由于认知网络内用户物理分布具有非均匀特性，各认知用户与相同主用户之间的距离及其信道环境存在较大差异，导致不同用户判决结果的可靠性程度不同。传统的硬融合算法往往忽视用户检测可靠性差异，同等对待所有本地判决，使得合作检测性能达不到最优。针对该问题，提出一种改进的基于K秩硬判决的多用户合作频谱感知算法。融合中心执行融合算法时，引入信任度区别对待各认知用户的本地判决，采用迭代方法，不断根据用户局部判决相对总判决的正确度，自适应更新信任度，并适时调整全局判决门限。算法简单有效，开销小。实验结果表明，该算法相比传统的K秩硬判决算法，具有更好的合作检测性能。%Owing to the non-uniform feature of physical distribution of users in the cognitive network,the distance and channel condition between each cognitive user and the same primary user are quite different,resulting in different levels of detection reliability among differ-ent users. Traditional hard fusion algorithms tend to ignore the difference of detection reliability,equally treat all local decisions,making the cooperative detection performance beyond optimal. To solve this problem,present an improved cooperative spectrum sensing method which is based on K-rank hard judgment. In implementing the fusion algorithm,the fusion center introduces credibility to differentiate the way to treat different cognitive user's local judgment,adopts iterative method to update credibility based on the accuracy of the user's par-tial judgment relative to the total judgment,and timely adjusts the global decision threshold. The algorithm is simple and effective and low overhead is introduced. The experimental results show that compared with the traditional K rank hard decision algorithm,the improved al-gorithm achieves better cooperation detection performance.
展开▼
