一种基于改进加权序贯检测的协作频谱感知算法

摘要

The sequential detection can improve the accuracy of cooperative spectrum sensing,but when the spectrum sensing data falsification node exists in the environment,the system performance will decline dramatically.In order to solve the prob-lems above,this paper proposed a spectrum sensing fusion algorithm that based on weighted sequential probability ratio test which optimized the system performance by giving reputation weights to each node and set a reasonable reputation count scheme to distinguish SSDF nodes.Simulation results show that the proposed algorithm reduce the influence brought by SSDF nodes, improve detection accuracy and stability of the system.%通过序贯检测可以提高协作频谱感知的准确度，但是在具有频谱感知数据窜改（spectrum sensing data falsification，SSDF）节点的环境下，系统感知性能急剧下降。为了解决上述问题，提出了一种基于加权序贯检测（weighted sequential probability ratio test，WSPRT）的频谱感知融合算法。通过给每个节点赋予信誉度权值，设置合理的信誉度奖惩方案来区分 SSDF 节点，从而优化系统感知性能。仿真实验表明，所提出的算法减少了 SSDF节点带来的影响，提高了系统的检测准确率和稳定性。