认知无线电中基于分簇的能量序贯协作检测算法

CLUSTERING-BASED ENERGY SEQUENTIAL COLLABORATIVE DETECTION ALGORITHM IN COGNITIVE RADIO

在认知无线电系统中,频谱感知技术是保证频谱资源分配以及重构网络参数的前提条件。针对复杂信道环境低信噪比下传统频谱感知技术感知性能恶化、感知时间变长和对授权用户位置不固定适应性差的问题,提出基于分簇的并行能量序贯检测的方法。该方法按照地理位置信息划分簇,按照减少传输误差和节省传输功率原则选取簇头,各簇分段并行计算似然比统计量的方法进行序贯检测,任一簇内判决成功即可结束检测。仿真结果表明,在低信噪比情况下,该算法相较现有的多节点分段序贯检测和单节点序贯检测可以有效缩短检测时间。分析分簇数对所提算法系统综合性能的影响,结果表明信噪比越低,选择分簇数越大,越能降低检测时间。

In cognitive radio system,spectrum sensing technology is the precondition to ensure spectrum resource allocation and reconstruct network parameters.In view of the deterioration of traditional spectrum sensing technology in complex channel environment with low signal-to-noise ratio (SNR),long sensing time and poor adaptability to the location of authorized users,this paper proposed a clustering-based parallel sequential detection method.We classified clusters according to geographic location information,and the cluster head was selected according to the principle of reducing transmission error and saving transmission power.Cluster head was sequentially detected by parallel computation of likelihood ratio statistics in segments.The detection could be completed if the intra-cluster decision was successful.The simulation results show that the proposed algorithm can effectively shorten the detection time compared with the existing multi-node subsection sequential detection and single-node sequential detection under low signal-to-noise ratio.The impact of cluster number on the overall performance of the proposed algorithm system is analyzed emphatically.The results show that the lower the signal-to-noise ratio,the larger the number of clusters selected,the lower the detection time.