Adaptive equalisation of frequency selective multipath fading channels based on sample selection

Abstract

The behaviour of the frequency selective multipath fading channel has been studied for normalised RMS delay spread over the range from 0.01 to 1.0. This corresponds to 1ns-100ns RMS delay spread for a system operating at 10Mb/s. Over this range, errors occur in bursts because of fading, intrasymbol interference (IaSI) when the channel spread is low and intersymbol interference (ISI) when the channel spread is high, When SNR is sufficiently high, the channel is rarely in deep fade, and equalisers in such channels may improve system performance. It is shown that conventional decision feedback equalisers (DFE) with fixed midpoint sampling do not perform well in the range of RMS delay spread considered for realising a BER less than or equal to 10(-3). A new equalisation technique is introduced, namely, the sample-diversity (SD) DFE. To realise sample diversity at the DFE input, N samples are taken from each symbol and these samples are used to drive N different DFEs. A simple selection mechanism is introduced to select the best sampling phase and DFE to be used in the tracking mode. Through simulations, the SDDFE is shown to have much better performance than conventional DFE. The SD technique is also applied to coherent-QPSK (CQPSK) detection where N parallel detectors are used. Being a combination of diversity and equalisation, the SD approach is shown as a promising technique for symbol detection in multipath fading channels.