Research On Barrier Coverage Of Wireless Sensor Network

Wireless sensors network(WSN) is composed of a mass of micro sensors, it can be deployed in all sorts of complex environment, which is not easy to access for human beings, to perform the monitoring tasks. Wireless sensor network is deeply affecting the military, industrial and civilian areas, however, it also faces with many technical challenges. Coverage problem is one of the key problems in wireless sensor network, it reflects the physical monitoring service quality which the sensor network can provide. Good coverage properties can provide more reliable data source for sensor networks to prepare for the data analysis. In this context, aiming at deal with the barrier coverage problem in wireless sensor network, this thesis do some in-depth study on two direction: monitor and invasion. The main contribute of this thesis as following:Firstly, it introduces the related knowledge of coverage problem in wireless sensors network. The related concepts and techniques of barrier coverage is presented from the angle of monitoring and invasion, respectively.Secondly, the basic concept of Vicsek model is introduced in detail and analyzed its convergence from the angles of theory and experiment. An existing 1-barrier construct algorithm(Distributed Self-Deployment Algorithm for the Coverage of Mobile Wireless Sensor Network, DSDAC) which is based on Vicsek model is introduced, and the feasibility of Vicsek model applied in the barrier coverage is verified on the basis of simulation, then the performance of the algorithm is also analyzed, which provides the theoretical basis for proposing the K-barrier coverage construct algorithm following.Then, considering the traditional barrier coverage algorithms using static sensor are inefficiency and cause resource waste and also regardless of the connectivity of network, inspired by DSDAC, this thesis study a K- barrier coverage algorithm base on Vicsek model and motion sensors from the perspective of monitoring. The whole algorithm consists of three stages, the movement of sensors in each stage can be regarded as preparation for the next phase. Before each movement, sensors according to its local information to determine its motion parameters using the Vicsek model. And sensors will consider the connectivity with other sensors at the same time. At the end of the final stage, sensors network can form a K barrier in the area. The algorithm is simulated by MATLAB, and the verification results show that the feasibility and correctness of the algorithm.Finally, from the angle of the invasion this thesis study the optimal path selection algorithm based on the detection probability. Considering the actual situation which sensors do the detection behavior periodic, the algorithm discrete the path pass through the area and analysis the detection probability of the path. On the premise of guaranteeing the detection probability of the path satisfy constraints, in order to make the path shortest, the algorithm introduces a function which is about the length and detection probability of the path, and use the heuristic algorithm based on linear aggregation to solve the optimal path selection problem. The result of the MATLAB simulation experiment proves the feasibility of the algorithm and also shows the effects of various parameters on the performance of the algorithm.