segment
J
, the more data blocks can be processed each time, so the better the PAPR performance. However, when
J
is larger than
L
, the performance improvement will not be significantly any more. That is because the length of each filtered data block is
LT
, and only the overlapped part of adjacent data blocks can have influence on the former data blocks. The PAPR performances with different
J
and
O
are given in Section 4.
Specifically, when
J=1
,
O=0
, it means each segment is the same length as one data block and there is no overlap between two segments. Under this special circumstance, the modified OSclipping scheme is just equal to the conventional clipping scheme.
After
J
and
O
are defined, the number of the segments
K
can be obtained by:
where ⌈⋅⌉ means the round up operation. As
M
represents the number of input symbols and is usually set larger than four figures,
L
is usually set as 4. Considering
M
>>
L
,
L
can be ignored and the simplified expression of
K
can be obtained by:
From equation (15) we can conclude that the number of the segment is primarily determined by
J

O
, which represents the time offset between two adjacent data blocks.
Then the filtered signals are divided into
K
segments and the modified OSclipping scheme is conducted on each segment. Assume that the previous
k
−1 (1≤
k
≤
K
) segments have already been processed, then extract the signal
s_{k}
(
t
) in the
k
th segment from
s_{m}
(
t
), and
s_{k}
(
t
) can be expressed as:
Operate the clipping scheme on the signal
s_{k}
(
t
) in the
k
th segment. If the amplitude of
s_{k}
(
t
) exceeds the preset threshold, it will be clipped to
A
; otherwise,
s_{k}
(
t
) will remain unchanged. Then the clipped signal can be written as:
where
ϕ_{k}
(
t
) is the phase of
s_{k}
(
t
). It can be seen form (17), the phase information of
s_{k}
(
t
) remain unchanged after clipping operation.
After the signal in the
k
th segment has been clipped, increase
k
by 1 and repeat the procedure until all
K
segments have been processed.
When operating the conventional clipping scheme directly in an OFDMOQAM system with
N
subcarriers, the
LN
(
L
is the length factor of the prototype filter) input signals in each filtered data block need to be compared with the threshold
A
and the computational complexity can be seen as
LN
complex additions. As for the proposed OSclipping scheme, the data in the overlapped part need to be compared twice, which will result in
ON
more complex additions (
O
is the overlapped factor and usually set less than
J
). However, in an OFDMOQAM system, (
L
+
1
)
MN
(
M
is the number of the input data blocks) real multiplications are required to create the filtered signals
[21]
, so the increase of complexity in the modified scheme can be ignored in the OFDMOQAM system. Besides, we can also set a reasonable value for
O
so the computational complexity and the PAPR reduction performance can be well balanced.
4. Simulation Results
In this section, the PAPR reduction performance and the BER performance of the proposed OSclipping technique are conducted for comparison. The OFDMOQAM system employs 64 subcarriers with 5000 input data blocks. The length of the prototype filter
h
(t) is set as 4
T
, which means the time duration of the filtered signals is about four times of that of the input signals. The Complementary cumulative distribution function (CCDF) is applied as measurement of PAPR reduction performance in the simulations.
The PAPR performance of the OSclipping scheme with different O
Firstly, we compare the PAPR performance of the modified OSclipping scheme with different lengths of the overlapped part
O
. In the simulation, the clipping ration is
CR=1.0
, the length of each segment is set as
3T
, i.e.,
J= 3
; and the length of the overlapped parts is
O=0, 1, 2
, where
O=0
means there is no overlap between adjacent segments. It can be seen from
Fig. 5
that the OSclipping scheme is effective in reducing the PAPR for the OFDMOQAM system, and a larger
O
can improves the performance of the OSclipping scheme a little.
The PAPR performance of the OSclipping scheme with different J and O
The PAPR performance of the OSclipping scheme with different
J
is shown in
Fig. 5
. It Is obvious that the overlapped segment structure is quite effective in reducing the PAPR and among all simulated conditions, the best PAPR reduction performance is achieved with
J=5
,
O=4
. It can also be concluded from the results that the larger the length of each segment
J
, the better the PAPR reduction performance. When
J
is increased to 4, the performance difference can be ignored. Besides, the increasing of
O
can also help to improve the PAPR performance, but the performance improvement is not as evident as that of when
J
increases. From equation (14) we can know that the number of the segments
K
is determined by (
JO
), and a larger
O
means higher computational complexity. As a result, we will use
W=4
,
O=2
in the following simulations to reduce the calculation costs.
Fig. 6
shows the PAPR performance of the proposed OSclipping technique and the conventional clipping scheme with different thresholds. The clipping ratios are:
CR1=1.0
,
CR2=1.1
,
CR3=1.2
. For the OSclipping scheme, the length of each segment is
J=4
, and the length of the overlapping part is set as
O=2
. It can be concluded that the PAPR performance of both schemes improves as
CR
decreases. When
CR=1.1
, the modified OSclipping scheme has a PAPR that exceeds 4.5dB for less than 0.1%, which is 0.7dB better than the 5.2dB of the conventional clipping scheme.
The PAPR performance of the OSclipping with different thresholds
In addition, the PAPR reduction performance of the OFDM system and the OFDMOQAM system are compared in simulation. Both of the two systems employ 64 subcarriers and 5000 input data blocks. The clipping ratio is set as
CR=1.1
in the conventional and the modified clipping schemes. From the simulation results shown in
Fig. 7
, we know that the conventional clipping scheme directly used in the OFDMOQAM system is not as effective as that used in the OFDM system. However, when the modified OSclipping scheme is applied, the PAPR reduction performance of the OFDMOQAM system is even better than that of the OFDM system with the conventional clipping scheme.
The PAPR performance of the OFDM and the OFDMOQAM system
The Bit Error Rate (BER) performance as a function of Signal to Noise Ratio (SNR) is shown in
Fig. 8
. As in the simulations above, the clipping ration of both clipping scheme is
CR=1.1
, and
J=4
,
O=2
is set for the modified OSclipping scheme. As can be seen from the results, the BER performance of the conventional clipping scheme decreases the SNRby about 0.6dB at BER=10
^{3}
, and the modified OSclipping scheme shows almost the same performance as the traditional clipping scheme. Considering the better PAPR reduction performance, we can conclude that the proposed OSclipping scheme can improve the performance of the whole OFDMOQAM system.
The BER performance of the OSclipping scheme
5. Conclusion
In this paper, a modified OSclipping scheme is proposed to reduce the PAPR of the OFDMOQAM system. Considering the overlapped structure of the OFDMOQAM signals, the modified OSclipping scheme divides the filtered signal into a number of segments and each segment is composed of multiple consecutive data blocks. Therefore, the influence of adjacent overlapped data block can be eliminate and the PAPR during a certain time interval can be reduced effectively. In addition, the overlapped structure of the adjacent segments can further improves the PAPR reduction performance. The simulation results showed that the OSclipping scheme could provide a better PAPR reduction performance than the OFDMOQAM system applying the conventional clipping scheme. Still, the BER performance remained unchanged. Besides, the OFDMOQAM system with the OSclipping scheme even outperformed the OFDM system with the conventional clipping scheme.
BIO
Jianping Wang received her PhD degree from the institute of Optoelectronic and Precision Instrument Engineering at Tianjin University in 2000. From 2000 to 2005, she is a senior lecture in the Department of Electronics Engineering, Tsinghua University. Now she is a professor in the University of Science and Technology Beijing. Her research interests are in optical fiber communication and Radio over Fiber technology.
Jing Yan received her bachelor’s degree from the School of computer and Communication engineering at the University of Science and Technology Beijing in 2012. Now she is pursuing the master degree in the Department of Communication Engineering, School of computer and Communication engineering at the University of Science and Technology Beijing. Her research interests lie in wireless network and optical fiber communication.
Zhen He received his master's degree from the School of computer and Communication engineering at Zhengzhou University of Light Industry in 2011. Now he is pursuing the PhD degree in the Department of Communication Engineering, School of computer and Communication engineering at the University of Science and Technology Beijing. His research interests lie in wireless network and optical fiber communication.
Zheng Hui
,
Shrestha Robin
,
Hwang Jaeho
,
Kim JaeMong
2012
“A Parallel Combinatory OFDM System with Weighted Phase Subcarriers”
KSII Transactions on Internet and Information Systems (TIIS)
Article (CrossRef Link)
6
(1)
320 
338
Siohan P.
,
Siclet C.
,
Lacaille N.
2002
“Analysis and design of OFDM/OQAM systems based on filterbank theory”
IEEE Transactions Signal Processing
Article (CrossRef Link)
50
(5)
1170 
1183
DOI : 10.1109/78.995073
Du Jinfeng
,
Signell Svante
2007
“Classic OFDM Systems and Pulse Shaping OFDM/OQAM Systems”
Technical Report of NGFDM Project, KTH/ICT/ECS
Article (CrossRef Link)
Bellanger M.
2010
“Physical layer for future broadband radio systems”
IEEE Conference on Radio and Wireless Symposium
Article (CrossRef Link)
436 
439
B¨oelcskei H.
2003
“Orthogonal frequency division multiplexing based on offset QAM”
in Advances in Gabor Analysis
Article (CrossRef Link)
321 
352
Skrzypczak A.
,
Siohan P.
,
Javaudin J. P.
2006
“Analysis of the peaktoaverage power ratio for OFDM/OQAM”
in Proc. of IEEE 7th Workshop on Signal Processing Advances in Wireless Communications
Article (CrossRef Link)
1 
5
Deng S.K.
,
Lin M. C.
2007
“Recursive Clipping and Filtering with Bounded Distortion for PAPR Reduction”
IEEE Transactions Communications
Article (CrossRef Link)
55
(1)
227 
230
DOI : 10.1109/TCOMM.2006.885102
Wang X. B.
,
Tihung T.
1999
“Reduction of peaktoaverage power ratio of OFDM systems using a companding technique”
IEEE Transaction on Broadcasting
Article (CrossRef Link)
45
(3)
303 
307
DOI : 10.1109/11.796272
Bauml R. W.
,
Fischer R. F. H.
,
Huber J. B.
1996
“Reducing the peak to average power ratio of multicarrier modulation by selected mapping”
IEEE Electronic Letters
Article (CrossRef Link)
32
(22)
2056 
2057
DOI : 10.1049/el:19961384
Müller S. H.
,
Huber J. B.
1997
“OFDM with Reduce. Peak–to–Average Power Ratio by Optimum Combination of Partial Transmit Sequences”
IEEE Electronic Letters
Article (CrossRef Link)
33
(5)
368 
369
DOI : 10.1049/el:19970266
Lim Hwan
,
Rhee Dai Ho
2012
“A low complexity PTS technique using threshold for PAPR reduction in OFDM systems”
KSII Transactions on Internet and Information Systems (TIIS)
Article (CrossRef Link)
6
(9)
Krongold S.
,
Jones D. L.
2004
“An activeset approach for OFDM PAR reduction via tone reservation”
IEEE Transactions on Signal Processing
Article (CrossRef Link)
52
(2)
495 
509
DOI : 10.1109/TSP.2003.821110
Sathananthan K.
,
Tellambura C.
2002
“Coding to reduce both PAR and PICR of OFDM signal”
IEEE Communications Letters
Article (CrossRef Link)
6
(8)
316 
318
DOI : 10.1109/LCOMM.2002.802065
Armstrong J.
2001
“New OFDM Peaktoaverage Power Reduction scheme”
in Proc. of IEEE VTS 53rd Vehicular Technology Conference
vol. 1, Article (CrossRef Link)
756 
760
Armstrong J.
2002
“Peaktoaverage power reduction for OFDM by repeated clipping and frequency domain filtering”
IEEE Electronic Letters
Article (CrossRef Link)
38
(5)
246 
247
DOI : 10.1049/el:20020175
Linfeng Liang
,
Jun Wang
,
Jian Song
2013
“Adaptive algorithm for OFDM PAPR reduction using clippingnoise guided signselection”
IEEE International Symposium on Broadband Multimedia Systems and Broadcasting (BMSB)
Article (CrossRef Link)
1 
3
Xiaodong Zhu
,
Wensheng Pan
,
Hong Li
,
Youxi Tang
2013
“Simplified Approach to Optimized Iterative Clipping and Filtering for PAPR Reduction of OFDM Signals”
IEEE Transactions on Communications
Article (CrossRef Link)
61
(5)
1891 
1901
DOI : 10.1109/TCOMM.2013.021913.110867
Wang L. Q.
,
Chintha T. A.
2005
“Simplified Clipping and Filtering Technique for PAR Reduction in OFDM Systems”
IEEE Signal Processing Letters
Article (CrossRef Link)
12
(6)
453 
456
DOI : 10.1109/LSP.2005.847886
Rahim M. U.
,
Stitz T. H.
,
Renfors M.
2009
“Analysis of clippingbased PAPRreduction in multicarrier systems”
in Proc. of IEEE 69rd Vehicular Technology Conference
Article (CrossRef Link)
1 
5
Skrzypczak A.
,
Siohan P.
,
Javaudin J. P.
2006
“Reduction of the peaktoaverage power ratio for OFDMOQAM modulation”
in Proc. of IEEE 63rd Vehicular Technology Conference
vol. 4, Article (CrossRef Link)
2018 
2022
Lu S.
,
Qu D.
,
He Y.
2012
“Sliding window tone reservation technique for the peaktoaverage power ratio reduction of FBMCOQAM signals”
IEEE Wireless Communication Letters
Article (CrossRef Link)
1
(4)
268 
271
DOI : 10.1109/WCL.2012.062512.120360
Qu Daiming
,
Lu Shixian
,
Jiang Tao
2013
“MultiBlock Joint Optimization for the PeaktoAverage Power Ratio Reduction of FBMCOQAM Signals”
IEEE Transactions on Signal Processing
Article (CrossRef Link)
61
(7)
1605 
1613
DOI : 10.1109/TSP.2013.2239991
Ye C.
,
Li Z.
,
Jiang T.
,
Ni C.
,
Qi Q.
2014
“PAPR Reduction of OQAMOFDM Signals Using Segmental PTS Scheme with Low Complexity”
IEEE Transactions on Broadcasting
Article (CrossRef Link)
60
(1)
141 
147
DOI : 10.1109/TBC.2013.2282732