Research On Node Deployment And Topology Reconfiguration Of Wireless Sensor Networks

Wireless sensor network is a self-organized network system, generally composed of a number of sensor nodes deployed in specific monitoring region and constituted by wireless communication, aims to collaborate sensing, acquiring and processing object attribute information in the network coverage region, and transmits to observer. To make a wireless sensor network work and effectively realize each function, the fundamental step required is deploying all the sensor nodes to form a wireless sensor network in a scientific manner. The placement of nodes largely influences the performance of networks. Moreover, sensor placement also affects the resource management in networks. However, the quality of service may no longer meet the application requirements after a period of network operation. To solve this problem, this thesis researches on the sensor deployment and the topology reconfiguration of wireless sensor networks. The main contributions are as follows:(1) The problems of sensor deployment and topology reconfiguration for wireless directional sensor network whose sensing ability is heterogeneous are studied. For deterministic network and random network, centralized Greedy Optimal Sensor Deployment Algorithm and Distributed Node Orientation Adjustment Algorithm are proposed respectively to make the network meet the coverage requirement while reducing network configuration cost and the number of active sensor nodes, further reducing energy consumption and improving communication quality. Moreover, Distributed Redundant Node Scheduling for network topology reconfiguration is designed to balance the network energy consumption.(2) This paper presents an event driven deployment strategy aimed at monitoring uncertain events by underwater sensor networks. In the initial deployment phase, random depth-adjustment method is adopted to make the nodes uniformly distributed, and thus more events can be captured. In the redeployment phase, according to the detected event information, nodes move based on virtual forces method. Meanwhile, through the use of clustering, motion range of each node is limited in its cluster. Thereby the size of reconstruction is reduced, and the connectivity of network is ensured. And a corresponding network reconfiguration mechanism is given for mobile events.(3) In the absence of geographical location information, redundant node scheduling algorithm based on embracing is designed. Firstly, based on the embracing approach, the discrimination rules of redundant nodes and nodes on a coverage-hole’s boundary are addressed. Secondly, the redundant node scheduling scheme is proposed, which includes node sleeping scheduling and coverage-hole patching. The scheme can maintain the initial coverage quality of networks, while effectively reduce the number of active nodes.Lastly, network reconfiguration mechanism based on the redundant node scheduling algorithm is proposed.(4) As it is difficult to satisfy the requirement of full coverage in sensor networks, we consider the problems of detecting and patching trap coverage holes, and propose a trap hole-detection method and a trap hole-patching method for them. The trap hole-detection method using the conception of effective arc exactly finds out the borderline of all coverage holes in a decentralized way. Therefore the network can get the information of these holes, and determine if they are trap holes. The trap hole-patching method is based on the conception of edge-weighted graph. The locations of new nodes for patching the trap holes are estimated by running the two algorithms for several rounds.(5) To handle the connectivity problem, this paper introduces an in-advance mechanism to prevent network partitioning in the initial deployment phase. The approach is implemented in a distributed manner, and every node only needs to know local information of its 1-hop neighbors, which makes the approach scalable to large networks. The goal of the proposed mechanism is twofold. First, critical nodes are locally detected by the critical node detection algorithm (CND) based on the concept of maximal simplicial complex, and backups are arranged to tolerate their failures. Second, under a greedy rule, topological holes within the maximal simplicial complex are patched step by step.