Research On Bussgang Blind Equalization Algorithm

In modern communication system, ISI (Inter-channel Interference) and ICI (Inter-symbol Interference) which are two primary factors caused by distortion and aberration of limited bandwidth channel, can influence the communication quantities dramatically. Effectual channel equalization techniques have to be adopted to eliminate such unwilling factors. Blind equalization is a novel adaptive technique without training sequence, which utilizes the prior information of transmitted signal to equalize the channel characteristic. Using blind equalization, the output sequence of equalizer can be approximated with the transmitted signal as accurate as possible. This paper reviewed the background and history of blind equalization, and status of research of blind equalization was also been studied. The basic theory, principles of classification and structure of blind equalization is discussed further. The basic algorithm of blind equalization was researched in this paper, which focused on the optimization methods.The theory and property of Bussgang algorithm and systematically analyzes constant modulus blind equalization algorithm (CMA) are presented, which is one of the Bussgang algorithms. We also simulate the algorithm in Matlab language and introduce and analysis the inherent limitations of CMA, which lead to a tradeoff between convergence rate and misadjustment. The series of new variable step-size CMA were studied, and the theoretical analysis and the simulation results show that the new algorithms can solve the antinomy between the convergence rate and accuracy because of using the fixed step in traditional CMA.The CMA have a greater misjudgment for quadrature amplitude modulation (QAM) signal. In order to overcome the problem, this paper introduced the dual-mode blind equalization algorithm of a combination of CMA and decision directed (DD) algorithm. We also implemented the CMA+DD algorithm and propose leakage CMA+DD algorithm and momentum CMA+DD algorithm. The simulation results show that the performance of the proposed two new algorithms is better than CMA+DD algorithm. Finally, this paper focused on several new dual-mode blind equalization algorithms with phase-correction ability, and made analysis in length to dual-mode MCMA+DD algorithm. Through the comparison of simulation, an apparent conclusion was got that the dual-mode MCMA + DD algorithm had the best convergence property.