WSN中一种基于混合CS的分簇数据收集方案

针对现有数据收集方案的不足, 提出了一种基于混合CS的分簇数据收集方案。首先网络被划分为多个簇, 簇内节点的数据通过最短路径路由直接发送到簇头上, 而簇头基于压缩感知（CS）对收到的数据进行融合, 并通过一棵骨干路由树将各个簇头的数据发送到sink上, 以最小化数据传输总次数为优化目标, 对簇的大小与数据传输次数之间的关系进行了理论分析, 得到了可以保证数据传输次数最小的簇大小。进一步地, 簇内的数据传输问题被建模为K-median问题, 簇间的数据传输问题被建模为Steiner树问题, 并分别提出了集中式和分布式算法来解决该问题。仿真实验结果表明, 该方案是有效的, 在降低数据传输次数方面要优于传统的方法。

Clustering data gathering scheme based on hybrid CS in WSN

Aiming at the disadvantages in existing data gathering schemes, this paper proposed a clustering data gathering scheme based on the hybrid CS. Firstly, it divided the network into multi-clusters, and sent the original data of nodes directly to the cluster head via the shortest path route. Then the cluster head aggregated the received data using CS and forwarded them to the sink via a backbone routing tree. The minimizing the number of transmission as the optimal objective among them, this paper theoretically analyzed the relationship between the size of clusters and number of transmissions in the hybrid CS method, and it obtained the optimal size of clusters that led to minimum number of transmissions. Furthermore, it modeled the intra-cluster data transmission as the K-median problem, and modeled the inter-cluster data transmission as the Steiner tree problem, and proposed the centralized and distributed algorithms to solve them. The simulation results show that the proposed scheme is effective, and is better than the traditional methods in terms of reducing the number of data transmission.