Nouvel algorithme pour le raccourcissement de canal UWB

Abstract

The Ultra Wide Band (UWB) is a radio transmission technique that uses signals whose spectrum extends over a wide band of frequencies, typically in the range of 500 MHz to several GHz. It was first used for radar applications and then transposed to telecommunications applications, thus arousing a growing interest in the academic and industrial community. The main features of this technology give it the status of a promising candidate for high-speed short-range radio communications systems. This thesis is part of the study of the application of the Channel Shortening Equalizer (CSE) method to the Ultra Wide Band called the Maximum Shortening Signal to Noise Ratio (MSSNR). The latter consists in maximizing the signal-to-noise ratio so that the UWB system is immunized against intersymbol interference (ISI) and the complexity of the Rake receiver is reduced. This thesis deals with: - The analysis of the UWB signal, the main characteristics and their applications. - The presentation of the different channel models used in the UWB, the CSE equalization techniques that reduce the effects of interference and thus increase the performance of the multi-access TH-UWB transmission system. - Proposition of a new algorithm based on the energy concentration of the effective channel impulse response (Channel-CSE) in a desired window using the optimization method "Singular Value Decomposition (SVD)". This method gives better BER performance compared to the optimal MSSNR method and further reduces the computational complexity of the P-Rake receiver compared to the S-Rake and All-Rake receivers. - Proposal of another method of optimization of a CSE equalizer based on the maximization of the ratio between the useful signal and the sum of the noise, the ISI and the multiple access interference (MAI).