Research On Coverage Control In Wireless Multimedia Sensor Networks

The demand for completely monitoring the target environment with the multimedia information and the availability of low-cost and small-scale hardware such as imaging sensors, CMOS cameras, microphones, etc., all of which can ubiquitously capture multimedia content, have brought on the emergence and rapid development of Wireless Multimedia Sensor Networks (WMSNs). As a new and emerging type of sensor networks, WMSNs have received the immediate and significant attention of the research from community and industry. In sensor network applications, COVERAGE reflects the ability of monitoring the physical world, often as a description of the standards of Quality of Service (QoS). The purpose of COVERAGE CONTROL is optimazing network resource, in order to accomplish the environmental awareness and capture the information as better as possible, which is the foundation of keeping on the whole monitoring task. Compared with the coverage control in traditional sensor networks that has been studied and analyzed intensively, many existing methods are not suitable effectively for novel WMSNs, because of the characteristic with directional sensing ability. Thus, WMSNs demand a series of innovative solutions, especially for coverage control. This dissertation studies some fundamental issues in WMSNs, such as sensing model, deployment model and coverage control. Aiming at the problems mentioned above, we propose a series of new models and algorithms, and carry out the performance evaluation and simulation analysis. The main works and contributions of this dissertation are as follows.1) Based on the existing directional sensing models, we design a novel rotatable directional sensing model of WMSNs. Aiming at the boundary effect on account of the bounded deployment region, we modify the formula of coverage evaluation under random deployment in directional sensor networks, which is accurate and effective compared to the existing formula through theoretical analysis and simulation evaluation. Then based on the modified formula, we calculate the minimum number of sensors to achieve the desired coverage, and prove that the probability of full area coverage can represent the quality of coverage.2) Relying on the basic model of WMSNs, we design different methods of the coverage control to meet corresponding demands, which are detailed as follows.(1) Demanding for the quality of coverage as the most, we propse the Hybrid Intelligent Algorithm-based Coverage Optimization (HIACO). Currently, it has been proved that the optimal coverage problem is NP-complete, so according to the characteristic that the orientation of a node can be rotated continuously, we use the particle swarm optimization algorithm to optimize the coverage, combined with the simulated annealing algorithm. HIACO can search out the approximate optimal solution and the approximate optimal value from continuous optimization space. HIACO is divided into three phrases, namely, the network initialization, the optimal coverage calculation with hybrid intelligent algorithm and the sensing orientation adjustment, where the second phrase is the core of HIACO.(2) To solve that the existing algorithms of coverage control are only adapted to one type of scenarios with one immutable sensing model, we propose the Coverage Control Strategy (CCS) that can apply to kinds of scenarios with different sensing models. CCS deduces each change of sensing orientations from the partial derivative of expected coverage, which can form a scheme of orientations scheduling and a distributed algorithm. In CCS, each node can only adjust orientation according to the information of neighbors within its communication region. Based on two scenarios that the events at any point occur with the same probability used binary sensing model and that the events occur with the different probabilities used probabilistic sensing model, simulations show that distributed CCS is accurate and effective.(3) In order to lessen nodes with reducing a little of coverage, we propose two approximation algorithms based on Voronoi diagram, which both can redeploy network. The same goal of two algorithms is to cover maximal area while activating as few nodes as possible, called the Optimal Deployment Problem (ODP), and we prove that ODP is NP-complete. Because the maximal exposure path must lie on the line segments of Voronoi diagram, they are coverd to satisfy the worst case coverage. Aiming at that, we present Voronoi-based Centralized Approximation (VCA) algorithm and Voronoi-based Distributed Approximation (VDA) algorithm, which both cover the line segments of Voronoi diagram as more as possible. The experiments illustrate that two algorithms lessen different number of nodes with reducing the corresponding quality of coverage under the same scenario.