ISSN 1450216X Vol.75 No.4 (2012), pp. 624642
© EuroJournals Publishing, Inc. 2012
http://www.europeanjournalofscientificresearch.com
A Survey on CF Method, PTS Approach, Companding
Technique and Time Domain Methods for
PAPR Reduction in OFDM Systems
Arvind Chakrapani
Assistant Professor, Electronics and Communication Engineering Department
Tamilnadu College of Engineering, Coimbatore, India
Email: arvichakra@gmail.com
Tel:+914212332544; Fax: +914212332244
V. Palanisamy
Principal, Info Institute of Engineering, Coimbatore, India
Email: vpsamyin@gmail.com
Tel: +91422 236 3700 ;Fax: +91 422  265 4888
Abstract
Orthogonal Frequency Division Multiplexing (OFDM) is widely used modulation
technique in wireless communication systems due to its ability to transmit high data rates
over multipath and frequency selective fading radio channels effectively. But a large Peak
toAverage Power Ratio (PAPR) causes significant performance degradation due to the non
linear effects of power amplifier; hence reduction of PAPR is mandatory in the design of
OFDM systems. This paper presents a survey on the Clipping and Filtering(CF) Method,
Partial Transmit Sequence (PTS) approach, Companding Technique (CT) and time domain
methods to reduce the PAPR of an OFDM signal. Based on the literary analysis performed
on the above stated methods, five proposals for future work were identified to limit PAPR
in a OFDM system namely (i) Iterative Clipping and time domain filtering technique based
on convex optimization technique (ii) DFT based PTS approach (iii) Differential Evolution
Algorithm assisted PTS scheme (iv) Gaussian Transform based Companding Technique
and (v) A time domain based PAPR reduction method.
Keywords: Clipping and Filtering Technique, Companding Technique, Orthogonal
Frequency Division Multiplexing, Partial Transmit Sequence, Peakto
Average Power ratio
1. Introduction
Orthogonal Frequency Division Multiplexing(OFDM) is a multicarrier modulation or multiplexing or
spread spectrum technique where a high data rate stream is subdivided into narrow lower data rate
streams, each of which modulates the individual orthogonal subcarriers such that the data is transmitted
simultaneously over the sub carriers. Basically an OFDM signal is the sum of individual orthogonal
subcarriers that are individually modulated by using either Quadrature Amplitude Modulation (QAM)
A Survey on CF Method, PTS Approach, Companding Technique and
Time Domain Methods for PAPR Reduction in OFDM Systems 625
or Phase Shift Keying (PSK) or Quadrature Phase Shift Keying (QPSK). Since OFDM provides high
spectral efficiency, power efficiency, robustness to multipath fading and immunity to frequency
selective fading channel, it is widely used for high data rate wireless communication systems and in
many wireless communication standards such as Digital Audio and Video Broadcasting systems,
Wireless Local Area Network (WLAN), VeryHighSpeed Digital Subscriber Loop (VDSL) and
Asymmetric Digital Subscriber Line (ADSL). In worst case the OFDM signal may add all the
sinusoidal signals constructively in phase and amplitude so that the amplitude is N times the average
power[40].
1.1. Peak to Average Power Ratio (PAPR) Problem
Tasadduq and Rao (2002) stated that in OFDM transmission Systems when the number of subcarriers
increases the Gaussian distributed signal approaches to that of a sample function, Peaktoaverage
power ratio (PAPR) is a good measure of the resulting occasional peaks. PAPR of an OFDM signal
is defined as the ratio of the maximum instantaneous signal power to the average signal power.
av
NT t
P
t x
t x PAPR
]  ) ( [
)} ( , {
2
0
max
≤ ≤
= (1)
In eqn. (1), P
av
is the average power
T is the time period of an OFDM symbol
Even though large peaks occur very rarely, one of the main drawbacks of the OFDM system is
the high Peak – to – average Power ratio which leads to increase in complexity of Analog to Digital
Converters (ADC) and Digital to Analog Converters (DAC) while introducing intermodulation
distortion, spectral spreading and undesired outof band radiation [20]. Large PAPR also leads to
Adjacent Channel Interference, degradation of Bit Error Rate (BER) performance and variation of
OFDM signal constellation [17]. The above mentioned issues can severely harm the OFDM system
performance and demands expensive transmitters for normal operation. Since the positive features of
OFDM can support the design of highly effective wireless commmunication systems, extensive
research activities were carried out in the late 1990s to study the distribution of PAPR and its reduction
in OFDM systems. Eventhough coding schemes for reducing peak power were employed by Wulich
(1996) in multicarrier modulation systems and Eetvelt et al., (1996) in QPSKOFDM systems, a major
break through in PAPR reduction methods started with Bauml et al., (1996) work on Selected mapping
(SLM) technique, Muller and Huber’s (1997) Partial Transmit Sequence (PTS) approach, Li and
Cimini’s (1998) Clipping and filtering technique and Wang et al., (1999) article on Companding
Technique. The research community have proposed innovative methods to reduce PAPR of OFDM
signal, recently PAPR reduction is achieved in time domain directly with reduced computational
complexity avoiding the need for FFT operation.
he rest of the paper is organized as follows. Section 2 presents a survey on Clipping and
filtering methods which includes iterative clipping and filtering methods and noniterative clipping
methods. Section 3 elaborates on Partial Transmit Sequence (PTS) approach with a brief review on
phase optimization methods, extended PTS approach, PTS based on DFT property and Algorithm
assisted PTS approach. Section 4 discusses on Companding schemes with emphasis on nonlinear and
linear companding schemes. Section 5 examines the time domain based methods for PAPR reduction.
Section 6 concentrates on overview of survey and Scope for future work followed by conclusion in
Section 7.
626 Arvind Chakrapani and V. Palanisamy
2. Clipping and Filtering Method
The clipping operation can be viewed as the product of OFDM signal and a rectangular window
function that equals one if the OFDM signal amplitude is below a threshold and less than one if the
peak amplitude needs to be clipped. The conventional hard clipping introduces nonlinear distortion
and out – of – band radiation which leads to spectral regrowth [16].
Li and Cimini (1998) introduced a post filtering operation to reduce the effects of out of band
radiation and spectral regrowth due to the clipping process, however the possibility of spectral
regrowth in the time domain is not addressed. Their work was the first to introduce filtering operation
after the clipping process inorder to reduce the spectral regrowth. Inorder to reduce the out of band
radiation, spectral regrowth and clipping noise involved in the PAPR reduction process, repeated
clipping and filtering methods were introduced while noniterative clipping and filtering methods
reduced the computational complexity significantly.
2.1. Iterative Clipping and Filtering Methods
Armstrong (2002) performed a repeated clipping operation followed by an FFT  based frequency
domain filtering of an oversampled time domain OFDM signal to achieve a PAPR reduction with only
moderate level of clipping noise and no increase in out – of – band power. This method distorts the in
band spectral components to shrink the signal constellation and adds noise. While Leung et al., (2002)
proposed an Iterative Clipping and Filtering(ICF) technique in time domain to achieve similar PAPR
reduction performance as that of Armstrong’s method with reduced complexity requires no FFT/IFFT
operation is presented in Table 1.
Table 1: Computational complexity of Armstrong’s ICF and Leung’s ICF technique.
Operation Armstrong’s ICF technique Leung’s ICF technique
Computational Complexity for K
iterations
Computation time in seconds 169 10
Where L – oversampling rate
Deng and Lin (2007) introduced a Repeated Clipping and Filtering (RCF) method, where the
number of recursions is reduced by employing Smart Gradient Projection (SGP) algorithm and also
bounds or limits the distortion on each tone after each recursion to reduce the error rate and PAPR of
an OFDM signal.
The iterative clipping process makes it difficult to estimate the BER performance. Bae et al.,
(2010) based on a noise enhancement factor suggested analytical expressions for the estimation of
attenuation factor, BER and Error Vector Magnitude (EVM). Their work also characterizes clipping
noise for iterative processes and the first to report an effective tradeoff between PAPR reduction and
BER performance in an ICF technique.
Wang and Luo (2011) proposed a ICF technique where each iteration is expressed as a convex
optimization technique and the optimal frequency response filter is designed to minimize the signal
distortion so that PAPR is reduced for each of the OFDM symbol. This method achieves a PAPR
reduction in just one or two iterations while the same performance in a conventional technique requires
eight to sixteen iterations.
A Survey on CF Method, PTS Approach, Companding Technique and
Time Domain Methods for PAPR Reduction in OFDM Systems 627
2.2. Non – Iterative Clipping Methods
A non – iterative constrained clipping technique by Baxley et al., (2006) satisfies the in – band metric
called Error Vector Magnitude(EVM) and the maximum permissible out – of – band spectral
constraints, so that PAPR of an OFDM signal is reduced.
A computationally efficient single iteration clippingfiltering technique proposed by Wang and
Tellambura (2005) exploits the fact that the clipping noise obtained after several iterations of clipping
and filtering is approximately equal to that generated in the first iteration. The simplified technique
scales the clipping noise generated during the first iteration and with just three FFT/IFFT operations it
achieves the same PAPR reduction as that of the existing iterative techniques with (2K+1) FFT/IFFT
operations where K is the number of iterations.
Clipping methods introduce inband and out of band radiation while the filtering process
reduces the out of band radiation due to clipping but cannot reduce the inband distortion [29]. Even
though the ICF process reduces spectral expansion and eliminates peak re  growth, it is time
consuming and increases the complexity of the transmitter.
3. Partial Transmit Sequence (PTS) Approach
Muller and Huber (1997) proposed a PTS approach where the subcarrier block is decomposed into
multiple disjoint sub blocks and each(other than the first sub block) of which is multiplied by a
common phase factor to generate a statistically independant new signal vectors that are optimally
combined to reduce PAPR of an OFDM signal. An exhaustive search over all the possible combination
of permissible phase factors is mandatory to find the optimal Phase factor but the search complexity
increases with the number of sub blocks exponentially [52]. The candidates generated using PTS
approach are interdependant and the required number of bits for transmitting the side information is
more compared to the Bauml et al., (1996) SLM scheme.
Figure 1: PTS based PAPR reduction in OFDM system (Muller and Huber, 1997)
628 Arvind Chakrapani and V. Palanisamy
3.1. PTS Schemes Based on Phase Optimization Methods
Muller and Huber (1997) introduced an effective PAPR reduction scheme called Optimal Binary Phase
Sequence (OBPS) for OFDM systems with arbitrary number of subcarriers and unconstrained signal
set by optimally combining the partial transmit sequences.
Subcarrier block A
µ
is subdivided into V pair wise disjoint carrier sub blocks
) (i
A
µ
where i =
1,2, … V. An optimization parameter called rotation factor or phase factor { } π ∈ φ =
µ
φµ
µ
2 , 0 ,
) ( ) ( ) ( i i j i
e b is
introduced to each sub block i and the peak power optimized PTS in time domain is
) (
1
) ( i
v
i
i
a b
µ µ ∑
=
⋅ = (2)
The above scheme works with almost vanishing redundancy and has a transmitter complexity
which increases exponentially with the number of sub carriers.
Tellumbura (1998) proposed a phase optimization criterion for several block phase factor in
PTS approach for PAPR reduction in OFDM system with increase in complexity.
The new optimized criterion is derived as
∑
−
=
=
1
0
3 2
 ) (  min arg ] ,
ˆ
,....
ˆ
,
ˆ
[
N
k
v
k ρ φ φ φ
Where ρ(k) is the aperiodic autocorrelation of the information vector.
Hill et al. (2000) introduced an adaptation to OBPS to reduce PAPR by combining cyclic shift
of the IFFT sub block output with Partial Transmit Sequences. Cyclically shifted PTS increases the
number of alternative transmit sequences with trivial operations by cyclically shifting the data before
or after they are phase rotated
) (
1
) ( ~ i j
v
i
i
e a a
φ
µ µ ∑
=
⋅ = (3)
Cyclic shift in Time Domain is performed as
∑
=
∂ +
⋅ =
v
i
i j i
v
e a a
1
) ( ) (
) (
~ φ
µ µ
(4)
In eqn. (4), ) (v δ is cyclic shift in time domain.
Cimini and Sollenberger(2000) introduced suboptimal iterative flipping algorithm (IFA) for
combining partial transmit sequences to reduce PAPR but it has a performance gap with ordinary PTS
technique, in other words the algorithm performs worse than conventional PTS with reduced
complexity.
Tellambura (2001) computed an optimal set of quantized phase factors and it achieves better
performance than the exhaustive search process of PTS approach. For small number of sub blocks the
proposed algorithm performs better than OBPS and when the number of sub blocks is large it performs
similar to OBPS.
Han and Lee (2004) suggested a gradient descent search algorithm to compute the phase factors
which leads to reduced PAPR statistics than the IFA with reduced search complexity and little
performance degradation.
3.2. Extended PTS Approach
Kang et al. (1999) proposed a concatenated pseudorandom Sub Block partition Scheme (SPS) for
PTS, it has the same PAPR reduction performance as that of the conventional pseudorandom SPS with
extensively reduced computational complexity. For M sub block partitioning, this method produces
candidates.
Chen and Pottie (2002) proposed a novel orthogonal projection based on PTS approach that
achieves significant PAR reduction at low redundancy.
Equation (2) can be rewritten in matrix form as
A Survey on CF Method, PTS Approach, Companding Technique and
Time Domain Methods for PAPR Reduction in OFDM Systems 629
] ,....... , [
~
2 2 1 N
T T T T
m b m b m b M b a = =
µ
(5)
Where m
n
– nth column of Matrix M,
b – Column phase vector of length V, b=[b
1
,b
2
… … ..b
v
]
This geometrical interpretation in eqn. (5) forms the basis for finding the optimal phase rotation
vector ‘b’ based on orthogonal vector function of the columns . 1 , N n m
n
≤ ≤
Schenk et al. (2005) proposed a Spatial shifting based PTS scheme to transmit the partial
transmit sequences of the OFDM signal on the transmission branches with minimum PAPR, so that
significant PAPR reduction with limited complexity and signaling overhead is achieved. While Xiao et
al.(2007) presented a Low Complexity PTS(LC – PTS) for reduction of PAPR by exploiting the
correlation among the candidate signals, the computational complexity of LCPTS and PTS is given in
Table 2.
Table 2: Computational complexity of PTS and LCPTS
Operation PTS LCPTS Percentage
Complex Addition
Complex Multiplication
Computations
Where L – Number of candidate signals
M – Number of sub blocks
N – Number of sub carriers
W – Number of phase factors
P – Number of highest amplitude positions
In 2010, Ghassemi and Gulliver employed the Autocorrelation function of PTS sub blocks to
develop a new PTS sub blocking scheme using errorcorrecting codes. This method minimizes the
number of repeated sub carriers within a sub block and provides better PAPR reduction than
pseudorandom or msequence sub blocking with low complexity.
3.3. PTS Approach Based on DFT Property
DFT property is employed to generate alternate candiadates of an OFDM signal, among which the
candidate with lowest PAPR is selected for transmission.
Lu et al. (2006) suggested to apply certain transformations on PTS (TPTS) of a OFDM signal
to generate alternate frequency domain signals and transmit the signal which has a reduced PAPR than
that of the actual signal. Basic operations like complex conjugation, frequency reversal, circular shift
and their combinations are used either individually or jointly on different sub blocks to perform the
transformation. The alternative frequency domain signal can be expressed from eqn. (2) as
[ ]
∑
=
µ µ µ
=
v
t
i i i
A T b A
1
) ( ) ( ) (
ˆ
(6)
Where [ ]
) ( ) ( i i
A T
µ
is a certain predefined transformation made on
) (i
A
µ
. The time domain signal
of eqn.(6) is given as
[ ] { }
) ( ) ( )
~
(
1
~ i i i v
t
A T IDFT b a
µ µ = µ ∑
= (7)
TPTS outperforms OBPS method for the same number of sub blocks and exhibits similar performance
for the same number of carriers with reduced complexity.
630 Arvind Chakrapani and V. Palanisamy
A sub – optimum PTS scheme introduced by Wang and Cao, (2008) combines an alternate
optimization method and linearity property of IDFT to reduce the computational complexity while
increasing the number of candidate signals, so that PAPR reduction performance of OPTS is attained
with dramatically reduced computational complexity.
Five novel simple transformations are proposed by Zhu et al., (2008) namely (i) circular time
shift of
) (i
a
µ
(ii) circular frequency shift of
) (i
A
µ
(iii) time reversal of
) (i
a
µ
(iv) complex conjugate of
) (i
a
µ
(v) complex conjugate of
) (i
A
µ
employs DFT property on PTS in an iterative manner to reduce PAPR.
For small number of sub blocks, some of the proposed transformations achieve better performance than
the IFA.
A PTS method by Yang et al., (2011) recursively combines the cyclically shifted sub block
sequences based on the linearity property of IFFT to generate a set of candidate signals in the time
domain with different phase constellation without employing multiplication. The phase detector
recovers the OFDM signal at the receiver. This method achieves PAPR reduction and maintains bit
error rate performance as that of the conventional PTS(OPTS) in both AWGN and Rayleigh fading
channel but requires a detector at the receiver.
When compared to conventional PTS approach, the DFT assisted PTS approach generates more
candidates of a OFDM signal and provides the liberty to select the candidate with lowest PAPR for
tramsmission.
3.4. Algorithm Assisted PTS Approach
Nguyen and Lampe (2008) preprocessed the data stream ahead of PAPR reduction so that side
information is embedded with minimal possible redundancy, maintaining the BER without causing
peak regrowth. The complexity in the search process of phase factor is formulated as a combinatorial
optimization problem enabling us to (i) unify different search strategies proposed in the PTS literature.
(ii) adapt different optimization algorithms known from the literature to assist the PTS approach.
Chen (2009) reduced the search complexity of PTS approach by combining the PTS with cross
– entropy method and achieved the same PAPR reduction perfromance as that of Exhaustive Search
Algorithm (ESA). The PAPR problem is rewritten as a score or fitness function that is further
translated to a stochastic approximation problem which is effectively solved using a stochastic
optimization technique called cross – entropy method.
Chen (2010a) expressed the PAPR reduction problem as a combinatorial optimization problem
which is effectively solved using Quantum Inspired Evolutionary Algorithm to find the optimal phase
factors that achieves significant PAPR statistics. While the search for optimal phase factors in a PTS
scheme is rewritten as a global optimization problem and is solved using a population based search
method called Electromagnetism  like (EM) Algorithm. The EM based PTS scheme by Chen (2010b)
follows a stochastic optimization approach and employs attraction – repulsion mechanism to search the
optimal phase rotation factor so that the desired PAPR statistics is achieved with reduced complexity.
Wang et al, (2010) proposed a sub optimal method by combining a numeric function
optimization algorithm called Artificial Bee Colony Algorithm with PTS (ABCPTS) to reduce the
search complexity of the allowable phase factors, such that PAPR reduction is achieved.
Taspinar et al., (2011) proposed an iterative heuristic search method in which Parallel Tabu
search Algorithm is used in conjunction with PTS (Parallel TSPTS) approach eliminates the iterations
that visits the solution obtained recently and selects the optimal phase rotation vectors to optimize the
PAPR statistics. The parallel information exchange between Tabu search algorithm is based on
crossover operator used in Genetic Algorithm (GA) combines good features of the parent to achieve
better performance. PAPR reduction and search complexity comparison of various PTS schemes
performed on a 16 QAM modulated OFDM signal with 256 subcarriers is presented in Table 3.
A Survey on CF Method, PTS Approach, Companding Technique and
Time Domain Methods for PAPR Reduction in OFDM Systems 631
Table 3: Comparision of various PTS schemes based on PAPR reduction
PTS schemes Number of searches (S) PAPR (dB)
Original OFDM 0 11.26
OPTS
6.74
IFA 1000 7.13
ABCPTS 1000 6.98
Parallel TSPTS 1000 6.93
Since Genetic Algorithm converges rapidly in changing channel conditions and provides
solution for Gradient based search method, Lixia and Murroni (2011) used it along with PTS approach
to reduce PAPR in multicarrier modulation(MCM) systems. Based on their study on OFDM and
Wavelet packets multicarrier modulation (WP – MCM) in Additive White Gaussian Noise (AWGN)
channel, GA applied to PTS reduces PAPR more effectively in OFDM rather than for WP – MCM.
4. Companding Technique
The first companding scheme was introduced based on the similarity between OFDM signal and
speech signal that large signal occurs infrequently by Wang et al. in 1999. He applied the idea of
companding in speech processing to reduce PAPR of an OFDM signal and to improve its transmission
performance. Companding scheme effectively compresses the large signals and enhances the small
signals to achieve PAPR reduction.
Figure 2: Block Diagram of a Companded OFDM system (X Wang et al., 1999)
4.1. Non Linear Companding Scheme
Wang et al., (1999) introduced a µLaw companding technique to generate optimal companding
coefficients to limit PAPR of a OFDM signal and improves the BER performance. A companded
signal increases the transmitter signal power while the noise power remains constant and makes the
High Power Amplifier(HPA) to operate in the non linear region. Mattsson et al. (1999) commented on
their work that Companding leads to spectral regrowth and raised a question whether the improvement
in BER is due to companding or increase in the transmit signal power. Later Wang et al. (1999)
compared the performance of a companded signal with that of the non companded signal with constant
transmission power to infer that in the region of high SWR the SER of a companded signal achieves
better results than that of an uncompanded signal. He also stated that the spectral regrowth due to
companding is minimal and is insensitive to the variations of companding coefficients. The
companding scheme shows better performance than clipping, but ignored the nonlinear operation of
power amplifiers which leads to spectral side lobes growth.
632 Arvind Chakrapani and V. Palanisamy
Huang et al., (2001) stated that the increase in the average input power of µLaw companding
technique can be avoided and maintained constant by transforming the OFDM signal based on their
power distribution, so that the HPA operates in linear region. In µlaw companding transform, by the
proper selection of the companding form and its corresponding inflexion point the PAPR can be
reduced with reduced complexity and moderate degradation.
Jiang and Zhu (2004) introduced a Nonlinear Companding Transform described by a single
valued function exploits the statistical characteristics of the OFDM signal to limit PAPR and exhibits
good BER performance in a AWGN channel.
Huang et al. (2004) proposed the design criteria of Companding Transform (CT) based on the
statistical characteristics of OFDM signal to enable an effective tradeoff between PAPR reduction and
BER performance. He also discussed about the performance of four companding schemes namely (i)
Linear Symmetrical Transform (LST) (ii) Linear Nonsymmetrical Transform (LNST) (iii)Nonlinear
Symmetrical Transform (NLST) (iv) Nonlinear nonsymmetrical Transform (NLNST). The inflexion
point in NLST treats the large and small amplitude signals on different scales and achieves better
performance than the other three schemes and the clipping method.
Jiang et al., (2005) proposed an Exponential Companding (EC) technique that converts the
amplitude statistics of an OFDM signal into a uniformly distributed signal to limit the PAPR while
maintaining the average signal power constant.
The Exponential Companding function is given by
d
x
a x x h
(
¸
(
¸


¹

\

σ
− − =
2
2
exp 1 ) sgn( ) ( (8)
In equation (8), sgn(x) is sign function
[ ]
2
2
2
2
2
exp 1
d
d
n
n
s
E
s E
a









¹

\

(
(
(
¸
(
¸
(
(
(
¸
(
¸
¸


¹

\

σ
− −
= maintains the average signal power constant.
The decompanding function at the receiver side is given by


¹

\

α
− σ − =
−
d
e
x
x x h 1 log ) sgn( ) (
2 1
(9)
The companding functions, eqn. (8) and eqn. (9) enhances the small signals and compresses the
large signals at the same time which is desirable when compared to the µ  law companding scheme
which enlarges the small signal and ignores the signal peaks. Hence the nonlinear companding scheme
attains better PAPR reduction, BER, Power Spectrum and Phase error than the µlaw companding
scheme. The main liability of EC scheme is that its performance remains unchanged for different levels
of companding.
Jiang et al. (2006) inferred that at the receiver side, a companded signal when undergoes only
inverse companding transform (ICT) the resultant spectrum exhibits severe out of band, inband
distortion and peak regrowth due to excessive channel noise. Hence they proposed an iterative receiver
with slight increase in complexity to cancel out the channel noise and companding noise. Later in
2007, they proposed two novel nonlinear companding functions for MCM signals to transform the
Gaussiandistributed OFDM signal into a Trapezoidaldistributed signal. By proper selection of
parameters this method achieves PAPR reduction and maintains average power constant.
The two novel companding functions introduced by Jiang et al. are
A Survey on CF Method, PTS Approach, Companding Technique and
Time Domain Methods for PAPR Reduction in OFDM Systems 633
1 0 ,
2
. 3 ) (
1
≤ ≤


¹

\

σ
σ = x
x
erf x C is a LNST (10)


¹

\

σ
σ =
2
. 2 ). ( ) (
2
x
erf x sgn x C is a NLNST (11)
Where
n n n
dt t p t
h
h
) ( ) (
2
1
1 2
−
∫ = σ , maintains average power constant.
LNST scheme in eqn. (10), C
1
(x) achieves better tradeoff between BER and PAPR reduction
while NLNST scheme in eqn. (11), C
2
(x) provides result consistent with non linear companding
scheme for PAPR reduction. Table 4 compares the PAPR reduction and BER performance of the
Jiang’s method with that of the EC scheme.
Table 4: PAPR reduction and BER performance comparison of the Jiang’s method and EC scheme.
PAPR (dB)
PAPR (dB)
PAPR (dB)
Original OFDM 6.98 8.58 10.80
EC 9.67 11.42 4.80
EC* 7.43 9.08
Jiang’s method 7.90 9.43 4.25
Jiang’s method* 7.38 9.01
* without decompanding operation at the receiver
4.2. Linear Companding Scheme
Aburakhia et al., (2009) proposed a Linear Companding Transform (LCT) with two inflexion points to
scale different signal levels independent of one another. Based on the simulation results of his
proposed LST and NLST on an AWGN channel it was inferred that LCT performs better in terms of
PAPR reduction and BER performance. The average value of PAPR reduction is 50% for LNST and
70% for the proposed LCT.
Hou et al. (2009) introduced a Companding Scheme based on transformation of a Gaussian
distributed OFDM signal to a Trapezoidaldistributed signal with proper selection of parameters
achieves a better PAPR reduction and BER performance than the EC scheme.
Jiang (2010) proposed a new Companding Transform based on a smooth function called airy
special function which is given by
[ ] ) . ( ) 0 ( ). ( . ) ( x a airy airy x sign x f − β = (12)
In equation (12), airy(.) is an airy function of first kind
α – parameter to control the degree of companding
[ ]
(
¸
(
¸
−
= β
2
2
. ( ) 0 ( x a airy airy E
x E
maintains the average signal power constant.
The decompanding function is given by
(
¸
(
¸
β
− =
− −
x
airy airy x sign
x
x f ) 0 ( ). ( .
1
) (
1 1
(13)
Eqn. (13) is computed based on Lookup table.
The proposed Transform based on airy function is flexible, the degree of companding is varied
by changing α which is reflected in the performance and this feature overcomes the liability of EC
scheme. The performance comparison of the Airy function based CT, EC technique and µlaw CT on
634 Arvind Chakrapani and V. Palanisamy
four times oversampled 64 QPSK modulated OFDM signal in terms of Out of Band Interference
(OBI), PAPR and Signal to Noise Ratio (SNR) is given in Table 5.
Table 5: Performance Comparison of the Airy function based CT, EC scheme and µlaw CT.
Parameters Airy function based CT EC Scheme µlaw CT
OBI (dB) 47.5 45 39.7
PAPR (dB) 3.88 2.37 5.78
(dB) 8.9 10.4 11.7
* required to reach a BER of
While Hou et al. (2010) introduced a companding scheme which mainly compresses only the
large signals and even without decompanding function it maintains a good BER performance. The
PAPR reduction capability and Power spectrum obtained based on this method is superior to that of EC
scheme.
Companding Transform can be used to limit PAPR statistics for arbitrary number of carriers,
irrespective of the frame format and constellation type [28]. Companding schemes achieve good PAPR
and BER performance with reduced complexity and no bandwidth expansion [29].
5. Time Domain Based PAPR Reduction Techniques
PAPR reduction is performed by employing some properties or methods in time domain to the original
OFDM signal so that the alternate candidate signals are generated and the candidate with lowest PAPR
is selected for transmission.
Lu et al. (2007) generated the candidate signals in time domain directly by computing the
product of circular convolution of the OFDM data and IFFT of optimized cyclically shifted phase
sequences (OCSPS). Table 6 compares the PAPR reduction performance of OCSPS with SLM and
PTS.
Table 6: Comparison of various PAPR reduction methods on a OFDM signal with 1024 subcarriers
Method Number of Main Computations PAPR (dB)
Subblocks candidates
SLM = 4
10.14
PTS 2
10.4
4 9.08
OCSPS L = 4
One Npoint IFFT 9.2
Where N refers to number of subcarriers and S refers to finite set of each element.
While a Low Complexity(LC) time domainbased PAPR reduction technique by Alsusa and
Yang, (2008) uses a linear symbol combining technique to consecutive OFDM symbols to create
several time domain representation of each OFDM symbol at the transmitter. The scheme requires one
IFFT block per OFDM symbol while PTS requires N IFFT blocks per OFDM symbol where N is the
number of sub sets. But Forward Error Correction(FEC) coding has to be performed on side
information to overcome the noise in the Channel, and to improve the BER performance at low SNR
values, which demands additional processing at the receiver. The BER performance is slightly reduced
due to its dependency on side information, symbols and multiblock combination.
Yang et al., (2008) proposed a LC – SLM scheme using Time Domain Sequence
Superposition(TDSS) to limit PAPR of an OFDM signal. Here two phase sequences generated are
multiplied with the input symbol to produce two intermediate sequences. Of this one time domain
sequence is fixed and linearly combined with the cyclically shifted versions of the other sequence to
A Survey on CF Method, PTS Approach, Companding Technique and
Time Domain Methods for PAPR Reduction in OFDM Systems 635
produce new candidates. The computational complexity comparison of TDSS and SLM scheme is
presented in Table 7.
Table 7: Computational complexity of Bauml et al. SLM scheme and LC – SLM TDSS scheme.
Operation Bauml et al. SLM Scheme LC – SLM TDSS scheme
Number of Multiplications
Number of Additions
When M = 4, 8, 16 and 32
PAPR reduction better than LC – SLM
TDSS scheme
PAPR is 0.2 dB less than the
conventional SLM scheme
Where M refers to the number of phase sequences.
A selective time domain filtering technique suggested by Du et al., (2009) employs a filter bank
to generate candidate signals with different PAPR. Since scrambling is performed in time domain, the
need for additional IFFT block is eliminated.
Wang et al., (2009) linearly combined the OFDM signal and its cyclically shifted version with
various allowable phase and time delays to produce candidate signals. The scheme has reduced
complexity than that of SLM scheme with some degradation in BER performance.
6. Overview of Survey and Scope for Future Work
The comparative analysis on Iterative clipping and filtering technique, PTS Approach based on DFT
property, Algorithm assisted PTS scheme, Companding schemes and time domain based methods are
presented in Table 8, Table 9, Table 10, Table 11 and Table 12 respectively along with their scope for
future work.
Table 8: Analysis of Iterative Clipping and Filtering Technique for PAPR reduction.
Author Methodology Merits Demerits Remarks
Armstrong
(2002)
Repeated clipping
and frequency
domain filtering
No increase in out
of band radiation
Two FFT operations
used
Peak regrowth
reduced
Moderate level of
clipping noise is
generated due to
distortion of in band
frequency components
Tradeoff between PAPR
reduction and BER
performance is not
addressed
Leung et al.,
(2002)
Repeated clipping
and time domain
filtering
Requires no FFT
operation
Clipping threshold is
increased by 5%
Deng et al.
(2007)
Distortion bounded
for each iteration
Reduces error rate
Demands more reserved
tones to achieve better
error rate and PAPR
reduction
Occurrence of Constellation
shrinkage is not considered
Wang and Luo
(2011)
Convex
optimization
technique
PAPR reduction is
achieved in 1 or 2
iterations
Computational
complexity is high.
Processed OFDM symbols
are distortionless and
provides better out of band
radiation
Baxley et al.
(2010)
Noise enhancement
factor
Estimates BER,
EVM and
attenuation factor
Suppression of out of
band radiation by filtering
leads to peak regrowth
Reports tradeoff between
BER performance and
PAPR reduction
636 Arvind Chakrapani and V. Palanisamy
6.1. Scope for Future Work I
The proposed PAPR reduction algorithm based on ICF technique will combine the positive features
like time domain filtering, distortion bounded for each iteration, expressing PAPR problem as a convex
optimization problem so that FFT operation is eliminated, BER is reduced and number of iterations is
drastically reduced while simultanously concentrating on clipping threshold, number of reserved tones
and computational complexity. This method is expected to achieve a significant tradeoff between
PAPR reduction, BER performance and reduction in computational complexity by reducing the
number of iterations.
Table 9: Analysis of PTS approach based on DFT property for PAPR reduction.
Author Methodology Merits Demerits Remarks
Lu et al.,
(2006)
Transformation of
PTS using DFT
property
Complexity reduced
compared to PTS
scheme for same
number of sub blocks
Complexity is same as
that of PTS for same
number of candidates
Only PAPR
reduction is
considered, BER
performance is
ignored
Wang and Cao
(2008)
Linearity property of
IDFT
Complexity less
compared to Lu and
PTS
Oversampled OFDM
signal is used to
capture the peaks
Zhu et al.,
(2008)
Five novel
transformation based
on Greedy algorithm
Performance better
than iterative flipping
algorithm for small
number of sub blocks
Performance is poor
for large number of
sub blocks
Yang et al.,
(2011)
Phase constellation
varied
PAPR reduction and
BER performance
similar to that of PTS
For large M, the
PAPR reduction
performance is
degraded.
For M sub blocks,
independant
candidates are
produced
6.2. Scope for Future Work II
It is planned to generate candidates of a OFDM signal by combining the linearity and circular time
shift property, with a slight increase in computational complexity this method is expected to achieve
significant PAPR reduction. Since the DFT property based transformations are less complex, when
different sub block undergo different transformations, the additional candidates generated can out
perform the conventional PTS scheme with less candidates.
Table 10: Analysis of Algorithm assisted PTS schemes for PAPR reduction.
Author Methodology Merits Demerits Remarks
Chen
(2009)
Cross Entropy
Algorithm
Performance similar
to ESA with low
complexity
OFDM system with
64, 128 subcarriers
and QPSK modulated
signals are considered
for simulation.
Tradeoff between
PAPR and BER
performance is
ignored.
Chen
(2010a)
Quantum Inspired
Evolutionary
Algorithm
PAPR reduction better
than conventional
PTS with low
computational
complexity
Chen
(2010b)
Electromagnetism –
like (EM) Algorithm
Unlike Genetic
Algorithm based –
PTS approach,
Encoding and
decoding process is
not required.
At least 1200 samples
is needed to achieve
better PAPR reduction
performance
These schemes are
simulated for a
specific number of
subcarriers and
modulation type.
A Survey on CF Method, PTS Approach, Companding Technique and
Time Domain Methods for PAPR Reduction in OFDM Systems 637
Table 10: Analysis of Algorithm assisted PTS schemes for PAPR reduction.  continued
Wang et al.,
(2010)
Artificial Bee Colony
(ABC) Algorithm
PAPR reduction better
than conventional
PTS with low
computational
complexity
OFDM system with
256 subcarriers and 16
QAM modulated
symbols are
considered for
simulation.
Taspinar et al.,
(2011)
Parallel Tabu search
Algorithm
Better PAPR
reduction and BER
performance
compared to
conventional PTS and
ABC algorithm.
Input Back off (IBO)
value of HPA and its
relation to the BER
performance is
mentioned.
6.3. Scope for Future Work III
a) The PAPR problem is expressed as a optimization problem and different algorithms are
employed to assist the PTS scheme. Hence it is planned to represent the PAPR problem as a
combinatorial optimization problem and to find the optimal phase factors using a combinatorial
algorithm like Differential Evolution which is expected to achieve better PAPR statistics and
BER performance.
b) To assign the permissible phase rotation factors generated based on PTS approach to the nodes
of a tree structure and to propose an algorithm which performs a search to compute the optimal
factors effectively such that PAPR reduction is achieved.
Table 11: Analysis of Companding Techniques for PAPR reduction.
Author Methodology Merits Demerits Remarks
Wang et al.,
(1999)
Combines the
advantage of clipping
and companding
scheme
Performs better than
clipping method
Average input power
increases
Ignored the non –
linear operation of
High power
amplifier(HPA)
Spectral side lobes are
generated.
Requires large dynamic
range for HPA
Companded OFDM
signal exhibits
quasi – Gaussian
distribution.
Huang et al.,
(2001)
Transformation is
based on Power
distribution
Performs better than
CF scheme with
reduced complexity
Slight degradation in
PAPR reduction
performance
HPA operates in
linear region
Tradeoff between
PAPR and BER is
ignored
Huang et al.,
(2004)
Inflexion point treats
the large and small
amplitude signals
differently
Tradeoff between
PAPR reduction and
BER performance is
achieved
NLST companding
scheme performs better
than LST, LNST,
NLNST and µ  Law
companding scheme
The efficiency of
Power amplifier is
increased with
decreased power
back off but no
evidence for
increased back off
638 Arvind Chakrapani and V. Palanisamy
Table 11: Analysis of Companding Techniques for PAPR reduction.  continued
Jiang et al.,
(2005)
Gaussian distributed
Amplitude statistics is
transformed into
uniformly distributed
signal
Achieves better
PAPR reduction,
BER performance,
power spectrum and
phase error than the
µ  Law
Companding
Technique.
Average input power is
maintained constant
PAPR reduction
remains unchanged
for different levels
of companding
Uniform companding
increases the distribution
of large amplitude
signals which in turn
degrades the BER
performance when HPA
operates in non linear
region.
Aburakhia et al.,
(2009)
Two inflexion points
for scaling different
signals independently
PAPR reduction and
BER performance
better than NLST
When input signal
crosses the inflexion
threshold, the
transformed signal
jumps abruptly and
results in degraded
Power Spectral Density.
Uses 5 control
parameters while
NLST uses just two
parameters.
Hou et al.,
( 2009)
Transforms Gaussian
distributed signal to
Trapezoidal
distribution
PAPR reduction
performance better
than EC scheme.
BER is degraded by 0.5
dB
Proper selection of
companding
parameters
provides effective
tradeoff between
PAPR reduction
and BER
performance.
Jiang
( 2010)
CT using airy special
function
PAPR reduction
performance better
than µ  Law
companding scheme.
PAPR reduction is 1.5
dB inferior to EC
scheme
Degree of
companding is
reflected in the
PAPR reduction
performance.
Better BER and out
of band radiation
performance than
EC and µ  Law CT.
Hou et al.,
(2010)
Mainly compresses
only large signals
PAPR and BER
performance better
than EC scheme.
BER performance is
slightly degraded
compared to the original
OFDM signal.
Even without
decompanding
operation better
BER performance
is achieved.
6.4. Scope for Future Work IV
a) The existence of Gaussian transform and its ability to compress Gaussian distributed signals is
suggested by Alecu et al.,(2006) hence PAPR reduction of an OFDM signal by employing
Gaussian Transform in companding scheme is feasible. Since the amplitude of an OFDM signal
is Gaussian distributed, this proposal is expected to achieve better tradeoff between PAPR
reduction and BER performance.
b) In Companding schemes, the Gaussian distributed OFDM signal is transformed into uniform
distribution by Jiang et al., (2005) and Trapezoidal distribution by Hou et al., (2009) to reduce
PAPR of OFDM signal. The above mentioned transformation opens avenues to find other
possible distribution based transformation methods which can be applied to OFDM signal so
that PAPR reduction is achieved without compromising on BER performance and Power
Spectral Density.
A Survey on CF Method, PTS Approach, Companding Technique and
Time Domain Methods for PAPR Reduction in OFDM Systems 639
Table 12: Analysis of Time domain based PAPR reduction methods.
Author Methodology Merits Demerits Remarks
Alsusa et al.,
(2008)
Linear Symbol
combining
Technique
One FFT operation /
OFDM symbol
BER performance is
slightly degraded
Achieves significant PAPR
reduction
Generates large number
of candidates
Latency is more due to
large number of symbol
storage
Number of multiplications
is reduced drastically
compared to PTS, SLM and
TR methods.
Yang et al.,
(2008)
SLM based
sequence
superposition
Two FFT operation /
OFDM symbol hence
computational
complexity is less
Inferior BER
performance compared
to conventional SLM
PAPR reduction is degraded
by about 0.2 dB compared
to conventional SLM
scheme
Du et al.,
(2009)
SLM based
Time domain
filtering
No FFT operation and
complex multiplication
required
PAPR reduction is
slightly degraded by 0.1
– 0.75 dB compared to
conv. SLM scheme
Demands one third to one
fourth of the number of
computations compared to
conventional SLM scheme
Requires only complex
additions
Wang et al.,
(2010)
Linear
combination of
cyclically
delayed signals
Single FFT operation
required
PAPR reduction is
slightly degraded
around 0.2 dB
compared to SLM
Requires half the
complexity compared to LC
– SLM scheme for similar
PAPR reduction and BER
performance
Requires additional
memory to store the
mapping sequences
6.5. Scope for Future Work V
Time domain transformations such as complex conjugation, time scaling, time reversal or their
combinations can be employed with existing time domain based PAPR reduction methods[54, 56, 57]
so that additional candidates can be generated and the candidate signal with lowest PAPR is selected
for transmission.
7. Conclusion
Most of the work discussed in the literature concentrates on parameters like PAPR reduction,
computational complexity, in band radiation, out of band radiation, peak regrowth, BER performance
and EVM in OFDM systems individually but an effective tradeoff between PAPR reduction and BER
performance which contributes to the overall betterment of the system is rarely addressed. In future
work, it is planned to perform research on the above mentioned proposals and to present a comparative
study based on their result while maintaining the PAPR reduction and its corresponding BER
performance.
References
[1] Aburakhia S. A., Badran E. F. and Mohamed D. A. E., 2009. “Linear Companding Transform
for the reduction of peaktoaverage power ratio of OFDM signals”, IEEE transactions on
Broadcasting, Vol. 55, No. 1, pp. 155160.
640 Arvind Chakrapani and V. Palanisamy
[2] Alecu T. I., Voloshynovskiy S., 2006. “The Gaussian Transform of Distributions: definition,
computation and application,” IEEE transactions on Signal Processing, Vol. 54, No. 8, pp. 2976
– 2985.
[3] Alsusa E. and Yang L., 2008. “A LowComplexity TimeDomain Linear Symbol Combining
Technique for PAPR Reduction in OFDM Systems”, IEEE transactions on Signal Processing,
Vol. 56, No. 10, pp. 4844  4855.
[4] Armstrong J., 2002. “Peaktoaverage power ratio for OFDM by repeated clipping and
frequency domain filtering”, Electronics Letters, Vol. 38, No. 8, pp. 246247.
[5] Bae K., Andrews J. G. and Powers E. J., 2010. “Adaptive active constellation extension
algorithm for peaktoaverage ratio reduction in OFDM”, IEEE Communication Letters, Vol.
14, No. 1 pp. 3941.
[6] Baxley R. J., Zhao C. and Zhou G. T., 2006. “Constrained Clipping for Crest Factor Reduction
in OFDM”, IEEE Transactions on Broadcasting, Vol. 52, No. 4, pp. 570575.
[7] Bauml R. W., Fischer R. F. H. and Huber J. B., 1996. “Reducing the peaktoaverage power
ratio of multicarrier modulation by selected mapping”, IEEE Electronics Letters, Vol. 32, No.
22, pp. 20562057.
[8] Chen H. and Pottie G. J., 2002. “An Orthogonal Projectionbased approach for PAR reduction
in OFDM”, IEEE Communications Letters, Vol. 6, No. 5, pp.169171.
[9] Chen J. –C., 2009. “Partial Transmit Sequences for PeaktoAverage Power Ratio Reduction of
OFDM Signals with the CrossEntropy Method”, IEEE Signal Processing Letters, Vol. 16, No.
6, pp. 545 – 548.
[10] Chen J. –C., 2010a. “Application of QuantumInspired Evolutionary Algorithm to Reduce
PAPR of an OFDM Signal Using Partial Transmit Sequences Technique”, IEEE transactions on
Broadcasting, Vol. 56, No. 1, pp. 110 – 113.
[11] Chen J. –C., 2010b. “Partial Transmit Sequences for PAPR Reduction of OFDM Signals with
Stochastic Optimization Techniques”, IEEE Transactions on Consumer Electronics, Vol. 56,
No. 3, pp.1229 – 1234.
[12] Cimini Jr. L. J., and Sollenberger N. R., 2000. “Peaktoaverage power reduction of an OFDM
signal using Partial Transmit Sequences”, IEEE Communications Letters, Vol. 4, No. 3, pp. 86
88.
[13] Deng S. –K. and Lin M. –C., 2007. “Recursive Clipping and Filtering With Bounded Distortion
for PAPR Reduction”, IEEE Transactions on Communications, Vol. 55, No. 1, pp. 227230.
[14] Du Z., Beaulieu N. C. and Zhu J., 2009. “Selective TimeDomain Filtering for Reduced
Complexity PAPR Reduction in OFDM”, IEEE transactions on Vehicular Technology, Vol. 58,
No. 3, pp. 1170 – 1176.
[15] Eetvelt P. V., Wade G. and Tomlinson M., 1996. “Peak to average power ratio reduction for
OFDM schemes by selective scrambling”, Electronics Letters, Vol. 32, pp. 19631964.
[16] Fischer R. F. H. and Siegl C., 2009. “ReedSolomon and Simplex Codes for Peaktoaverage
power ratio reduction in OFDM”, IEEE transactions on Information Theory, Vol. 55, No. 4, pp.
1519–1528.
[17] Ghassemi A. and Gulliver T. A., 2010. “PAPR reduction of OFDM using PTS and error
correcting code subblocking”, IEEE transactions on Wireless Communications, Vol. 9, No. 3,
pp. 980989.
[18] Han S. H. and Lee J. H., 2004. “PAPR reduction of OFDM signals using reduced complexity
PTS technique”, IEEE Signal Processing Letters, Vol. 11, No. 11, pp. 887890.
[19] Hill G. R., Faulkner M. and Singh J., 2000. “Reducing the peaktoaverage power ratio in
OFDM by cyclically shifting partial transmit sequence,” Electronics Letters, Vol. 36, No. 6, pp.
560561.
[20] Hong E. and Har D., 2010. “Peaktoaverage power ratio reduction in OFDM Systems using
AllPass Filters”, IEEE transactions on Broadcasting, Vol. 56, No. 1, pp. 114119.
A Survey on CF Method, PTS Approach, Companding Technique and
Time Domain Methods for PAPR Reduction in OFDM Systems 641
[21] Hou J., Ge J. H. and Li J., 2009. “Trapezoidal companding scheme for peaktoaverage power
ratio reduction of OFDM signals”, Electronics Letters, Vol. 45, No. 25, pp. 1349  1351.
[22] Hou J., Ge J., Zhai D. and Li J., 2010. “PeaktoAverage Power Ratio Reduction of OFDM
Signals with Nonlinear Companding Scheme”, IEEE transactions on Broadcasting, Vol. 56, No.
2, pp. 258 – 262.
[23] Huang X., Lu J., Zheng J., Chuang J. and Gu J., 2001. “Reduction of peaktoaverage power
ratio of OFDM signals with companding transforms”, IEEE Electronics Letters, Vol. 37, pp.
506507.
[24] Huang X., Lu J., Zheng J., Letaief K. B. and Gu J., 2004. “Companding Transform for
reduction in peaktoaverage power ratio of OFDM signals”, IEEE transactions on Wireless
Communications, Vol. 03, No. 6, pp. 20302039.
[25] Jiang T. and Zhu G., 2004. “Nonlinear companding transform for reducing peaktoaverage
power ratio of OFDM signals”, IEEE transactions on Broadcasting, Vol. 50, No. 3, pp. 342
346.
[26] Jiang T., Yang Y. and Song Y., 2005. “Exponential companding transform for PAPR reduction
in OFDM systems”, IEEE transactions on Broadcasting, Vol. 51, No. 2, pp. 244248.
[27] Jiang T., Yao W. B., Guo P. and Qu D. M., 2006. “Two novel nonlinear companding schemes
with iterative receiver to reduce PAPR in multicarrier modulation systems”, IEEE transactions
on Broadcasting, Vol. 52, No. 2, pp. 268273.
[28] Jiang T., Xiang W., Richardson P. C., Qu D., and Zhu G., 2007. “On the Nonlinear
Companding Transform for reduction in PAPR of MCM Signals”, IEEE transactions on
Wireless Communications, Vol. 06, No. 6, pp. 2017  2021.
[29] Jiang T. and Wu Y., 2008. “An overview: PeaktoAverage power ratio reduction techniques
for OFDM signals,” IEEE transactions on Communications, Vol. 54, No. 2, pp. 257268.
[30] Jiang Y., 2010. “New Companding Transform for PAPR Reduction in OFDM”, IEEE
Communication Letters, Vol. 14, No. 4, pp. 282 – 284.
[31] Kang S. G., Kim J. G. and Koo E. K., 1999. “A novel sub block partition scheme for partial
transmit scheme OFDM”, IEEE transactions on Broadcasting, Vol. 65, No. 3, pp. 333338.
[32] Leung S. H., Ju S.M. and Bi G.G., 2002. “Algorithm for repeated clipping and filtering in peak
toaverage power reduction for OFDM”, Electronics Letters, Vol. 38, No. 25, pp. 17261727.
[33] Li X. and Cimini Jr. L. J., 1998. “Effects of clipping and filtering on the performance of
OFDM”, IEEE Communication Letters, Vol. 2, No. 5, pp. 131133.
[34] Lixia M. and Murroni M., 2011. “Peaktoaverage power ratio reduction in multicarrier system
using genetic algorithms”, IET Signal Processing,Vol. 5, No. 3, pp. 356–363.
[35] Lu G., Wu P. and Logothetis C., 2006. “Peaktoaverage power ratio reduction in OFDM based
on transformation of partial transmit sequences”, IEEE Electronics Letters, Vol. 42, No. 2, pp.
105106.
[36] Lu G., Wu P. and Aronsson D., 2007. “Peaktoaverage power ratio reduction in OFDM using
cyclically shifted phase sequences”, IET Communications, Vol. 1, No. 6, pp. 1146 – 1151.
[37] Mattsson A., Mendenhall G. and Dittmer T., 1999. “Comments on: Reduction of peakto
average power ratio of OFDM system using a companding technique, Vol . 45, No. 3, Sept.
1999”, IEEE transactions on Broadcasting, Vol. 45, No. 4, pp. 418419.
[38] Muller S. H. and Huber J. B., 1997. “OFDM with reduced peaktoaverage power ratio by
optimum combination of partial transmit sequences”, Electronics Letters, Vol. 33, No. 5, pp.
368369.
[39] Nguyen T. T. and Lampe L., 2008. “On partial transmit sequence for PAR reduction in OFDM
systems,” IEEE transactions on Wireless Communications, Vol. 7, No. 2, pp. 746755.
[40] Prasad R., OFDM for Wireless Communications Systems. Artech House, Inc., 2004
642 Arvind Chakrapani and V. Palanisamy
[41] Schenk T. C. W., Smulders P. F. M. and Fledderus E. R., 2005. “Peaktoaverage power
reduction in space division multiplexing based OFDM systems through spatial shifting”,
Electronics Letters, Vol. 41, No. 15. pp. 860861.
[42] Tasadduq I. A. and Rao R. K., 2002. “PAPR reduction of OFDM signals using multiamplitude
CPM”, Electronics Letters, Vol. 38, No. 16, pp. 915 – 917.
[43] Taspinar N., Kalinli A. and Yildirim M., 2011. “Partial Transmit Sequences for PAPR
Reduction using Parallel Tabu Search Algorithm in OFDM Systems”, IEEE Communication
Letters, Vol. 15, No. 7, pp. 974 – 976.
[44] Tellambura C., 1998. “Phase Optimization criterion for reducing peak to average power ratio in
OFDM”, Electronics Letters, Vol. 43, No. 2, pp. 169170.
[45] Tellambura C., 2001. “Improved phase factor computation for PAR reduction of an OFDM
signal using PTS”, Communication Letters, Vol. 5, No. 4, pp. 135–137.
[46] Wang J. S., Lee J. H., Park J. C., Song I. and Kim Y. H., 2010. “Combining of Cyclically
Delayed Signals: A LowComplexity Scheme for PAPR Reduction in OFDM Systems”, IEEE
transactions on Broadcasting, Vol. 56, No. 24, pp. 577 – 583.
[47] Wang L. and Tellambura C., 2005. “A simplified clipping and filtering technique for PAR
reduction in OFDM systems”, IEEE transactions on Signal Processing, Vol. 12, No. 6, pp. 453
456.
[48] Wang L. and Cao Y., 2008. “Sub – optimum PTS for PAPR reduction of OFDM signals”,
Electronics Letters, Vol. 44, No. 15, pp. 921 – 922.
[49] Wang X., Tjhung T. T. and Ng C.S., 1999. “Reduction of peaktoaverage power ratio of
OFDM system using a companding technique”, IEEE transactions on Broadcasting, Vol. 45,
No. 3, pp. 303307.
[50] Wang X., Tjhung T. T., and Ng C. S., 1999 “Reply to comments on Reduction of peakto
average power ratio of OFDM system using a companding technique, Vol . 45, No. 3, Sept.
1999”, IEEE transactions on Broadcasting, Vol. 45, No. 4, pp. 420422.
[51] Wang Y., Chen W. and Tellambura C., 2010. “A PAPR Reduction Method Based on Artificial
Bee Colony Algorithm for OFDM Signals”, IEEE transactions on wireless communications,
Vol. 9, No. 10, pp. 2994 – 2999.
[52] Wang Y., Chen W. and Tellambura C., 2010. “PAPR reduction method based on Parametric
minimum Cross entropy for OFDM signals”, IEEE Communication Letters, Vol. 14, No. 6, pp.
563565.
[53] Wang Y. –C. and Luo Z. –Q., 2011. “Optimized iterative clipping and filtering for PAPR
reduction of OFDM signals”, IEEE transactions on Communications, Vol. 59. No. 1, pp.3337.
[54] Wulich D., 1996. “Reduction of peak to mean ratio of multicarrier modulation using cyclic
coding”, Electronics Letters, Vol. 32, No. 5, pp. 432433.
[55] [55] Xiao Y., Lei X., Wen Q. and Li S., 2007. “A class of low complexity PTS techniques
for PAPR reduction in OFDM systems”, IEEE Signal Processing Letters, Vol. 14, No. 10, pp.
680  683.
[56] Yang L., Soo K. K., Siu Y. M. and Li S. Q., 2008. “A low complexity selected mapping
scheme by use of time domain sequence superposition technique for PAPR reduction in OFDM
systems”, IEEE transactions on Broadcasting, Vol. 54, No. 4, pp. 821824.
[57] Yang L., Soo K. K., Li S. Q. and Siu Y. M., 2011. “PAPR Reduction Using Low Complexity
PTS to Construct of OFDM Signals Without Side Information”, IEEE transactions on
Broadcasting, Vol. 57, No. 2, pp. 284–290.
[58] Zhu X., Jiang T. and Zhu G., 2008 “Novel Schemes Based on Greedy Algorithm for PAPR
Reduction in OFDM Systems”, IEEE Transactions on Consumer Electronics, Vol. 54, No. 3,
pp. 1048–1052.