基于压缩感知的星型拓扑结构无线传感网络信号检测算法

A Compressed Sensing based Detection for Star Topology WSNs

针对星型拓扑结构无线传感网络中,不能及时地处理多个传感器节点同时采集到紧急事件并且立即向中心发送数据请求的问题,提出了一种有效的检测算法。因为同时激活的传感器节点数目相对于整个网络中传感器节点的总数来说是比较稀疏的,因此压缩感知理论能够用于这种情况的信号检测。此外,考虑到在整个传感网络,传感器节点数目非常大,因此,可采用李雅普诺夫中心极限定理估计每一个传感器节点是否被激活的状态概率,进而根据该概率估计选择可靠的支撑集元素并有效的克服误差传播的影响。最后,通过仿真实验验证本文提出算法的性能。

In traditional star topology WSNs,during aguaranteed time slots（GTSs）,only one sensor node is allowed to communicate with the AP.However,this scheme fails to deal with the scenario where more sensor nodes simultaneously gather urgent accidents and all of accidents require response from AP immediately.To address this problem,we consider the case where multiple sensors simultaneously send signal to the AP and propose an effective detector to distinguish multi-sensors signals in star WSN topology.Since the active sensor nodes are relatively sparse compared with the total number of sensor nodes,compressed sensing has been taken into account.Moreover,the number of sensor node is usually large in star topology WSNs,which satisfies the condition of Lyapunov central limit theorem.So,it is possible to estimate the probability of the state of each sensor node.Using the estimated probability,we can select more reliable support elements against the error-propagation effect.Simulation results substantiate the performance of the proposed algorithm.