一种基于ZP-OTFS的低复杂度SSOR检测算法  

A Low Complexity SSOR Detection Algorithm Based on ZP-OTFS

作　　者：何茂恒 张薇[1] HE Maoheng;ZHANG Wei(College of Information and Communication Engineering,Harbin Engineering University,Harbin 150001,China)

机构地区：[1]哈尔滨工程大学信息与通信工程学院,哈尔滨150001

出　　处：《电讯技术》2025年第2期223-230,共8页Telecommunication Engineering

基　　金：电子信息系统复杂电磁环境效应国家重点实验室资助课题(CEMEE2021K0101A)。

摘　　要：针对高速移动场景中正交时频空间(Orthogonal Time Frequency Space, OTFS)系统线性最小均方误差(Linear Minimum Mean Square Error, LMMSE)检测复杂度过高而难以快速有效实现的问题,利用零填充(Zero Padding, ZP)OTFS系统时域信道矩阵呈块对角稀疏特性提出一种逐块迭代的对称逐次超松弛(Symmetric Successive over Relaxation, SSOR)迭代算法,在降低系统复杂度的同时获得与LMMSE检测近似的性能。仿真结果表明,与逐次超松弛(Successive over Relaxation, SOR)算法相比,所提算法对松弛参数不敏感且具有更快的收敛速度,在迭代次数为10次时误码性能几乎达到LMMSE误码性能,显著降低了检测器的复杂度。To solve the problem that the linear minimum mean square error(LMMSE)detection complexity of orthogonal time frequency space(OTFS)system is too high to be realized quickly and effectively in high-speed mobile scenes,a block-by-block Symmetric Successive over Relaxation(SSOR)iterative algorithm is proposed based on the block-diagonal sparsity of the time-domain channel matrix of zero-padding(ZP)OTFS system.The system complexity is reduced and the performance is similar to that of LMMSE detection.Simulation results show that compared with Successive over Relaxation(SOR)algorithm,the proposed algorithm is not sensitive to the relaxation parameter and has faster convergence speed.When the number of iterations is 10,the bit error performance of the proposed algorithm almost reaches the LMMSE bit error performance,which significantly reduces the complexity of the detector.

关 键 词：ZP-OTFS 线性最小均方误差(LMMSE) 信号检测 SSOR迭代检测 

分 类 号：TN929.5[电子电信—通信与信息系统]