Research Of Dynamic Coverage In Wireless Sensor Networks

With the rapid development of technologies such as energy, sensing and communication, applications based on Wireless Sensor Networks(WSNs) have been widely used. WSNs can expand people’s ability to interact with the physical world. One of applications is the battlefield surveillance. A bulk of sensors are organized to monitor the battlefield. Once events we focus emerge, the surveillance process is triggered by one or more sensors in the vicinity which detects these events and reports them to the sink node. The data collection for the area of interest is also applied by a WSN. A number of sensors are deployed and form a network. They collect the primary data at regular intervals and send them to the specified node. Actually, in a broad sense, the essential of most applications is the coverage issue.However, abundant existing works just assume that the area is static. For the dynamic area, the existing studies are very rare. Taking the environment surveillance system for example, in the contaminated area of diffusing poisonous materials, mobile sensors need transform positions due to the deformation of an area such that they can keep collecting the data or warning the border line of interest. In this thesis, we focus on the Dynamic Area Coverage(DAC) problem, which is to cover the dynamic area using mobile sensors. The main points are as follows:First, we formulate the DAC problem from different aspects. Then, the movement strategy should be designed to help each sensor find its destination such that the different movement distance requirements can be satisfied. In addition, we derive several subproblems from the DAC problem and prove two of them are NP-completeness problems.Second, we propose the Shrink Wrap Coverage(SWC) method for deploying sensors to get the full coverage of an area. Furthermore, we prove that this method can minimize the sum of overlay sensing area and useless sensing area in local coverage. After that, we summarize the features of movement patterns and propose three algorithms.Last, the exploration experiments with different parameters are presented in order to simulate the reality. We propose three kinds of proportions to evaluate the performance of the SWC method. For making the results general, each performance result comes from multiple tests. And we also compare with the performance of algorithms about the movement distance.