DocumentCode
3253832
Title
Improved Equalization For Coded, Zero-Padded OFDM (ZP-OFDM) Systems
Author
Ghosh, M.
Author_Institution
Philips Res. USA, Briarcliff Manor
fYear
2007
fDate
24-28 June 2007
Firstpage
4263
Lastpage
4268
Abstract
Zero-padded OFDM (ZP-OFDM) systems have recently been adopted in certain wireless systems, such as the multiband-OFDM based WiMedia standard, instead of the more commonly used cyclic-prefix OFDM (CP-OFDM). ZP-OFDM offers potentially superior performance in multipath channels only if more complex equalization, such as minimum-mean- squared-error (MMSE) or least-squares (LS) is performed at the receiver instead of the more commonly used overlap-add (OLA) equalizer. These complex schemes usually involve inversion of a non-diagonal, channel-dependent matrix that cannot be precomputed. In this paper we first develop two simplified algorithms for equalization of uncoded OFDM signals that improve performance in channels with spectral nulls compared to the OLA equalizer. We then develop a novel, iterative metric-calculation method for coded OFDM that with proper initialization requires only two iterations to approach the performance of a LS equalizer, without the need for matrix inversions. Simulation results will be presented to demonstrate the performance of the proposed algorithms.
Keywords
OFDM modulation; equalisers; least mean squares methods; multipath channels; wireless channels; WiMedia standard; coded OFDM systems; complex equalization; cyclic-prefix OFDM; iterative metric; least-squares equalizer; minimum-mean-squared-error; multiband-OFDM; multipath channels; wireless systems; zero-padded OFDM systems; Communications Society; Convolution; Equalizers; Iterative algorithms; Iterative decoding; Iterative methods; OFDM; Signal processing algorithms; Telephony; USA Councils;
fLanguage
English
Publisher
ieee
Conference_Titel
Communications, 2007. ICC '07. IEEE International Conference on
Conference_Location
Glasgow
Print_ISBN
1-4244-0353-7
Type
conf
DOI
10.1109/ICC.2007.703
Filename
4289374
Link To Document