酉空时调制系统中最大似然多符号差分检测的性能分析

Performance Analysis of Maximum-Likelihood Multiple Symbol Differential Detection for Unitary Space-Time Modulation

在酉空时调制系统中,针对连续衰落信道下最大似然多符号差分检测给出了一种误符号率截断联合界的数值计算方法。采用Gauss-Chebyshev求积公式对成对错误概率进行数值计算,并利用提取主要错误事件技术对误符号率的联合界进行了化简。数值与仿真结果表明,采用该方法计算出的截断联合界可以在低信噪比下提供误符号率的一个上界,在较高信噪比下获得它的良好估计。在性能分析基础上进一步提出了一种近似最大似然度量,分析与仿真说明,采用该度量的多符号差分检测算法的性能非常接近最大似然检测。

A numerical method for calculating truncated union bound of symbol error rate （SER） is presented for maximum-like- lihood （ML） multiple symbol differential detection （MSDD） over continuous channel in unitary space-time modulation systems. Gauss- Chebyshev quadrature is applied to pairwise error probability calculation. Dominant error events are used to simplify the union bound of SER. Numerical and simulation results show that the truncated union bound obtained by the proposed method provides an upper bound of SER at low signal-to-noise ratios （SNRs）, while yields a good approximation of SER at moderate to high SNRs. Moreover, an approximate ML metric is proposed. Analysis and simulation results verify the performance of MSDD based on the approximate ML metric is very close to that of ML detection.