Research On Channel Equalization Technique For OFDM System

The signals of the communication system have various kinds of interference in thefading channel. Meanwhile, the non-ideal characteristics and the limited features of thetransmission bandwidth exist in the wireless channel. These effects can make the signalsproduce a serious distortion. Finally it cannot recover the transmission signal completelywhich carries the amount of information. The feature of channel equalization algorithm hasthe ability to compensate the system feature losses effectively. It can also eliminate theinter-carrier interference (ICI) and the inter-symbol interference (ISI) produced by themulti-path fading channel.The channel equalization technology has played an important role in the mobilecommunication systems. Now the convergence rate, steady-state error, the structuralcomplexity and the amount of calculating of the traditional equalization algorithms need tobe improved. With the development of the equalization technology, these problems will beencountered by the corresponding solutions. For the orthogonal frequency divisionmultiplexing system (OFDM), the channel compensation equalization algorithms and theimproved algorithm are researched. The paper includes:1) The research summarizes present situation and significance of the equalizationtechnology which are studied, and the background of the traditional equalization algorithmsis introduced. The environmental characteristics of the radio channel and some parametersof fading channels are analyzed. Then the basic principles and the advantages anddisadvantages of the OFDM system are introduced.2) The corresponding channel compensation equalization algorithm is studied for theOFDM system by comparing four traditional equalization algorithms. The advantages anddisadvantages of the respective algorithms are analyzed. The Bit Error Rate (BER)performance and the mean square error performance of these algorithms are also compared,and the simulation results are relatively good.3) The Least-Mean-Square (LMS) equalization algorithm whose step size is controlledby the SNR parameter is proposed. The improved algorithms are studied in theory. TheRecursive Least-Square (RLS) equalization algorithm based on the subtraction factor isproposed. The basic principles of the improved equalization algorithm are analyzed. A lot ofcomputer simulations are completed by the improvement of the important parameter stepsize and the forgetting factor. Simulation results show that the two improved algorithmsbased on the parameters can significantly reduce the minimum mean square error values andthe error rate through the appropriate parameter setting comparing with the originalalgorithm in the time-varying fading channel environment. And to some extent, a good optimization of the performance of the transceiver signals can be received in the OFDMsystem.