摘要
利用目前方法对短距离无线通信网络异常信号进行识别时,没有对网络异常信号特征进行分类,存在异常信号识别准确率不高、不同信噪比下的识别正确率低和网络信号覆盖率差的问题。为此提出短距离无线通信网络异常信号识别方法,首先采用相像系数法对网络异常信号进行特征提取,同时归一化处理异常信号,从中获取了网络异常信号特征向量,将其输入到最小二乘支持向量机中进行分类,在最小二乘支持向量机分类建模期间,取得的超参数会对识别结果产生影响,所以采用灰狼优化算法对超参数进行优化,最终获取了最优解,完成对网络异常信号的识别。实验结果表明,通过上述方法进行异常信号识别准确率测试、不同信噪比下的识别正确率测试和网络信号覆盖率测试,验证了所提方法的有效性强,可靠性高。
During the abnormal signal recognition of short-range wireless communication network, the traditional methods have low signal recognition accuracy and poor coverage, because the characteristics of network abnormal signal are not classified. Therefore, the method of abnormal signal recognition in short-range wireless communication network was studied in the paper. The similarity coefficient method was utilized to extract the characteristics of network abnormal signals. Concurrently, these abnormal signals were normalized for obtaining the characteristic vector of network abnormal signals. The obtained feature vectors were input into the least squares support vector machine and then classified. In the process of least squares support vector machine classification modeling, because the recognition results were influenced by the super parameters, the gray wolf optimization algorithm was adopted to optimize the super parameters to obtain the optimal solution. At last, the network abnormal signals were identified. The experimental results show that this method has high signal recognition accuracy and signal coverage.
作者
张燕
刘磊
ZHANG Yan;LIU Lei(Chengdu College of University of Electronic Science and Technology of China,Chengdu Sichuan 611731,China)
出处
《计算机仿真》
北大核心
2022年第6期197-200,205,共5页
Computer Simulation
关键词
短距离
无线通信网络
异常信号识别
最小二乘支持向量机
灰狼优化算法
Short distance
Wireless communication network
Abnormal signal recognition
Least square support vector machine
Gray wolf optimization algorithm
