基于向量加权平均优化的盲源分离LTE-M同频干扰检测  

Blind Source Separation LTE-M Co-Channel Interference Detection Based on Vector Weighted Average Optimization

作　　者：王乃鑫 赵恒凯[1] 何庆军 郑国莘[1] 

机构地区：[1]上海大学特种光纤与光接入网重点实验室,上海先进通信与数据科学研究院,上海200444 [2]上海中铁通信信号测试有限公司,上海200072

出　　处：《工业控制计算机》2025年第2期45-47,共3页Industrial Control Computer

基　　金：2022-2025上海市科委重点项目(22ZR1422200):轨道交通高可信车地协同与高可靠车车通信研究。

摘　　要：轨道交通LTE-M(Long Term Evolution-Metro,基于轨道交通的长期演进)同频干扰检测关乎列控信号传输的可靠性,提出一种基于INFO(weIghted meaNoFvectOrs,基于向量加权平均)算法的盲源分离方法,即INFO-BSS。该方法以混合信号的最大化负熵为目标函数,用INFO优化算法替代牛顿迭代法,解决了牛顿迭代法初始参数易设置不当以及容易陷入局部最优的问题。仿真结果对比表明,在不同分辨率带宽、不同信干比等条件下,INFO-BSS的检测性能都要优于常规算法。Long Term Evolution Metro(LTE-M)co frequency interference detection in rail transit is crucial for the reliability of train control signal transmission.This paper proposes a blind source separation method based on the INFO(weIghted meaN oF vectOrs)algorithm,namely INFO-BSS.This method takes the maximization of negative entropy of mixed signals as the objective function and replaces the Newton iteration method with the INFO optimization algorithm,solving the problems of improper initial parameter setting and easy falling into local optima in the Newton iteration method.The comparison of simulation results shows that the detection performance of INFO-BSS is superior to conventional algorithms under different resolution bandwidth,signal-to-interference ratio,and other conditions.

关 键 词：同频干扰 LTE-M 盲源分离 向量加权平均优化算法 

分 类 号：TN97[电子电信—信号与信息处理] U285.21[电子电信—信息与通信工程]