On the wavelet families for OFDM system comparisons over AWGN and Rayleigh Channels

Abstract

In the study of OFDM systems, discrete wavelet transforms have been reported to perform better than Fast Fourier Transform in multicarrier systems (MCS) - in terms of spectral efficiency because they can operate without a cyclic prefix, have reduced side-lobes and improved BER. However all of the wavelet families do not perform alike. This study has investigated various wavelet families such as Daubechies, Symlet, Haar (or db1), biorthogonal, reverse-biorthogonal and Coiflets for OFDM system design over an AWGN and multipath channels. Results show that Daubechies, Symlet, Haar and Coiflet wavelet families perform considerably better than other families considered, thus these families could be better in OFDM.