See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/245462758

Complement block coding scheme for reducing peak-to-average power ratio

of OFDM systems

Article  in  Journal of Electronics (China) · September 2004

DOI: 10.1007/BF02687943

CITATIONS READS

6 189

2 authors:

Tao Jiang Guangxi Zhu

Huazhong University of Science and Technology Huazhong University of Science and Technology
378 PUBLICATIONS   6,038 CITATIONS    275 PUBLICATIONS   1,739 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Security with Physical layer Infomation View project

Massive MIMO View project

All content following this page was uploaded by Tao Jiang on 01 February 2015.

The user has requested enhancement of the downloaded file.

JIANG LAYOUT 8/22/05 11:15 AM Page 51

Complement Block Coding for

Reduction in Peak-to-Average
Power Ratio of OFDM Signals
TAO JIANG AND GUANGXI ZHU, HUAZHONG UNIVERSITY OF SCIENCE AND
TECHNOLOGY

Abstract large linear range. Otherwise, nonlinear signal distortion

occurs, and leads to high adjacent channel interference and
In this article a new coding scheme, complement block poor system performance. In order to reduce the PAPR of
coding (CBC), is proposed to reduce the PAPR of OFDM sig- OFDM signals, many schemes have been proposed, such as
nals. This method utilizes the complement bits that are added [1, 2], among which coding schemes [3–7] are the most attrac-
to the original information bits, which can effectively reduce tive due to their inherent error control capability and simplici-
the PAPR of OFDM signals with random frame size N and ty of implementation. In 1994 a simple block coding scheme
the coding rate R ≤ (N – k)/N, where k is a positive integer was introduced by Jones et al. [3]. The basic idea of this
and k ≤ N/2. The performance results obtained with CBC are scheme is that mapping 3 bits of data into 4 bits of codeword
given and compared with that of some well known schemes, by adding a simple odd parity code (SOPC) at the last bit
such as simple block coding, modified simple block coding, across the channels reduces the PAPR, and can reduce PAPR
simple odd parity code, and cyclic coding, for the same pur- by 3.54 dB. The main disadvantage of this method is that it
pose. The results show that at the same coding rate 3/4, CBC can only reduce PAPR for a 4-bit codeword. This is because if
can achieve almost the same performance as SBC and MSBC, SOPC is used for longer codes, the high-powered codewords
but with lower complexity, and the same performance can be will not be eliminated. Later, Fragiacomo et al. proposed a
obtained with a higher coding rate using CBC. The PAPR modified and efficient version of the simple block coding
reductions of CBC with coding rate (N – 1)/N are almost the (SBC) scheme to reduce the PAPR of OFDM signals [4], but
same as with a coding rate less than (N – 1)/N, but almost the this method is not effective when the frame size is large [8].
twice as these of SOPC when N ≥ 16. Moreover, we can find Subsequently, a modified SBC (MSBC) scheme was presented
that PAPR is the lowest for all block codes using CBC with [9] and combined with the subblock processing technique to
coding rate 3/4. So modified CBC (MCBC) is also proposed make SBC effective for OFDM systems with large frame sizes
and analyzed, combined with the subblock processing tech- [10]. But in practice, compared to SBC, the performance of
nique to make CBC effective for OFDM systems with large the MSBC scheme is no improvement when the coding rate R
frame sizes. The flexibility in coding rate choice and low com- > 5/6 [9]. Furthermore, the performance of OFDM systems
plexity make the proposed CBC more suitable for random can only be improved by MSBC under the lower coding rate
frame size with high coding rate and can also provide error when the frame size is large. In this article a complement
detection. block coding (CBC) scheme is proposed and analyzed; modi-
fied CBC (MCBC), which is combined with the subblock pro-
INTRODUCTION cessing technique, is also presented. CBC and MCBC can
effectively reduce the PAPR of OFDM systems with random
Due to its several significant advantages: frame sizes according to the required coding rate. The simula-
• Robust to multipath fading, intersymbol interference, co- tion results demonstrate that the CBC and MCBC can pro-
channel interference, and impulsive parasitic noise vide more or almost the same PAPR reduction as SBC,
• Lower implementation complexity than the single-carrier MSBC, SOPC, and CC, but with higher coding rate and low
solution complexity. This means that the bandwidth efficiency of
• High spectral efficiency in supporting broadband wireless OFDM will be improved, which is very important to wireless
communications communication. For comparing with some well-known results
orthogonal frequency-division multiplexing (OFDM) is conveniently, we select binary phase shift keying (BPSK)
believed to be a suitable technique for broadband wireless OFDM systems.
communications and has been adopted in many wireless This article is organized as follows. We briefly review
standards, such as digital audio broadcasting (DAB), terres- OFDM systems, and the PAPR problem is formulated. Then
trial digital video broadcasting (DVB-T), the European particular descriptions of CBC and MCBC schemes are given.
Telecommunications Standards Institute (ETSI) HIPER- To confirm the effectiveness of the proposed methods, com-
LAN/2 standard, the IEEE 802.11a standard for wireless puter simulations that have been performed and their results
local area networks (WLANs), and the IEEE 802.16a stan- are shown, followed by conclusions.
dard for wireless metropolitan area networks (WMANs).
The main disadvantage of OFDM is its large peak-to-aver- OFDM and PAPR
age power ratio (PAPR). When N signals are added with the
same phase, they produce a peak power, which is N times the In OFDM systems, total bandwidth is divided into many over-
average power. Of course, not all code words result in a bad lapping orthogonal subchannels, which are transmitted in par-
PAPR. High PAPR requires sophisticated (expensive) radio allel synchronously. The complex envelope of the transmitted
transmitters with high-power amplifiers operating in a very OFDM signals can be written as

IEEE Radio Communications • September 2005 0163-6804/05/$20.00 © 2005 IEEE S17

JIANG LAYOUT 8/22/05 11:15 AM Page 52

6 6

5 5

4 4

3 3

2 2

1 1

0 0
0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1
(a) (b)

6 6

5 5

4 4

3 3

2 2

1 1

0 0
0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1
(c) (d)

FIGURE 1. Reduction of the PAPR of OFDM using CBC (BPSK, N = 6): a) [111111]; b) [111101]; c) [111110]; d) [111001].

coding can be directly implemented before modulation. If the

∞ N −1
power in each subcarrier is normalized to 1 W, Pavg is N W.
s(t ) = ∑ ∑ dm,n e j 2 π fnt g(t − nT ), (1)
So the PAPR of uncoded BPSK OFDM systems with a frame
m =−∞ n = 0
——
size of N is equal to 10 log10(N) (dB).
where j=√–1, T is the time duration of one OFDM frame, N
is the number of subcarriers (i.e., one OFDM frame size), Complementary Block Coding
dm,n is the data symbol modulated by the nth subcarrier of the
mth OFDM frame, fn is the frequency of the nth subcarrier, In order to illustrate the principle of the CBC technique, we
and g(t) is a rectangular pulse shape of duration T. give some intuitive results. One example is given below, where
Let us consider the 0th block interval. Then Eq. 1 can be the frame size is chosen as N = 6 (i.e., the number of subcar-
simply expressed as riers is 6). The envelope of the OFDM signal for the uncoded
sequence with all bits equal to 1 (length equals 6) is shown in
N −1
Fig. 1a, where the abscissa is t/T (0 ≤ t ≤ T). Choosing the
s(t ) = ∑ d n e j 2 π fn t . (2)
length of information sequence as 5 instead of 6 (coding rate
n=0
equals 5/6), and encoding the all ones sequence (length is 5)
Let p(t) be the envelope power of the OFDM signal. Since with CBC, one gets the frame codeword 111101. The enve-
s(t) is a complex valued, p(t) = s(t) ⋅ s*(t), where * denotes a lope of this codeword is shown in Fig. 1b. Compared to the
conjugate of a complex number. uncoded all ones sequence of length 6 (Fig. 1a), the PAPR
The PAPR (in dB) of the OFDM signal can be defined as reduction with CBC is 3.45 dB (Fig. 1b). Figure 1c is the
 P  envelope of the frame codeword 111110, and the PAPR
PAPR = 10 log10  peak  , reduction with CBC is 3.41 dB. The PAPR can be reduced
(3)
 Pavg  further by appending two complement bits. Choosing the
length of the information sequence as four with two comple-
where Pavg is the average power consumed by each frame, and ment bits for one frame (coding rate 2/3), and encoding with
Ppeak is the maximum of power for one OFDM frame. When CBC, one gets 111001 as the frame codeword. The envelope
BPSK modulation is used on each subcarrier, binary block of this codeword is also shown in Fig. 1d.

S18 IEEE Radio Communications • September 2005

JIANG LAYOUT 8/22/05 11:15 AM Page 53

Complement
(a) (b)

Original bits 101010101010 100110110101

d1 ... dk-n ... dk-1 ... c1 ... cn ... dN-n

POS.1 1010↓↓↓↓↑↑↑↑1010 1001↓↓↓↓↑↑↑↑0101
d: data bit c: complement bit

(a) POS.2 101↓↓↓↓0↑↑↑↑1010 100↓↓↓↓1↑↑↑↑0101

... ... ... ... ↓: Positions of IBs for constructed CBS ↑: Positions of CBs

TABLE 1. Structures of the frame with different positions of CBs.

Transmitter
IFFT

Parallel-to-serial conversion
(a) (b)
Channel

Serial-to-parallel conversion Original bits 101010101010 101110111011

Receiver
FFT 3 CBs 10101010↑↑↑1010 10111011↑↑↑1011

-1 -1 4 CBs 10101010↑↑↑↑1010 10111011↑↑↑↑1011

5 CBs 10101010↑↑↑↑↑1010 10111011↑↑↑↑↑1011

TABLE 2. Structures of the frame with different numbers of CBs.

... ... ...

IBs IBs used for CBs CBs codes are composed of information bits (IBs) and complement
bits (CBs). Code length N is the number of subcarriers, and dk
(b) is a certain information bit. A certain complement bit dk+i (i ≤
n, and (i+k) ≤ N), which is applied to the (i+k)th subcarrier,
FIGURE 2. CBC technique: a) code sequence with CBC; b) using CBC in is the inverse of the information bit dk–i. If we use n CBs, they
an OFDM system. are the complementary of the n IBs, so the number of infor-
mation bits in a block code sequence is (N – n).
Then we expatiate a general illustration of the CBC tech- Although the CBC technique increases bit redundancy and
nique in OFDM systems. All illustrative frame structure is the coding rate is lower than that of an uncoded sequence, the
given in Fig. 2a, and Fig. 2b shows how to use the CBC tech- maximum peak power of the OFDM signal can be reduced by
nique in OFDM systems. This method utilizes the comple- these CBs. This is because the code sequence with all bits
mentary sequence, which is added to the midst of the original equal and all bits alternating is not generated in a sequence
sequence (called the information sequence), so the whole from these CBs. The complemental bit dk+i is a recomplemen-

16 9
POS.1 POS.1
POS.2 POS.2
14 8

7
12

6
10

5
8
4
6
3

4
2

2
1

0 0
0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1

(a) (b)

FIGURE 3. Results of the different positions of CBs with n = 4.

IEEE Radio Communications • September 2005 S19

JIANG LAYOUT 8/22/05 11:15 AM Page 54

14 8
3 CBs 3 CBs
4 CBs 4 CBs
5 CBs 5 CBs
12 7

6
10

5
8

4
6
3

4
2

2
1

0 0
0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1
(a) (a)

FIGURE 4. Results of different numbers of CBs but beginning at the same position.

tation bit at the receiver side and is compared with the corre- information on CBs is shown in Table 2. Ordinarily, the num-
sponding information bit dk–i to detect the stronger one. So ber of CBs can be chosen according to the coding rate.
the CBs can also provide error detection, namely CBC tech- Moreover, it has been found that when the frame size is
nique has correction capability. large (the information sequence is long), high PAPR reduc-
There are two key questions in CBC. Where is the opti- tion can still be obtained when the long information sequence
mum selection for a series of IBs to construct the CBs is the is divided into several subblocks, and each subblock is encod-
first key question. One example is shown below, where the ed with CBC. This is the basic idea of MCBC, which con-
frame size is chosen as N = 16 and the number of CBs is cho- structs each frame sequence with CBC coded subblocks to
sen as n = 4. The results of different selections for IBs are reduce the PAPR. At the end, a frame codeword is formed
shown in Fig. 3, and detailed information on IBs and CBs is from the coded subblocks. The MCBC scheme is quite effec-
given in Table 1. We can easily find that the results of POS.2 tive for multicarrier systems with large frame size.
are better than that of POS.1. Generally, POS.2 is selected for
avoiding symmetry, as in Fig. 3a, and we should choose k = Simulations and Results
N/2 in the original bits when N is an even number, whereas if n
is an odd number, k = (N + 1)/2 in Fig. 2. The second key In this section a BPSK OFDM system is assumed to be used
question is how to confirm the number of CBs. Figure 4 shows for CBC and MCBC to compare with some well-known results
two examples to see the PAPR reduction for CBC when the of SBC, MSBC, SOPC, and CC. It is an example of the PAPR
number of CBs is chosen as n = 3, 4, 5, respectively, and the reduction improvement for CBC in OFDM systems with N =
original frame size is chosen as 12: (N – n) = 12. Detailed 16 and coding rate R = 3/4, as shown in Fig. 5, where the

30
12 SBC with coding rate R = 7/8
SBC CBC with coding rate R = 7/8
MSBC MCBC with coding rate R = 7/8
CBC 25 CBC with coding rate R = 31/32
10 MCBC CBC with coding rate R = 15/16

20
8

15
6

10
4

5
2

0
0 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
FIGURE 6. Performance improvement by using CBC with different coding
FIGURE 5. Performance improvement by using CBC with the original rate, with original frame 0000000000000000000000000000 loaded and
frame 110110011011 loaded, but for MCBC with two subblocks. two subblocks for MCBC.

S20 IEEE Radio Communications • September 2005

JIANG LAYOUT 8/22/05 11:15 AM Page 55

original frame load is 110110011011. The PAPR reduc-

tion êPAPR = 6.45 (dB) when CBC is used and a PAPR
PAPR (dB)
reduction of êPAPR = 6.97 (dB) can be obtained by
N n R
MCBC with two subblocks, but êPAPR = 5.35 (dB) for
CBC SBC MSBC SOPC CC

MSBC [9], êPAPR = 2.97 (dB) for SBC [9, 10]. 4 1 3/4 2.48 2.48 2.48 2.48 2.48
We can define coding rate R = (N – n)/N, where N is
the total size of the whole codeword (composed of the IBs 1 7/8 6.46 6.53 6.53
6.53 5.36
and CBs), and n is the number of complemental bits. Figure 8
(R = 7/8) (R = 3/4)
6 shows the results when N = 32 and n = 4 (i.e., R = 7/8). 2 3/4 6.38 5.33 6.38
The results of PAPR in an OFDM system obtained with
SBC and MSBC are the same, as shown in Fig. 6. In order 1 15/16 9.28 10.90 10.90
to make comparisons, some results of PAPR and PAPR
2 7/8 9.28 9.54 9.57
reduction obtained with SOPC [3, 10], SBC [4, 9], MSBC 16
10.88 8.30
[9], and CC [5] are presented. When R ≥ 3/4, some results 3 13/16 9.28 — —
(R = 15/16) (R = 3/4)
of PAPR and PAPR reduction obtained with CBC, SBC,
MSBC, SOPC, and CC coding schemes are given in Tables 4 3/4 9.28 9.08 8.94
3 and 4, respectively, where N is the whole codeword length,
R denotes the coding rate, and n is the number of CBs in 1 31/32 13.91 14.51 14.51
each frame. From these results, it can be observed that
under the same coding rate 3/4, no matter what the frame 2 15/16 13.91 13.91 13.91
size, the performance of the proposed CBC scheme is
almost the same as that of SBC and MSBC. But the same 3 29/32 13.27 — —
14.49
performance as that of SBC, MSBC, and SOPC can be 32 —
(R = 31/32)
obtained with a higher coding rate by using CBC. When R 4 7/8 12.57 12.56 12.56
= (n – 1)/N, the performance of CBC is almost the same as
5 27/32 12.22 — —
that of SOPC. Furthermore, the PAPR reductions of CBC
when R = (N – 1)/N are almost the same as that when R < 8 3/4 12.22 12.07 11.38
(N – 1)/N. The PAPR maximums of different frame sizes
obtained by MCBC with the same coding rate R = 3/4, and TABLE 3. PAPR comparisons for CBC, SBC, MSBC, SOPC, and CC.
each OFDM frame divided to different subblocks are shown
in Table 5. From these results, it can be observed that under
the same coding rate 3/4, the performance of MCBC is PAPR (db)
almost the same as that of CC. But under the same coding N n R
rate, such as 3/4, the PAPR reduction obtained by MCBC is CBC SBC MSBC SOPC CC
better than the rest (SBC, MSBC, etc.).
So the proposed CBC and MCBC are more suitable 4 1 3/4 3.56 3.56 3.56 3.56 3.56
for OFDM systems with large frame sizes. However,
CBC and MCBC have lower complexity than SBC, 1 7/8 2.59 2.52 2.52
MSBC, SOPC, and CC, especially for decoding. In addi- 2.52 3.66
8
tion, CBC and MCBC also have high coding ratio and (R = 7/8) (R = 3/4)
2 3/4 2.67 3.72 2.67
correction capability.
1 15/16 2.74 1.16 1.16
Conclusions
2 7/8 2.74 2.52 2.49
In this article a new block coding scheme with low com- 1.18 3.742
16
plexity, complement block coding, is proposed to reduce (R = 15/16) (R = 3/4)
3 13/16 2.74 — —
the PAPR of OFDM signals and detect transmission
errors. It was found that almost 3 dB PAPR reduction 4 3/4 2.74 2.98 3.12
can be obtained when coding rate R > (N – 2)/N by using
CBC with long frame size. It is also shown that the 1 31/32 1.16 0.55 0.55
PAPR reductions obtained with CBC when coding rate R
= (N – 1)/N are almost the same as that when R < (N – 2 15/16 1.16 1.16 1.16
1)/N. In other words, when the frame size is large, besides
having lower complexity, the proposed CBC scheme can 3 29/32 2.75 — —
provide a higher coding rate than SBC, MSBC , SOPC, 0.58
32 —
and CC with the same performance (i.e., PAPR reduc- (R = 31/32)
4 7/8 2.50 2.51 2.51
tion). In addition, when coding rate is 3/4, more than 3
dB more PAPR reduction can be obtained using MCBC 5 27/32 2.75 — —
than the other schemes with any frame size. The flexibili-
ty in coding rate choice and low complexity make the 8 3/4 2.75 3.00 3.69
proposed CBC and MCBC schemes attractive for OFDM
systems with large frame sizes and high coding rates. TABLE 4. PAPR reduction comparisons for CBC, SBC, MSBC, SOPC, and CC.

ACKNOWLEDGMENT
This work was supported in part by the National High REFERENCES
Technology Research and Development Program of China
under grant no. 2001AA123014. Thanks are given to Dr. [1] R. van Nee, “OFDM Codes for Peak-to-Average Power Reduction and Error
Correction,” Proc. IEEE GLOBECOM ’96, Nov. 1996, pp. 740–44.
Joseph Mitola III and anonymous reviewers for their valuable [2] T. Jiang and G. Zhu, “Nonlinear Companding Transform for Reducing
comments and suggestions, which helped to improve the pre- Peak-to-Average Power Ratio of OFDM Signals,” IEEE Trans. Broadcast-
sentation of the article. ing, vol. 50, Sept. 2004, pp. 342–46.

IEEE Radio Communications • September 2005 S21

JIANG LAYOUT 8/22/05 11:15 AM Page 56

[6] C. Tellambura, “Use of m-sequences for OFDM Peak-to-average Power

PAPR (dB) Ratio Reduction,” IEE Elect. Lett., vol. 33, July 17, 1997, pp. 1300–01.
[7] X. Li and J. Ritcey, “M Sequences for OFDM Peak-to-average Power
N Ration Reduction and Error Correction,” IEE Elect. Lett., vol. 33, Mar.
CBC Number of subblocks MCBC 27, 1997, pp. 554–55.
[8] C. Tellambura, “Comment: Multicarrier Transmission Peak-to-average
Power Reduction Using Simple Block Code,” IEE Elect. Lett., vol. 34,
8 6.38 2 6.24 Aug. 20, 1998, pp. 1646–47.
[9] Y. Zhang et al., “OFDM Peak Power Reduction by Sub-Block-Coding and
Its Extended Versions,” Proc. IEEE, VTC, Houston, USA, May. 1999, pp.
2 8.56
695–99.
16 9.28 [10] Y. Zhang, A. Yongacoglu, and J. Y. Chouinard, “Reducing Multicarrier
4 8.49 Transmission Peak Power with a Modified Simple Block Code,” Proc.
ICCT, Beijing, China, Aug. 2000, pp. 578–80.
2 10.26
24 11.03 BIOGRAPHIES
3 7.80 T AO J IANG (unique_jt@yahoo.co.uk) received a Ph.D. degree in informa-
tion and communication engineering from Huazhong University of Sci-
ence and Technology, Wuhan, P. R. China, in 2004. From August 2004
2 11.52 to August 2005, he was with Electronic and Computer Engineering,
32 12.22 School of Engineering and Design, Brunel University, Uxbridge, United
4 11.31 Kingdom. His current research interests include the areas of wireless and
mobile communications, coding theory, modulation, and signal process-
ing for communications, especially for OFDM, UWB, and cross-layer tech-
TABLE 5. PAPR comparisons for CBC and MCBC with niques. He is an expert reviewer in the research areas of DSP, broadband
the same coding rate 3/4 but different frame sizes. communications, and multicarrier modulation for IEEE Transactions on
Broadcasting, the Journal of Communications and Networks (Korean),
and others.
[3] A. E. Jones, T. A. Wilkinson, and S. K. Barton, “Block Coding Scheme for
Reduction of Peak-to-Average Envelope Power Ratio of Multicarrier G UANGXI Z HU (gxzhu@mail.hust.edu.cn) received a B.S. in information
Transmission Systems,” IEE Elect. Lett., vol. 30, Dec. 8 1994, pp. and communication engineering from Huazhong Institute of Technolo-
2098–99. gy, P.R. China, in 1969. Since 1974 he has held a variety of academic
[4] S. Fragicomo, C. Matrakidis, and J. J. O’Reilly, “Multicarrier Transmission positions at Huazhong University of Science and Technology, P.R. China.
Peak-to-Average Power Reduction Using Simple Block Code,” IEE Elect. His research interests include multimedia communication, signal pro-
Lett., vol. 34, May 14, 1998, pp. 953–54. cessing, and engineering education, and he has published over 100
[5] D. Wulich, “Reduction of Peak to Mean Ratio of Multicarrier Modulation research papers. He is currently a professor at Huazhong University of
Using Cyclic Coding,” IEE Elect. Lett., vol. 32, Feb. 29, 1996, pp. 432–33. Science and Technology.

CALL FOR PAPERS

IEEE
R ADIO
C OMMUNICATIONS
COMPONENTS, SYSTEMS, and NETWORKS
A QUARTERLY SUPPLEMENT TO IEEE COMMUNICATIONS MAGAZINE
SUPPLEMENT EDITOR: DR. JOSEPH MITOLA III, THE MITRE CORPORATION
IEEE Radio Communications will cover components, systems and networks related to radio frequency (RF) technolo-
gy. Articles will be in-depth, cutting-edge tutorials, emphasizing state of the art design solutions involving physical
and other lower-layer issues in radio communications, including RF, microwave and radio-related wireless communica-
tions topics.
IEEE Radio Communications will emphasize practical solutions and emerging research for immediate and practical
applications for innovative research, design engineers, and engineering managers in industry, government, and aca-
demic pursuits. Articles will be written in clear, concise language and at a level accessible to those engaged in the
design, development, and application of products, systems, and networks.
A rigorous peer review process will ensure that only the highest quality technical articles are published, keeping to
the high IEEE standard of technical objectivity that precludes marketing and product endorsements.
IEEE Radio Communications will bind into IEEE Communications Magazine on a quarterly schedule, on special paper
stock, ensuring in-depth readership by our paid and unduplicated circulation of more than 50,000 subscribers.

Manuscripts must be submitted through the magazine’s submissions Web site at PUBLICATION DATES
http://commag-ieee.manuscriptcentral.com/ MARCH, JUNE,
On the Manuscript Details page please click on the drop-down menu to select SEPTEMBER,
Radio Communications Supplement DECEMBER

S22 IEEE Radio Communications • September 2005

View publication stats