中文摘要：

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。

英文摘要：

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.