The Research On Equalization Algorithms Based On Reduced-State

Because of the complicated space propagation environment in mobile communication systems, there are path propagation loss, slow fading loss and various fast fading loss. This thesis analyzes and researches the equalization technique and algorithms which combat frequency selective fading in multi-decimal digital modulation systems. This thesis also discusses the problems of applying equalization algorithms under the ISI circumstance, and summarizes the proposed reduced-state sequence estimation (RSSE) and delayed decision-feedback sequence estimation (DDFSE) equalization algorithms based on reduced-state. And we illustrate the algorithms procedures and compare these two algorithms. In this thesis, we discuss the value of soft information output. When the soft information output is necessary, we discuss the problem of using soft-output Viterbi algorithm (SOVA) based on reduced-state in multi-decimal digital modulation systems. In order to solve the problem, we summarize different proposed soft information generation methods and propose the realization algorithm of one method. This thesis also proposes one improved algorithm and simulates the different algorithms. The performance of improved algorithm is comparable to the best soft information generation method while its computational complexity is lower. In addition, this thesis also analyzes the common minimum mean square error-decision feedback sequence estimator (MMSE-DFSE) structure in the multi-input multi-output (MIMO) sequence estimation. And we also summarize some equalization algorithms applied in this structure, such as vector Viterbi algorithm (VVA), partitioned Viterbi algorithm (PVA), decoupled Viterbi algorithm (DVA) and joint reduced-state sequence estimation (JRSSE) algorithm. In addition, we analyze their computational complexity.