Performance analysis of an over-sampling multi-channel equalization for a multi-band UWB system

Abstract

By dividing the available spectrum into multiple bands, multi-band ultra-wideband (UWB) systems possess many advantages, such as the flexibility to eliminate possible interference and coexistence problems with available wireless standards. UWB channels have large delay spread, causing inter- symbol interference (ISI) in high rate systems. The RAKE receiver is often used to capture multipath energy, but it is incapable of ISI compensation, causing severe bit error rate (BER) degradation in practical UWB channels. This paper proposes an over-sampling multi-channel equalizer per sub-band for multi-band UWB. The performance of the proposed receiver is compared with the RAKE receiver through analytical and simulation studies. The proposed equalizer has been shown to be able to handle ISI under any channel conditions or data rate presented, with an acceptable BER. In addition, the rich multipath diversity of the UWB channels is harnessed by the over-sampling scheme, for output SNR improvement. one. Therefore, equalization receiver based on adaptive algorithms seems to be an effective way to recover the transmitted data. With this consideration, we propose an over-sampling multi-channel equalizer per sub-band for a multi-band UWB system, which provides more effective inter-symbol interference (ISI) suppression compared with the conventional RAKE receiver when operating in channels with large delay spread. Through detailed theoretical analysis and extensive simulations, we are able to show that the proposed 4-arm equalizer scheme is superior to a 4-arm conventional RAKE receiver in the presence of ISI at all signal-to-noise ratios (SNRs). Based on the method presented in (4), the simulated performance is shown to conform to analytical predictions. The RAKE receiver is incapable of ISI compensation, and a moderate level of ISI is shown to cause severe bit error rate (BER) degradation. The proposed over-sampling equalizer, on the other hand, is able to handle ISI under any channel conditions or data rate, with an acceptable BER. Over- sampling is shown to be able to achieve diversity gain, and improves the output SNR.