基于可信度的加权协作频谱检测算法

发布时间：2018-05-13 18:27

本文选题：认知无线电 + 协作频谱感知　；参考：《数据采集与处理》2014年03期

【摘要】：协作频谱检测技术可以有效解决认知无线网络本地检测存在的信道衰落、隐藏终端等问题。本文研究了加权协作频谱检测技术,提出一种基于可信度的协作频谱检测算法。每个感知节点基于最大最小值特征值检测完成本地频谱检测,并与融合中心的全局检测结果进行比较,估计各自感知节点的频谱检测可信度;融合中心利用切尾平均法计算参与频谱协作检测的可信度门限,并选择可信度大于门限的感知节点参与协作频谱检测。该算法有效降低了认知网络协作检测的复杂性,提高了频谱检测性能,在噪声波动环境下具有良好的鲁棒性。仿真结果表明,算法频谱检测性能要优于其他加权算法1~3dB,节省系统开销43.75%左右。
[Abstract]:Cooperative spectrum detection can effectively solve the problems of channel fading and hidden terminals in cognitive wireless networks. In this paper, the weighted cooperative spectrum detection technique is studied, and a novel cooperative spectrum detection algorithm based on credibility is proposed. Each sensing node completes the local spectrum detection based on the maximum and minimum eigenvalue detection, and compares with the global detection results of the fusion center to estimate the reliability of the spectrum detection of each sensing node. The fusion center calculates the confidence threshold of the cooperative spectrum detection using the tail-cutting averaging method, and selects the perceptual nodes whose credibility is greater than the threshold to participate in the cooperative spectrum detection. This algorithm effectively reduces the complexity of cooperative detection in cognitive networks, improves the performance of spectrum detection, and has good robustness in noise fluctuation environment. The simulation results show that the spectrum detection performance of the algorithm is better than that of other weighted algorithms, and the system cost is reduced by about 43.75%.
【作者单位】：南通大学电子信息学院;
【基金】：国家自然科学基金(61371112)资助项目江苏省普通高校研究生科研创新计划(CXZZ12-0866)资助项目南通大学研究生科技创新计划(YKC12069)资助项目
【分类号】：TN925

【参考文献】