分布式宽带合作频谱感知算法

Research on the distributed cooperative wideband spectrum sensing

针对传统宽带合作频谱感知中集中式频谱重构带来的网络通信负担重、重构复杂度高的缺点,提出了一种基于分布式子空间估计的非重构宽带合作频谱感知算法.该算法直接从欠采样数据中估计信号子空间,利用子空间的正交性实现频谱感知.为了获得空间分集增益,采用基于扩散自适应的分布式估计算法求解全局信号子空间.理论分析与仿真结果表明,该算法无需重构原始信号的频谱,运算复杂度低,全分布式的多用户合作提高了频谱感知性能,且具有更少的节点通信量.

In order to solve the problems of high communication overhead and computational complexity for reconstructing the sparse frequency spectrum in conventional cooperative wideband spectrum sensing, a novel decentralized algorithm based on distributed subspace estimation is proposed. In the proposed method, the subspace is estimated directly from the sub-Nyquist samples and then the orthogonality property of the signal subspace and noise subspace can be exploited to complete spectrum sensing. To obtain the spatial diversity gain, the global signal subspace is estimated by using the distributed algorithm based on the diffusion adaptation cooperation scheme. Theoretical analysis and simulation results show that the method does not reconstruct the original signal spectrum, with lower computational complexity. The fully distributed cooperative scheme improves the performance of spectrum sensing and has a lower communication amount.