基于BP神经网络的认知无线电疑似跳转信道预测研究  

SHC Prediction Scheme based on BP Neural Network for Cognitive Radio

作　　者：项亚楠 徐以涛[1] 项楠 丁国如[1] 徐庆龙[1] 

机构地区：[1]解放军理工大学通信工程学院,江苏南京210007 [2]湖北移动设计中心,湖北武汉430000

出　　处：《通信技术》2015年第3期311-317,共7页Communications Technology

基　　金：国家自然科学基金项目(No.61172062;No.61301160);江苏省自然科学基金项目(No.BK2011116)~~

摘　　要：为了在认知无线电环境中进行跟踪干扰,需要预测认知用户在遭受干扰后进行频谱切换的目标信道——疑似跳转信道,以便在这些疑似跳转信道上对其进行跟踪检测和干扰。提出了一种在多信道多用户条件下,认知用户在感知得到的空闲信道上智能决策并选择合适信道接入的频谱模型,并用基于BP神经网络的预测算法预测疑似跳转信道,为衡量预测效果,提出了预测切准率和预测精度两个参考标准,仿真结果验证了该算法的可行性。To implement the tracking jam in CR（ cognitive radio）communication,it is necessary to predict the target channel—SHC（ Suspicious Hopping Channel）—that CR may hop to after suffering a certain interference,so as to track and jam it in SHC.A spectrum model with multiple channels and multiple users is proposed,in which the CR user can intelligently select and access the appropriate channel from the sensed idle channels.Then the algorithm based on BP neural network is used to predict the SHC.In order to measure the prediction effect,two reference standards—handoff accuracy and precession are proposed.Finally,Simulation results verify the feasibility of this prediction algorithm.

关 键 词：认知无线电 疑似跳转信道 频谱切换 频谱模型 BP神经网络 

分 类 号：TN97[电子电信—信号与信息处理]