812
3, MARCH 2010
Intercarrier Interference Cancellation Using
General Phase Rotated Conjugate Transmission for
OFDM Systems
Chin-Liang Wang, Senior Member, IEEE, and Yu-Chih Huang
AbstractIn this paper, we propose a general phase ro-
tated conjugate cancellation (PRCC) scheme for intercarrier
interference (ICI) cancellation in orthogonal frequency division
multiplexing (OFDM) systems. We also derive the optimal
phase rotation for this general scheme such that the carrier-
to-interference ratio would be maximized. It is shown that the
previous conjugate cancellation (CC) scheme is equivalent to a
special case of our proposed scheme. The general PRCC scheme
not only inherits advantages of the conventional CC scheme, such
as backward compatibility with the existing OFDM systems, low
receiver complexity, and two-path diversity, but also provides
better performance, especially at high frequency offset situations.
Index TermsFrequency offset, intercarrier interference (ICI),
ICI self-cancellation, orthogonal frequency division multiplexing
(OFDM).
I. I NTRODUCTION
O RTHOGONAL frequency division multiplexing
(OFDM) transmission [1] achieves high spectral
efficiency by splitting a high-rate data stream into several
lower-rate streams which are mapped onto mutually
overlapping subcarriers and transmitted simultaneously. By
inserting a cyclic prefix of length larger than the maximum
channel delay spread, the intersymbol interference can be
easily avoided and the demodulation of OFDM signals can be
realized by a one-tap equalizer. Motivated by these attractive
properties, OFDM has been widely adopted in transmission
standards such as Digital Video Broadcasting (DVB) [2],
Wireless LAN [3], WiMAX [4], and so on. Compared to
the conventional single-carrier systems, the OFDM systems
allowing spectral overlapping in an orthogonal way are highly
sensitive to frequency offset which is often caused by the
Paper approved by S. Batalama, the Editor for Spread Spectrum and
Estimation of the IEEE Communications Society. Manuscript received June
13, 2008; revised January 5, 2009 and May 26, 2009.
C.-L. Wang is with the Department of Electrical Engineering and the
Institute of Communications Engineering, National Tsing Hua University,
Hsinchu, Taiwan 30013, Republic of China (e-mail: clwang@ee.nthu.edu.tw).
Y.-C. Huang was with the Institute of Communications Engineering,
National Tsing Hua University, Hsinchu, Taiwan 30013, Republic of China.
He is now with the Department of Electrical and Computer Engineering,
Texas A&M University, College Station, TX 77843, USA (e-mail: cy-
eekk51@gmail.com).
This work was supported by the National Science Council of the Republic
of China under Grants NSC 95-2221-E-007-041, NSC 96-2628-E-007-005-
MY3, and NSC 97-2219-E-007-007. This paper was presented in part at the
2008 IEEE Wireless Communications and Networking Conference, Las Vegas,
Nevada, March 2008.
Digital Object Identifier 10.1109/TCOMM.2010.03.080288
0090-6778/10$25.00
Authorized licensed use limited to: Jeppiaar Engineering College. Downloaded on May 01,2010 at 03:49:16 UTC from IEEE Xplore. Restrictions apply.
IEEE TRANSACTIONS ON COMMUNICATIONS, VOL. 58, NO.
Doppler shift and/or mismatch between oscillators in the
transmitter and receiver. The frequency offset may damage
the orthogonality among OFDM subcarriers and introduce
intercarrier interference (ICI), which degrades the overall
system performance [5], [6].
The ICI effect can be effectively mitigated by the conven-
tional ICI self-cancellation schemes in [7] and [8] at the cost
of halving the spectral efficiency. One major problem reported
by [9] associated with this kind of methods is that the bit
error rate (BER) performance is limited due to the constant
phase error generated at the coherent detection process. The
approach proposed in [9] can avoid the constant phase error
problem, but it might not provide satisfactory performance at
high frequency offset situations. Another useful methods for
ICI suppression are based on performing correlative coding
before the inverse fast Fourier transform (IFFT) block of
an OFDM transmitter [10]-[13]. Correlative coding has also
been applied in OFDM for spectral shaping [14] and/or other
purposes (see, e.g., [14] and references therein). Despite the
ICI suppression ability and the versatility of the correlative-
coded OFDM systems, the error propagation problem makes
them more suitable for BPSK transmission. Recently, two-path
conjugate transmission was proposed for ICI self-cancellation
[15], [16], where the first path represents the standard OFDM
signal and the second one is formed by a conjugate of the first
paths. Such an ICI self-cancellation scheme was referred to as
conjugate cancellation (CC) in [16]. Similar to the approach
in [9], the CC scheme could not guarantee good performance
for high frequency offset situations.
In this paper, we propose a general phase rotated conjugate
cancellation (PRCC) scheme for ICI cancellation in OFDM
systems. We generalize the conventional CC scheme by in-
troducing an artificial phase rotation which is determined by
the frequency offset estimate in the training mode. To achieve
better performance, the optimal phase rotation is derived with
the criterion of maximizing the carrier-to-interference ratio
(CIR). Furthermore, we modify the proposed scheme for appli-
cations in a user cooperation environment. As compared to the
conventional CC scheme, the PRCC scheme achieves better
performance at both low and high frequency offset situations,
with only a little increase in the transmitters complexity.
The rest of the paper is organized as follows. Section II
describes the OFDM system model and the conventional CC
scheme proposed in [16]. In Section III, the general PRCC
scheme is presented and its application for a user cooperation
c 2010 IEEE
WANG and HUANG: INTERCARRIER INTERFERENCE CANCELLATION USING GENERAL PHASE ROTATED CONJUGATE TRANSMISSION . . . 813

protocol is discussed. Besides, optimal phase rotations of the
proposed scheme are also derived in this section. To evaluate
the proposed approach, a number of computer simulation
results are shown in Section IV. Finally, some conclusions
are given in Section V.
II. BACKGROUND
A. OFDM System Model
The complex baseband OFDM signal after the IFFT block
at the transmitter can be expressed as
1
1 Unlike other ICI self-cancellation algorithms employing
= 2/ , = 0, 1, . . . , 1 (1)
=0
where is the total number of subcarriers and is the
transmitted symbol on the kth subcarrier. At the receiver, the
time-domain received signal suffering from a frequency offset
and an additive white Gaussian noise (AWGN) can be written
as
= 2/ + , = 0, 1, . . . , 1 (2)
where represents the frequency offset normalized by the
frequency spacing of two adjacent subcarriers and is a
zero-mean AWGN. After the fast Fourier transform (FFT)
block, the frequency domain signal at the lth subcarrier is
given as
information symbol is not only attenuated by (), but
also interfered by other 1 subcarriers. This problem cannot
be solved by simply increasing the transmission power because
both the desired and ICI terms will be increased if a higher
transmission power is used. Therefore, the ICI cancellation
researches focus on reducing the sensitivity of OFDM signals
to frequency offsets (i.e., to diminish or remove the effect of
the second term in (5)).
B. The Conventional CC Scheme
data repetition [7], [8] or conjugate data repetition [9] within
the same OFDM symbol interval, the CC scheme [16] trans-
mits two independent paths such that the ICI weighting
coefficient for one path could have an opposite polarity of that
for the other path. At the receiver, the time-domain received
signal of the first path is the same as (2) and the signal of the
second path can be expressed as
= 2/ + , = 0, 1, . . . , 1 (6)
where (.) denotes the conjugate operation and is a zero-
mean AWGN. By employing a conjugate operation on (6) and
passing the resulting signal through the FFT block, we have
1

1 = ( ) 2/

= 2/ =0

1
=0
1 = ( + ) + . (7)

2/ 2/
= ( + ) =0

=0 Averaging the FFT results of the two transmission paths yields

1
1
1 1
= 2()/ = ( + )
2
=0 =0
1
1

1 = { (( ) + ( + )) + ( + )}.
+ 2/ . (3) 2
=0
=0 (8)
Let be the discrete Fourier transform of and
The CIR of the conventional CC scheme can be defined as

1
1 () + ()2
() = 2/ = . (9)
=0

1
( ) + ( + )2
sin (1 ) =1
= exp( ). (4)
sin(/ )
Simulation results in [16] showed that the CC scheme achieves
Then (3) can be rewritten as follows: pretty high CIR values when the frequency offset is small.

1 However, the CIR curve of the CC scheme declines more
() + ( ) + , rapidly than those of other ICI cancellation schemes. This
=0,= property makes the CC scheme undesirable at high frequency
= 0, 1, . . . , 1. (5) offset situations.

() is usually referred to as the ICI weighting coefficient

function [7]. The first term in (5) contains the desired in-
formation with amplitude attenuated by a factor (),
and the second term is regarded as the sum of interferences
resulted from multiplied by ( ), the ICI weighting
coefficient from the kth subcarrier to the lth subcarrier. If there
is no frequency offset, since the orthogonality is preserved
(i.e., ( ) = 0), the second term of (5) is zero. However,
in the presence of frequency offset, one can observe that the
III. T HE P ROPOSED G ENERAL PRCC S CHEME
The purpose of the proposed scheme is to make the CC
scheme provide better performance at both low and high
frequency offset situations. Like the CC scheme, this scheme
adopts two-path transmission, but introduces a new variable,
. It allows the transmitter to tune the transmitted signals so
that the ICI effects of the two-paths signals could be mutually
cancelled at the receiver.

Authorized licensed use limited to: Jeppiaar Engineering College. Downloaded on May 01,2010 at 03:49:16 UTC from IEEE Xplore. Restrictions apply.
814 IEEE TRANSACTIONS ON COMMUNICATIONS, VOL. 58, NO. 3, MARCH 2010

A. The Algorithm B. The CIR and Optimal Phase Rotation

For the proposed general PRCC scheme, the first path em- From (14), the CIR of the proposed general PRCC scheme
ploys the standard OFDM signal with artificial phase rotation can be given as
of , while the second path adopts the conjugate of the 2 ( ) + ( + )2
standard OFDM signal with artificial phase rotation of , =
1
.
i.e., and ( ) , respectively. In the presence of a 2 ( ) + ( + )2
=1
frequency offset , the received signals of the two transmission (15)
paths can be expressed as If the average symbol energies, [ 2 ], are equal, (15) can
be reduced as
(1) = 2/ + (1) , = 0, 1, . . . , 1 (10)
() + ()2
=
1
. (16)
and ( ) + ( + )2
=1
(2) = 2/ + (2) , = 0, 1, . . . , 1 (11) Fig. 1 depicts theoretical CIR comparisons of the standard
OFDM, the conventional CC, and the proposed general PRCC
(1) (2)
where and are AWGN. scheme, where simulated CIR curves for the proposed scheme
(1) (2)
The receiver first passes and the conjugate of are also included to demonstrate the validity of (16). The
through the FFT block to obtain the following frequency- simulation environment for Fig. 1 is with QPSK-OFDM
domain signals symbols of = 1024 under an AWGN channel. Now, we
define the optimal phase rotation for a given to be a phase

1 rotation resulting in the maximum CIR value as follows:
(1) (1)
= {(1)} = { ()}+ (12)
=0 = arg max (, ), [/2, /2). (17)

and Proposition 1:The optimal phase rotation for a given is

1 1
(2)
(2)
= . (18)
= {(2)} = {
( + )} +
=0 Proof : See Appendix.
(13) According to (18), the frequency offset information is
(1) (2)
where and are the discrete Fourier transforms of needed at the transmitter for obtaining the optimal phase
(1) (2)
and , respectively. Taking an average of the results rotation; therefore, the receiver is required to estimate the
of (12) and (13), we obtain frequency offset and feed it back in the training mode. Note
that, for multipath Rayleigh fading channels, the frequency-
1 (1) (2)
= ( + ) domain received signals described in (12) and (13) should be
2 modified by multiplying a channel gain to each subcarriers
1
1 symbol in the summations. If perfect channel information is
= { ( ( ) + ( + ))
2 given, these channel gains can be compensated by a frequency-
=0
(1) (2) domain equalizer and (14)-(18) can be derived in the same
+ ( + )}. way. Therefore, the phase rotation given by (18) is also optimal
(14) for multipath Rayleigh fading channels under perfect channel
equalization. So far, the detail of the proposed general PRCC
It is worth noting that the combining method described scheme has been presented. Fig. 2 shows a block diagram of
in (14) is still suitable when the two independent paths the proposed scheme. It can be observed that the receiver of
signals experience similar fading channels. If the two paths the proposed scheme does not need any information of the
signals suffer from significantly different fading channels, artificial phase rotation; therefore, the receivers complexity is
under the assumption that two paths signals can be separated as low as the conventional CC schemes. As to the transmitter,
perfectly at the receiver, the maximum ratio combing (MRC) the proposed scheme has a little increase in the computational
can be adopted to take advantage of the two-path diversity for complexity due to the artificial phase rotation on the transmit-
further performance improvement. For this approach, instead ted signals.
of compensating the channel effects directly, we multiply
the received signal of each path with the conjugate of the
corresponding channel gain before combining. Moreover, it C. Applying the Proposed Scheme to a User Cooperation
can easily be observed that the conventional CC scheme is a Protocol
special case of the proposed scheme, where = 0. Therefore, Recently, realizing the spatial diversity through multiple
the proposed general PRCC scheme has advantages of the users relaying for each other has attracted great attention
conventional CC scheme, such as backward compatibility with [17]-[19]. The frequency synchronization problem in this
the existing OFDM systems, low receiver complexity, and two- scenario is trickier than a non-cooperative scenario because
path diversity. the transmission links from different relays to the destination

Authorized licensed use limited to: Jeppiaar Engineering College. Downloaded on May 01,2010 at 03:49:16 UTC from IEEE Xplore. Restrictions apply.
WANG and HUANG: INTERCARRIER INTERFERENCE CANCELLATION USING GENERAL PHASE ROTATED CONJUGATE TRANSMISSION . . . 815

Transmitter
80
Standard OFDM 1st path

Calculator
e j TDM

Mapping
Proposed PRCC theoretical = /4

IFFT
X

S/P

P/S
60 Proposed PRCC simulated = /4 k
Proposed PRCC theoretical = /8
j 2nd path
( )
*
Proposed PRCC simulated = /8 e
Proposed PRCC theoretical = /16
40
Proposed PRCC simulated = /16
Conjugate Cancellation Receiver
CIR (dB)

1st path
20

S/P
De-mux

Equalizer
Mapping

Channel
Zk (+ )

FFT
De-
P/S
0
( )
*

2nd path
20

Fig. 2. Block diagram of the proposed general PRCC scheme.

40
0 0.1 0.2 0.3 0.4 0.5
{0, e j
xn }
xn Cooperative
Encoding
Fading
Channel
Frequency
offset 1 {y (1)
n , yn(2) } {Yl
(1)
, Yl ( 2) }
Fig. 1. CIR comparison of the standard OFDM system, the conventional
CC scheme, and the proposed scheme for three different phase rotations. + FFT Combining

xn Cooperative Fading Frequency

Encoding Channel offset 2

{(e
j
xn)*,0 }
{w (1)
n , wn(2) }
may experience distinguishable frequency offsets. Fig. 3 shows
the multi-transmitter cooperative OFDM transmission model AWGN

proposed in [20], which is considered as the cooperative

protocol in this paper, with two relays participating in the
cooperation. At the first epoch, the first relay transmits
and the second one transmits 0. Successively, the first relay
transmits 0 and the second one transmits ( ) at the
second epoch. The receiver combines the received signals on
the two epochs. Following the derivations from (10)-(16), we
can obtain the CIR of the proposed scheme under this protocol
as
(1 ) + (2 )2
=
1
( 1 ) + ( + 2 )2
=1
where 1 and 2 (1 = 2 ) are normalized frequency offsets
suffered by two different relays to the destination. The cost
function of determining the optimal phase rotation now be-
comes
= arg max (1 , 2 , ), [/2, /2).
Proposition 2: The optimal phase rotation of the above
scenario is
(1 + 2 )( 1)
= . (21)
2
Proof : Similar to the proof of Proposition 1.
IV. S IMULATION R ESULTS
To verify the effectiveness of the general PRCC scheme,
a number of computer simulations were conducted. Although
the derivations in Sections II and III were all in the AWGN
channel, the BER performance was evaluated in the simu-
lations under the multipath Rayleigh fading channel model
described in [21], which is adopted by the IEEE 802.11
Working Group for software simulations. The channel is
modeled by complex samples with the average power decaying
exponentially and the phase and amplitude being uniformly
and Rayleigh distributed, respectively. Note that, in real wire-
less environments, two-path conjugate transmission can be
Fig. 3. Multi-transmitter OFDM protocol.
achieved in a number of ways [16], where the two paths may
have correlated (similar) or uncorrelated multipath channel
structures. To investigate this issue, we performed simulations
on both the PRCC and the CC scheme with = 0.1 for two
extreme cases, one with channel correlation equal to 0 and the
other with channel correlation equal to 1. We can see from
(19) the simulation results shown in Fig. 4 that the PRCC scheme
outperforms the CC scheme in a similar manner for both cases;
similar results can also be obtained for another case with
channel correlation between 0 and 1. Therefore, for simplicity,
we considered only the case of uncorrelated multipath channel
structures in the remaining simulations. We also assumed that
the channel effects of both two paths are perfectly known and
can be perfectly compensated by the receiver. In the frequency
(20) domain, each OFDM symbol is composed of 1024 subcarriers
(i.e., = 1024) with QPSK-modulated data symbols. In the
time domain, each OFDM data symbol is appended by 64
samples of cyclic prefix.
A. Performance of the Proposed General PRCC Scheme
Fig. 5 shows the CIR performance of the general PRCC
scheme for five cases using different phase rotations. The
first curve in this figure represents the CIR for the case with
the optimal phase rotation calculated by (18). While doing
simulations for this curve, we optimally changed the phase
rotation with respect to . For the other four curves, the phase
rotations were fixed as -0.157, -0.628, -0.942, and -1.255,
which are the optimal phase rotations for = 0.05, 0.2, 0.3,
and 0.4, respectively. From this figure, we can see that the
proposed scheme with a fixed phase rotation not only achieves
the maximum CIR value at the corresponding mentioned
above, but also provides satisfactory CIR performance over a
pretty wide region around the specific . This property implies

Authorized licensed use limited to: Jeppiaar Engineering College. Downloaded on May 01,2010 at 03:49:16 UTC from IEEE Xplore. Restrictions apply.
816 IEEE TRANSACTIONS ON COMMUNICATIONS, VOL. 58, NO. 3, MARCH 2010

high tolerance to an imprecise frequency offset estimate which 0

10
would lead to a non-optimal phase rotation. It is also worth CC Corr = 0
PRCC Corr =0
mentioning that the property also suggests that the general 1
10 CC Corr = 1
PRCC Corr =1
PRCC scheme is not only able to cope with fixed frequency
offsets (caused by the mismatch between oscillators), but also 2
10

resistant to time-varying frequency offsets (caused by the

3
Doppler shift). 10

BER
Since the optimal phase rotation in the proposed general 4
10
PRCC scheme is calculated under the assumption that the Corr = 1

transmitter knows the frequency offset value, it is reasonable to 5

10
conjecture that the performance of the proposed scheme would
be affected by the frequency offset estimation error. Here, 6
10
we want to investigate how the frequency offset estimation Corr = 0

error influences the performance of the proposed scheme. 7

10
0 5 10 15 20 25 30
Fig. 6 shows the BER performance of the proposed scheme Eb/No (dB)

with imprecise frequency offset information in the transmitter

when the signal-to-noise ratio (SNR) is equal to 20dB and
the number of channel taps is equal to 6. The horizontal
axis represents the absolute value of the frequency offset
estimation error . Specifically, the optimal phase rotation
used for simulations is calculated by (18) with a frequency
offset estimate = + or = randomly. BER performance between the CC and the proposed scheme
One can see that the performance degradation resulted from
imprecise frequency offset information is very slight from
= 0 to 0.1 for = 0.05 and 0.3. Actually, the accuracy
of frequency offset estimation described in [20] is about 103
to 104 for SNR = 20dB; therefore, it is convincible to say
that the proposed scheme is workable in the presence of a
frequency offset estimation error.
Next, the performance of the proposed scheme under the
user cooperation scenario described in Section III.C is dis-
cussed. In the corresponding simulations, the SNR is fixed at
20dB, the number of channel taps is 6, and the frequency
offset suffered by the first path is 1 = 0.05 or 0.3. The
optimal phase rotation used for simulations is obtained by (21)
under the perfect frequency offset estimation assumption. The
simulation results are shown in Fig. 7, where the horizontal
axis represents the difference of frequency offsets experienced
by users 1 and 2, i.e., 1 2 . We can see that the BER
performance is insensitive to 1 2 , which means that the
proposed scheme is suitable for ICI cancellation in this user
cooperation scenario.
B. Comparison with Other Schemes
So far, we have presented performance evaluation of the
proposed PRCC scheme. Now, we want to compare it with
other existing methods. For the purpose of comparison, the
standard OFDM scheme, the symmetric conjugate symbol
repetition (SCSR) scheme in [9], and the conventional CC
scheme in [16] were also simulated. Note that, for the sake
of maintaining the same spectral efficiency among these four
schemes, the modulation types used are BPSK for standard
OFDM and QPSK for the others. Fig. 8 shows the correspond-
ing BER comparison under the aforementioned multipath
fading channel for fixed frequency offsets = 0.05 and 0.3.
We can observe that when the frequency offset is small (e.g.,
= 0.05), the proposed scheme can offer the best performance
because of the two-path diversity and the highest CIR among
Fig. 4. BER comparison of the conventional CC and the PRCC scheme with
= 0.1 for channel correlation equal to 0 or 1.
the schemes compared. It can be seen that the difference of
is indistinct when = 0.05. This is due to the fact that the
CC scheme is a special case of the proposed general PRCC
scheme with = 0 which is the optimal phase rotation for
very small frequency offsets. Note that 0.05 can be regarded
as a relatively small frequency offset in QPSK and BPSK
modulation forms. Nevertheless, when the frequency offset is
large (e.g., = 0.3), the proposed general PRCC scheme
shows its superiority over the others. Note that the SCSR
and the CC scheme have higher error floors than the standard
OFDM scheme because that these two are more sensitive to
the frequency offset than standard OFDM at high frequency
offset situations.
Fig. 9 depicts a BER performance comparison of different
schemes with both fixed and time-varying frequency offsets.
For simulating this scenario, we define the maximum time-
varying frequency offset as , and the frequency offset is
uniformly distributed from to + . We can see
that although the additional time-varying frequency offset
deteriorates the BER performance when = 0.05, the
proposed scheme still offers the best one among the schemes
compared. The above results demonstrate that the proposed
general PRCC scheme works not only in fixed frequency
offset environments but also in time-varying frequency offset
scenarios.
V. C ONCLUSIONS
In this paper, we have proposed a new ICI self-cancellation
approach, called the general PRCC scheme, for OFDM sys-
tems. The proposed method can be regarded as the general
form of the conventional CC scheme, where appropriate artifi-
cial phase rotation is performed on the signal to be transmitted.
It can offer better performance than the conventional CC
scheme, especially at high frequency offset situations. We
have also derived a closed form of the optimal artificial
phase rotation for a given frequency offset with which the
CIR value is maximized. Moreover, we have extended the

Authorized licensed use limited to: Jeppiaar Engineering College. Downloaded on May 01,2010 at 03:49:16 UTC from IEEE Xplore. Restrictions apply.
WANG and HUANG: INTERCARRIER INTERFERENCE CANCELLATION USING GENERAL PHASE ROTATED CONJUGATE TRANSMISSION . . . 817

SNR=20dB
0
80 10
Proposed PRCC = opt
70 1 = 0.05
Proposed PRCC = 0.157
1
Proposed PRCC = 0.628 10 1 = 0.3
60
Proposed PRCC = 0.942
50 Proposed PRCC = 1.255
2
10
40
CIR (dB)

BER
3
30 10

20
4
10
10

0
5
10
10

20 10
6
0 0.1 0.2 0.3 0.4 0.5 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1
| |
1 2

Fig. 5. CIR comparison of the proposed scheme for the case using the Fig. 7. BER performance of the proposed scheme under a user cooperation
optimal phase rotation and those using other phase rotations. scenario.

SNR=20dB 0
0 10
10
= 0.05
= 0.3
1
10 1
10

2
10
2
10
BER
BER

3
10
3
10 OFDM = 0.05
SCSR = 0.05
4
10 CC = 0.05
PRCC = 0.05
4
10 OFDM = 0.3
5
10 SCSR = 0.3
CC = 0.3
PRCC = 0.3
6
10 0 5 10 15 20 25
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1
Eb/No (dB)
||

Fig. 6. BER performance of the proposed scheme for different frequency Fig. 8. BER comparison of different schemes with a fixed frequency offset.
offset errors.

value once the absolute value of the difference is greater than

proposed scheme to a user cooperation protocol. As in the non-
cooperative scenario, the optimal artificial phase rotation has
been derived and computer simulation results have been given
to show the effectiveness. As compared to the conventional CC
scheme, the proposed general PRCC scheme involves the same
low receivers complexity and slightly higher transmitters
complexity. Besides, according to the frequency offset infor-
mation feedback in the training mode, the proposed scheme
can adapt itself to the frequency offset through adjusting the
artificial phase rotation. The adaptability can be achieved by
feeding the frequency offset estimate obtained at the receiver
back to the transmitter for recalculating the optimal phase
rotation. Since the CIR of the proposed scheme has a plateau
around the frequency offset we target (as shown in Fig. 5) and
the BER performance of the proposed scheme is not sensitive
to the frequency offset estimation error (as shown in Fig. 6),
there would be no need to do the feedback operations so often.
A reasonable feedback mechanism is suggested as follows:
1) check the difference between the current frequency offset
estimate and the last feedback value; 2) update the feedback
a threshold. The threshold could be a value smaller than the
maximum time-varying frequency offset . In general, the
smaller the threshold, the better the performance would be, but
the more often the feedback operations will be performed. We
have verified the suggested feedback mechanism for a time-
varying case with = 0.15 by using computer simulations,
where the system parameters used are the same as those for
Fig. 9. It can be seen from the simulation results shown in
Fig. 10 that the value 0.1 is a pretty good choice for the
threshold in this example. The adaptability feature makes the
proposed general PRCC scheme useful for applications with
time-varying frequency offsets.
A PPENDIX
Proof of Proposition 1
At first, we rewrite (16) as
(, )
(, ) =
1
(22)
(, )
=1

Authorized licensed use limited to: Jeppiaar Engineering College. Downloaded on May 01,2010 at 03:49:16 UTC from IEEE Xplore. Restrictions apply.
818 IEEE TRANSACTIONS ON COMMUNICATIONS, VOL. 58, NO. 3, MARCH 2010

10
0 After substituting (24) and (26) into (22), we can take partial
differentiation on (22) with respect to as follows:
(, )
=
10
1

1
(2 + 2( 1)/ ) 2

1
{4() ( ( )2
2
10
( )2 =1
BER

=1
OFDM = 0.05
SCSR = 0.05 + ( + )2 2( )( + ))}.
10
3
CC = 0.05 (27)
PRCC = 0.05
OFDM = 0.3
SCSR = 0.3
Since ( )2 + ( + )2 2( + )( ),
10
4 CC = 0.3 = 1, 2, . . . , 1, and the equality occurs when (
PRCC = 0.3
) = ( + ), we observe that (,) = 0 if
0 5 10 15 20 25
Eb/No (dB) (2 + 2 1 ) = 0, i.e., 2 + 2 1 = . It suffices
to consider 2 + 2 1 = 0; therefore, we conclude that
Fig. 9. BER comparison of different schemes with both fixed and time- the maximum CIR occurs when = 1 .
varying frequency offsets.
R EFERENCES
= 0.05 and = 0.15
10
0
D [1] R. van Nee and R. Prasad, OFDM for Wireless Multimedia Communi-
No Feedback (fixed corresponing to = 0.05 ) cations. Norwood, MA: Artech House, 2000.
No Feedback (fixed corresponding to = 0.1) [2] ETSI, Digital video broadcasting (DVB); framing structure, channel
Feedback with Threshold = 0.1
Feedback with Threshold = 0.01
coding and modulation for digital terrestrial television," ETSI EN 300
1
10
744 V1.4.1, Jan. 2001.
[3] IEEE, IEEE standard for local and metropolitan area networkspart 11:
wireless LAN medium access control (MAC) and physical layer (PHY)
10
2 specifications," IEEE Std. 802.11, Aug. 1999.
[4] IEEE, IEEE standard for local and metropolitan area networkspart 16:
BER

air interface for fixed broadband wireless access systems," IEEE Std.
3
802.16-2004, Oct. 2004.
10
[5] M. Speth, S. A. Fechtel, G. Fock, and H. Meyr, Optimum receiver
design for wireless broad-band systems using OFDMpart I," IEEE
Trans. Commun., vol. 47, no. 11, pp. 1668-1677, Nov. 1999.
10
4
[6] T. Pollet, M. V. Bladel, and M. Moeneclaey, BER sensitivity of OFDM
systems to carrier frequency offset and Wiener phase noise," IEEE
Trans. Commun., vol. 43, pp. 191-193, Nov. 1995.
[7] Y. Zhao and S.-G. Hggman, Intercarrier interference self-cancellation
0 5 10 15 20 25 scheme for OFDM mobile communication systems," IEEE Trans. Com-
Eb/No (dB)
mun.., vol. 49, no. 7, pp. 1185-1191, July 2001.
[8] J. Armstrong, Analysis of new and existing methods of reducing
intercarrier interference due to carrier frequency offset in OFDM," IEEE
Fig. 10. BER performance of the proposed scheme under a time-varying Trans. Commun., vol. 47, no. 3, pp. 365-369, Mar. 1999.
scenario with = 0.15 for different threshold values. [9] K. Sathananthan, C. R. N. Athaudage, and B. Qiu, A novel ICI
cancellation scheme to reduce both frequency offset and IQ imbalance
effects in OFDM," in Proc. IEEE 9th Int. Symp. Comput. Commun.,
where July 2004, pp. 708-713.
[10] Y. Zhao, J. D. Leclercq, and S.-G. Hggman, Intercarrier interference
(, ) = () + ()2 compression in OFDM communication systems by using correlative
coding," IEEE Commun. Lett., vol. 2, pp. 214-216, Aug. 1998.
= ()2 + ()2 + (2 ()() [11] C.-L. Wang, Y.-C. Huang, and P.-C. Shen, An intercarrier interference
suppression technique using time-domain windowing for OFDM sys-
+ 2 () ()) (23) tems," in Proc. 2006 IEEE Veh. Technol. Conf.Spring (VTC 2006
Spring), Melbourne, Australia, May 2006, pp. 2518-2522.
and [12] H. Zhang and Y. (G.) Li, Optimum frequency-domain partial response
encoding in OFDM system," IEEE Trans. Commun., vol. 51, no. 7, pp.
(, ) = ( ) + ( + )2 1064-1068, July 2003.
= ( )2 + ( + )2 [13] W. Bai and Y. Kim, Effects of cooperative coding on OFDM systems
for intercarrier interference suppression," IEEE Trans. Veh. Technol.,
+ 2( )( + )(2 + 2( 1)/ ). vol. 56, no. 6, pp. 3732-3738, Nov. 2007.
(24) [14] C.-D. Chung, Correlatively coded OFDM," IEEE Trans. Wireless
Commun., vol. 5, no. 8, pp. 2044-2049, Aug. 2006.
Since [15] H.-G. Yeh and C. C. Wang, New parallel algorithm for mitigating the
frequency offset of OFDM systems," in Proc. 2004 IEEE Veh. Technol.
() 1 Conf.Fall (VTC 2004Fall), Los Angeles, CA, Sept. 2004, pp. 2087-
() = () 2091.
{()/ }
2(1 )
[16] H.-G. Yeh, Y.-K. Chang, and B. Hassibi, A scheme for cancelling
= () (25) intercarrier interference using conjugate transmission in multicarrier
communication systems," IEEE Trans. Wireless Commun., vol. 6, no.
and ()2 = ()2 , (23) can be rewritten as 1, pp. 3-7, Jan. 2007.
[17] A. Sendonaris, E. Erkip, and B. Aazhang, User cooperation diversity
1 part I: system description," IEEE Trans. Commun., vol. 51, no. 11, pp.
(, ) = 2()2 + 2()(2 + 2 ). (26) 1927-1938, Nov. 2003.

Authorized licensed use limited to: Jeppiaar Engineering College. Downloaded on May 01,2010 at 03:49:16 UTC from IEEE Xplore. Restrictions apply.
WANG and HUANG: INTERCARRIER INTERFERENCE CANCELLATION USING GENERAL PHASE ROTATED CONJUGATE TRANSMISSION . . .
[18] A. Sendonaris, E. Erkip, and B. Aazhang, User cooperation diversity
part II: implementation aspects and performance analysis," IEEE Trans.
Commun., vol. 51, no. 11, pp. 1939-1948, Nov. 2003.
[19] J. N. Laneman, D. N. C. Tse, and G. W. Wornell, Cooperative diversity
in wireless networks: efficient protocols and outage behavior," IEEE
Trans. Inf. Theory, vol. 50, no. 12, pp. 3062-3080, Dec. 2004.
[20] X. Li, F. Ng, and T. Han, Carrier frequency offset mitigation in
asynchronous cooperative OFDM transmission," IEEE Trans. Signal
Process., vol. 56, no. 2, pp. 675-685, Feb. 2008.
[21] B. OHara and A. Petrick, The IEEE 802.11 Handbook: A Designers
Companion. IEEE Press, 1999.
Chin-Liang Wang (S85M87SM04) was born Outstanding Award of the 1995 Student Paper Contest sponsored by the
in Tainan, Taiwan, R.O.C., in 1959. He received the
B.S. degree in electronics engineering from National
Chiao Tung University (NCTU), Hsinchu, Taiwan,
in 1982, the M.S. degree in electrical engineering
from National Taiwan University, Taipei, Taiwan, in
1984, and the Ph.D. degree in electronics engineer-
ing from NCTU in 1987.
He joined the faculty of National Tsing Hua Uni-
versity (NTHU), Hsinchu, Taiwan, in 1987, where
he is currently a Professor of the Department of
Electrical Engineering and the Institute of Communications Engineering.
During the academic year 1996-1997, he was on leave at the Information
Systems Laboratory, Department of Electrical Engineering, Stanford Uni-
versity, Stanford, CA, as a Visiting Scholar. He served as the Director of
the Institute of Communications Engineering from 1999 to 2002 and the
Director of the Universitys Computer & Communications Center from 2002
to 2006. He was the Chair of the Wireless Networks Group of the National
Science & Technology Program for Telecommunications from 2004 to 2008,
and has been the Chair of the Access Technology Group of the Networked
Communications Program since 2009. He is also serving as the Director of the
Communications Engineering Program, National Science Council, R.O.C. His
current research interests are primarily in baseband technologies for wireless
communications and wireless sensor networks.
Authorized licensed use limited to: Jeppiaar Engineering College. Downloaded on May 01,2010 at 03:49:16 UTC from IEEE Xplore. Restrictions apply.
819
Dr. Wang was a recipient of the Distinguished Teaching Award sponsored
by the Ministry of Education, R.O.C., in 1992. He received the Acer Dragon
Thesis Award in 1987 and the Acer Dragon Thesis Advisor Awards in 1995
and 1996. In the academic years 1993 1994 and 1994 1995, he received the
Outstanding Research Awards from the National Science Council, R.O.C.
He received the HDTV Academic Achievement Award from the Digital
Video Industry Development Program Office, Ministry of Economic Affairs,
R.O.C., in 1996. He was also the advisor on several technical works that
won various awards in Taiwan, including the Outstanding Award of the 1993
Texas Instruments DSP Product Design Challenge in Taiwan, the Outstanding
Award of the 1994 Contest on Design and Implementation of Microprocessor
Application Systems sponsored by the Ministry of Education, R.O.C., the
Chinese Institute of Engineers, and the 1995 and 2000 Master Thesis Awards
of the Chinese Institute of Electrical Engineering. He served as an Associate
Editor for the IEEE T RANSACTIONS ON S IGNAL P ROCESSING from 1998 to
2000 and has been an Editor for Equalization for the IEEE T RANSACTIONS
ON C OMMUNICATIONS since 1998. He is also one of the Guest Editors for
the Special Issue on Model Order Selection in Signal Processing Systems of
the IEEE J OURNAL OF S ELECTED T OPICS IN S IGNAL P ROCESSING.
Yu-Chih Huang was born in Taipei, Taiwan,
R.O.C., in 1981. He received the B.S. degree in elec-
trical engineering from Chung Yuan Christian Uni-
versiy, Chungli, Taiwan, in 2003 and the M.S. degree
in communications engineering from National Tsing
Hua University, Hsinchu, Taiwan, in 2005. From
2005 to 2007 he was in his military service. After
that, he spent one year being a full-time research
assistant under the Computer & Communications
Center, National Tsing Hua University. Currently,
he is pursuing his Ph.D. degree in the Department
of Electrical and Computer Engineering, Texas A&M University, College
Station, Texas. His research interests include wireless communications and
information theory.