基于证据预处理的协作频谱感知方法

Cooperative spectrum sensing method based on evidence preprocessing

针对传统基于DS证据理论的协作频谱感知方法中,由于信道环境的复杂性和DS合成公式的弊端,导致的证据悖论问题,本文提出了基于证据预处理的协作频谱感知方法。该方法对证据值预处理,利用静态权重提高自身可靠性,并利用动态权重确定冲突数据并修正,提出的合成公式对预处理后的证据值进行融合并判决。仿真结果表明:该方法能有效解决证据悖论问题,提高检测性能,鲁棒性强,适用于故障感知用户少于正常感知用户的情况。

A cooperative spectrum sensing method based on evidence preprocessing is proposed to solve the evidence paradox problem caused by the channel environment complexity and the DS(Dempster-Shafer)composition formula shortcomings in the traditional cooperative spectrum sensing method based on DS evidence theory.This technique first preprocesses the evidence value,then employs static weights to improve its reliability,following the usage of dynamic weights to determine and correct the conflicting data,and finally proposes a new composition formula to fuse and judge the preprocessed evidence values.The simulation results reveal that the proposed method effectively addresses the evidence paradox problem,improves detection performance,and has strong robustness.It is suitable for situations with fewer fault sensing users than normal sensing users.