Multi-path Fading Channel Adaptive Equalization Technology Research

In digital microwave communication, inter-symbol interference (ISI) is inevitable because of multi-path fading, finiteness of channel bandwidth and faultiness of channel. However, adaptive equalization technique can decrease the effect of ISI and noise, reduce the bit error ratio and trace time-changing channel. So it is necessary to add an equalizer to the physical receiver. It can enhance the adaptability of broadband digital transmission apparatus and effectively decrease the performance requirement of channel and receiver, accordingly can reduce cost.The primary task of this dissertation is to design a hardware system of adaptive equalizer in the digital microwave communication system. First of all, the characteristics of different adaptive algorithms and equalizers are theoretically researched. Then digital microwave communication system model is build by Simulink. Based on this model, the convergent behavior and bit error rate performance of three equalizers (NLMS, Kalman and RLS) are compared in multi-path fading channel. Conclusions can be made that NLMS equalizer is the best among three equalizers in terms of convergent stability and calculating complexity, Kalman equalizer is the best among three equalizers in terms of bit error rate. And it turns out that the bit error rate of passband equalizer is smaller than that of baseband equalizer, but the convergent stability of baseband equalizer is better than that of passband equalizer. At the same time, baseband equalizer can be implemented in hardware more easily than passband equalizer.Finally, LMS linear equalizer is designed based on FPGA because of its good performance, simple structure and low calculating complexity. And behavior simulation, synthesis and implementation of this design are carried out.As a result, the design can be implemented in hardware. However, since the fixed-point calculation maybe cause cumulative quantization error, the convergent performance of fixed-point equalizer is not better than that of floating-point equalizer.