基于信息论考虑提出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噪声信道及衰落信道下多码元差分检测 MSDD(Multi- Symbol Differentail Detection)的低复杂实现 ,在基本保持其性能的条件下实现了检测复杂度随多码元观察长度的线性增长 ,而最佳 MSDD复杂度呈指数增长 Based on the information theory, this paper proposes a framework of low complexity implementation, analyzes the performance of the low complexity scheme for the detection structure of the maximum likelihood sequence and proposes two propositions to find the algorithm with low complexity. In addition, we successfully apply this low-complexity implementation framework to the low-complexity implementation of multi-symbol differential detection (MSDD) in Gaussian white noise channels and fading channels, and achieve complex detection with basically maintaining its performance The degree increases linearly with the length of the multi-symbol observation, and the optimal MSDD complexity increases exponentially