Adaptive Weighted Energy Detection Receivers For Noncoherent UWB Communication Systems

Ultra-wideband(UWB) technology has been currently considered as one of the most promising candidates for short-range wireless communication due to its low power density,high data rate and low cost. Receiver design is a key problem for the implementation of UWB systems. Therefore, the performance evaluation of the UWB receivers in fading channel is of great significance. In low-rate UWB communication systems, due to the considerations in terms of cost and power consumption, non-coherent(NC) receivers with low-complexity are generally adopted without the need of channel estimation due to their relaxed requirements of the sampling rate. At present, Energy Detection(ED) receiver and Autocorrelation Receiver(ACR) are two popular non-coherent UWB receivers. In this dissertation, we investigate the weighted ED(WED) receivers and our work is focused on its adaptive implementation. The main contributions of this dissertation are summarized as follows.1. Classical adaptive algorithms are evaluated for WED receivers, such as normalized least mean square(NLMS) algorithm, recursive least square(RLS) algorithm and LMS-Newton algorithm. That is, the WED receivers are further developed to perform updating the weighting coefficients adaptively via adaptive algorithms to optimize the weighting coefficients without the need of complicated matrix manipulation and channel estimation. Moreover, through the analysis of convergence behavior and the bit error rate(BER) performance of the three algorithms in UWB channel models with different numbers of weighting coefficients, we present the BER performance of WED receivers under the optimum number of weighting coefficients. Simulation results show that in terms of the BER performance and convergence speed, the RLS outperforms the LMS-based algorithms.Compared to the RLS, NLMS and LMS-Newton methods can obtain comparable BER performance with the penalty in the convergence speed. At the same time, the NLMS possesses the benefit of computational complexity among all investigated adaptive algorithms.Compared to the conventional ED receivers, the adaptive WED receiver can obtain about0.5~1.2 dB-gain in terms of the BER performance.2. We further developed the improved NLMS(INLMS) algorithm to get the parameteroptimized INLMS algorithm, and its convergence behavior and BER performance are analyzed in WED receivers. Our analysis and simulation results show that the parameter optimized INLMS algorithm outperforms the traditional NLMS algorithm in terms of the BER performance by 0.3 dB. Despite the advantages of INLMS algorithm in the BER performance is not significant, in terms of convergence speed, it has obvious advantages.