基于压缩感知和最小二乘的分布式协作频谱感知

针对认知无线电(CR)集中式频谱感知算法对融合中心要求高，而且对节点失效的容忍性也不高等缺点，提出了一种基于压缩感知的分布式多节点协作算法。认知无线电网络中每个CR节点在接收信号频谱后，首先根据压缩采样理论在本地获取压缩采样测量值，然后利用l1范数约束的最小二乘算法在本节点估计频谱，把在此节点估计的频谱传给下一相邻节点，以此进行迭代优化直到算法收敛。理论分析和仿真结果表明，所提算法不仅计算复杂度低，收敛速度快，而且精确重构性能好，可靠性较高。

Distributed cooperative spectrum sensing based on compressed sensing and least squares

For the shortcomings that centralized cognitive radio(CR) spectrum estimation sets strict requirement for fusion centers and has poor tolerance for node failure,this paper proposes a distributed multi-node cooperative algorithm based on compressed sensing. Each node of CR networks obtains the local compressed sampling according to compressed sampling theory firstly,then recovers the spectrum by exploiting l1 norm constrained algorithm. Finally,the spectrum estimated at the node is delivered to the next neighboring node until the algorithm converges. The theoretical analysis and simulation results show that this algorithm has not only low computational complexity and fast convergence speed,but also high accuracy and high reliability.