分级分布式无人机协作频谱感知方法

A Hierarchical Distributed UAV Cooperative Spectrum Sensing Method

无人机认知网络(Cognitive UAV Network,CUAVN)可以通过高精度的频谱感知提高频谱效率,但传统的集中式协作频谱感知不适用于无人机(Unmanned Aerial Vehicle,UAV)网络。UAV网络拓扑结构动态变化,使得全局信息很难快速收集至一个融合中心。针对上述问题,提出了一种基于K-means聚类算法的分级分布式协作频谱感知方法。首先,利用K-means聚类算法基于位置信息对UAV进行分簇;然后,采用两级分布式融合方案进行分层融合,在每个簇内先进行一致性信息融合得到K个融合结果,再进行第二次融合得到最终全局收敛结果;最后将最终收敛结果与检测阈值对比,得到最终决策。仿真结果表明,所提出的分级分布式融合方案具有较好的收敛性,且信息融合迭代次数比未分级融合方案更少。在不同权重因子的加权情况下,所提方案较未分级融合方案更好,且平均接收信噪比越大时检测性能越好。与未分级分布式融合方案相比,在UAV节点较多的情况下,该方案使CUAVN的频谱检测性能得到了提升。

Cognitive unmanned aerial vehicle network(CUAVN)can improve spectrum efficiency through high accuracy spectrum sensing,but traditional centralized collaborative spectrum sensing is not suitable for unmanned aerial vehicle(UAV)networks.The dynamic change of UAV network topology makes it difficult to collect global information into a fusion center quickly.For this problem,a hierarchical distributed collaborative spectrum sensing method based on K-means clustering algorithm is proposed.First,the UAV is clustered based on the location information using the K-means clustering algorithm.Then,a two-level distributed fusion scheme is used for hierarchical fusion,in which information is first fused within each cluster to obtain K fusion results,and then a second fusion is performed to obtain the final global convergence results.Finally,the final convergence results are compared with the detection threshold to obtain the final decision.The simulation results show that the proposed scheme has better convergence performance and the number of information fusion iterations is less than that of the un-hierarchical fusion scheme.In the case of weighting with different weighting factors,the proposed scheme is better than the un-hierarchical fusion scheme,and the detection performance is better when the average received signal-to-noise ratio(SNR)is larger.Compared with the un-hierarchical distributed fusion scheme,this scheme leads to the improved spectrum detection performance of CUAVNs in the case of more UAV nodes.