基于余弦代价函数的T/4分数间隔盲均衡算法

Cosine cost function based T/4 fractionally spaced blind equalization algorithm

针对常模算法(CMA)对非常模高阶QAM信号均衡后收敛速度慢、稳态均方误差大的缺点,以T/4分数间隔均衡器为例,提出一种基于余弦代价函数的T/4分数间隔盲均衡算法(T/4-FSE-CCF),该算法将常模的代价函数用构造的余弦代价函数来替代,新算法摆脱了常数模算法对统计模值R的依赖。最后对算法性能进行了理论分析和仿真实验,结果表明,该算法对高阶QAM信号不仅提高了收敛速度而且降低了稳态均方误差。

Aiming at the slow convergence rates and the large steady-state mean square error of constant modulus algorithm（CMA） for high-order QAM signal,taking the T/4 fractionally spaced equalizer for example,this paper proposed a T/4 fractionally spaced equalization algorithm based on cosine cost function（ T/4-FSE-CCF）. It used cosine cost function to replace the cost function of CMA,the new algorithm got rid of the dependence on statistical modulus value R compared with CMA.This paper analyzed the performance of the algorithm by theoretical analysis and simulation,and the results show that the T/4-FSE-CCF has not only the faster convergence rates but also the smaller steady-state mean square error for high-order QAM signal.