基于改进SVM的通信干扰识别

Communication jamming recognition based on improved SVM

目前直扩通信系统的抗干扰算法仅能够针对一种干扰类型进行抑制,因此有必要对各类干扰进行识别,从而选取相应的抑制算法。针对通信干扰的识别方法进行研究,建立基于SVM算法的通信干扰识别模型。常规SVM算法性能依赖于参数选取,而常规SVM算法参数选取多是通过人工经验选取,存在较大的随机性和盲目性,因此该文使用遗传免疫粒子群优化SVM算法,提高算法和识别模型的性能。使用Matlab针对研究的通信干扰识别模型进行仿真。研究结果表明,所建立的通信干扰识别模型性能优于常规SVM算法和BP算法建立的识别模型。

The anti jamming algorithm of DSSS communication system can only restrain a type of interference, so it is necessary to identify all kinds of interference, so as to select the corresponding suppression algorithm. The recognition methods of communication jamming are studied in this paper, A communication jamming recognition model based on SVM algorithm is estab- lished. Since the properties of conventional SVM algorithm depend on parameter selection, and the parameter selection of the conventional SVM algorithm is selected by means of artificial experience, there are certain randomness and blindness. There- fore, genetic immune particle group optimization SVM algorithm is used in this paper to improve the performances of the algo- rithm and recognition model. Simulation study on the communication interference recognition model is conducted with Matlab. The results show that the the performance of the communication interference identification model established in this paper is superior to that of the conventional SVM algorithm and BP algorithm.