Research On Connected Sensor Coverage Problem In Wireless Sensor Networks

In recent years, with the rapid development of micro technologies, theemergence of sensing, communication, data processing, storage capabilities sensorshave brought researchers’ strongly interest in wireless sensor networks (WSNs). In aWSN, the coverage reflects the physical sensing quality of the network duringmonitoring the target area, and the connectivity indicates the capability of sendingand receiving data in the WSN. As the two fundamental problems of the WSN,coverage and connectivity are the critical factors for energy consumption and thenetwork lifetime.During performing monitoring tasks in a WSN, if we select the involved sensornodes as few as possible, it can reduce the WSN energy consumption, improve theWSN energy efficiency, reduce the communication overhead and the redundancy ofthe sensing data among the proximity sensor nodes. Based on this consideration, thethesis considers the coverage and connectivity problem comprehensively in a WSN.The objective is to find a minimum connected sensor cover to meet covering theentire sensing region and the induced network is a connected network, thusextending the WSN lifetime.Firstly, this paper presents a node selection algorithm based on ConnectedBenefit to satisfy the requirements of the network coverage and connectivity. Byconsidering the contribution of nodes for the network coverage, this node selectionalgorithm first selects a set of nodes to meet the coverage condition, and then addsthe minimum number of nodes to the result set according to their Connected Benefit.Thus, we can select the minimum connected sensor cover to perform the monitoringtask.This paper then presents another new node selection algorithm based on theconcept of a minimum spanning tree (MST). At the connectivity phase, we choosenodes by constructing a minimum total weights MST to meet the coverage andconnectivity conditions.Finally, a Group Steiner Tree based on linear programming node selectionalgorithm is presented. By generating a Group Steiner Tree (GST), we use linear programming method to obtain a solution of the GST. Thus we can obtain theselected probability of each node and it will provide the correct direction to find theapproximate solution. At last, we use the random rounding technology to modify thesolution to meet the coverage and connectivity conditions.The algorithms based on Connected Benefit and MST both consider coverageand connectivity separately. The algorithm based on GST and linear programmingconsiders coverage and connectivity simultaneously. With different networkparameters, the performance of these algorithms is different. We verified theperformance of these algorithms through a simulation, and proved that thesealgorithms can ensure obtaining the least number of nodes to meet coverage andconnectivity conditions and present different efficient performance in distinctnetwork environments. In total, the proposed algorithms can reduce the WSN energyconsumption effectively and also extend the WSN lifetime.