直扩系统中基于SVM的干扰自动分类识别方法  

An Automatic Interference Recognition Method in DSSS Communication System Based on SVM

作　　者：于波[1] 邵高平[1] 孙红胜[1] 任孝民[1] 

机构地区：[1]解放军信息工程大学信息工程学院,河南郑州450002

出　　处：《信号处理》2010年第10期1539-1543,共5页Journal of Signal Processing

摘　　要：针对实际直接序列扩频系统中的干扰类型总是未知的,难以采用相应的干扰抑制措施有效提高系统性能的问题,提出了直接序列扩频系统中基于支持向量机的干扰自动分类识别方法。首先对直扩系统中八种主要干扰情况进行干扰信号建模;然后分别从时域、频域和时频域提取待识别干扰信号的特征参数;最后使用决策树支持向量机构造了一个5级干扰自动分类识别系统,实现了干扰信号的自动分类识别。仿真结果表明,干信比大于等于9dB时,分类器可以准确识别干扰类型。To solve the problem that we can＇t take the right interference suppression method to improve the performance of direct sequence spread spectrum（DSSS） communication system because of the unknown interference types,an algorithm to automatically recognize interferences is proposed.First,the mathematical models of 8 important interferences in DSSS communication system are made;Then,the features of interferences from time domain,frequency domain and time-frequency domain are extracted;And finally, a fifth-order automatic interference recognizer based on decision tree support vector machine is constructed,and the automatic recognition of the interferences is realized.Simulation results demonstrate that the proposed algorithm provides a very high recognition rate when ISR isn＇t lower than 9dB.

关 键 词：直扩系统 干扰信号 支持向量机 特征提取 分类识别 

分 类 号：TN911.72[电子电信—通信与信息系统]