移动传感器网络的安全性能智能预测  

Intelligent Prediction for Security Performance of Mobile Sensor Networks

作　　者：李振华 王涵 杜文才[2] LI Zhenhua;WANG Han;DU Wencai(Innovation and Entrepreneurship Institute,Yichun University,Yichun 336000,China;Institute of Data Science,City University of Macao,Macao 999078,China)

机构地区：[1]宜春学院创新创业学院,江西宜春336000 [2]澳门城市大学数据科学研究院,澳门999078

出　　处：《电讯技术》2021年第1期101-105,共5页Telecommunication Engineering

基　　金：国家自然科学基金资助项目(61901409);江西省自然科学基金(20202BABL212001)。

摘　　要：移动传感器网络的物理层安全问题日益复杂,已经成为了一个研究热点。为了及时处理网络安全事件,研究了移动传感器网络的安全性能预测,提出了一种基于灰狼优化广义回归(Grey Wolf Optimization-Generalized Regression,GWO-GR)神经网络的安全性能智能预测方法。该方法利用发射天线选择策略,推导了非零安全容量概率性能的精确闭合表达式。仿真比较了所提方法、反向传播神经网络、广义回归神经网络、支持向量机等方法,结果表明,所提方法可以实现更好的预测性能,提高安全性能预测的实时性。The physical layer security of the mobile sensor network is becoming complex,which has been a research hotspot.In order to deal with network security issues timely,the secrecy performance prediction is investigated,and a grey wolf optimization-generalized regression(GWO-GR)neural network-based secrecy performance prediction method is proposed.With transmit antenna selection(TAS)scheme,the exact closed-form expressions for the probability of strictly positive secrecy capacity is derived.Compared with the back propagation(BP)neural network,generalized regression(GR)neural network,and support vector machine(SVM)methods,the simulation results show that the proposed method can achieve higher prediction results and improve the real-time performance of prediction.

关 键 词：移动传感器网络 物理层安全 安全性能智能预测 GWO-GR神经网络 

分 类 号：TN929.5[电子电信—通信与信息系统]