Research Of Area Coverage Control In Wireless Multmedia Sensor Networks

The rapid development of technology in communication embedded computing and sensor promotes the rapid growth of Wireless Sensor Networks (WSNs). With the environment to be monitored is becoming increasingly complex and the information is becoming gradually rich which make bring images, audio, video and other multimedia with richer amount information to traditional wireless sensor networks to be the new requirement. Thus, Wireless Multimedia Sensor Networks (WMSNs) comes into being. As a new area needs to be explored, there are many issues that need to be done in WMSNs. Among these, coverage control mechanism is the basis to ensure the others. Coverage of the networks reflects the perception ability of physical environment from WMSNs, and has a directly impact on the Quality of Service (QoS) of the whole network. According to the requirement, a reasonable coverage control mechanism needs to be designed to improve the efficiency of the network. At the same time, the high consume of energy is the key factor which effects the service time of the networks. Perception model is the basis of studying network coverage control mechanism, which directly determined the sensing scope of the nodes. Different from nodes in WSNs, the wireless multimedia sensor nodes have the characteristics of directionality and angle of view. Most of the existing researches focus on sector perception model, but to video sensors which have an important indicator on wireless multimedia sensor nodes, whether the sharpness of the received images could be accepted or not is very important. Sector model does not fully satisfy the practical applications. In this thesis, through a set of optical definitions, a modified sector ring perception model is proposed and the following study of coverage control mechanism in WMSNs is based on this model.Some research results have already been found in applying the Virtual Potential Field Algorithm (VPFA) into the coverage problems of WSNs. VPFA has a series of advantages like low computational complexity, simplicity and easy to be implemented. It also can be applied into coverage deployment in WMSNs. But before the nodes reaching the optimum positions, nodes may be force-balanced and become local minimum points, which makes the nodes could not reach the desired effects or even fail in optimization process of VPFA. In this thesis, based on the thoughts of VPFA, the common coverage of neighbor nodes is introduced to the function. In order to take the effective use of boundary nodes into consideration, a Mixed Virtual Potential Field Algorithm (MVPFA) is proposed. In the coverage control process, through computing common coverage among neighbor nodes to improve repulsive force function by adding in relative repulsive force factor. MVPFA can break local minimum balance and improve the efficiency of the tradition algorithm. After the first time of optimization, the network determines the boundary nodes to enhance the utilization of boundary nodes, try to maximize the coverage of the edge. Through simulation experiments and an example of application illustrate the effectiveness of MVPFA.In nodes scheduling mechanisms of WSNs, using clustering algorithms is relatively common. However, due to the features of WMSNs, a novel clustering formation process is proposed, the Multi-membership Clustering Algorithm (MCA) which applies for WMSNs. Forming clusters according to the common coverage between nodes. If the node could meet more than one cluster, it can become a member of multiple clusters, which could facilitate the collaboration among clusters and reduce redundant data by collection. Comparing with Single membership Clustering Algorithm (SCA), the MCA can effectively reduce the number of clusters in densely deployed WMSNs.