一种低复杂实现框架及在多码元差分检测中的应用

本文基于信息论考虑提出了一种低复杂实现的框架，对最大似然序列检测的分组（block)检测结构分析了该低复杂实现方案的性能并提出两命题，可用于寻求低复杂实现算法，另外，我们将该低复杂实现框架成功应用于高斯白声信道及衰落信道下多码元差分检测（MSDDD，multi-symbol differential detection)的低复杂实现，在基本保持其的条件下实现了检测复杂度多码元观察长度的线性增长，而最佳MSDD复杂度呈指数增工。

A general framework for low complexity implementation and its application in muti-symbol differential detection

A general framework for low complexity block detection is proposed based on the information theory Two propositions are presented and may be employed to find low complexity algorithms We apply the proposed framework to the problem of finding low complexity algorithms for muti symbol differential detection(MSDD)over either AWGN of flat correlated Rayleigh fading channels and a class of fast algorithms are presented The complexity of new fast algorithms increases roughly linearly with block length N while optimum MSDD increases exponentially with N Simulation results show almost the same performance for practical N