Abstract:
Surveillance wireless sensor networks are deployed at border locations to detect unauthorized intrusions. For deterministic deployment of sensors, the quality of deployment can be determined sufficiently well by analysis in advance of deployment. However, when random deployment is required, determining the deployment quality becomes challenging. To assess the quality of sensor deployment, appropriate measures must be proposed. Determining the required number of sensors to be deployed initially is a critical decision. After deployment, temporal changes in the surveillance quality as the sensors die in time must be analyzed. The network lifetime definition must consider the surveillance performance of the network. In this thesis, to analyze the surveillance performance of the network, we propose deployment quality measures. We discuss the trade-off between the number of sensors and the deployment quality. We formulate the weakest breach path problem, and propose a method to determine the required number of sensors to be deployed. We propose the utilization of the watershed segmentation on the iso-sensing graph that reveals the equally sensed regions of the field of interest in a surveillance application. The watershed segmentation algorithm is applied on the iso-sensing graph to identify the possible breach paths. An algorithm is proposed to convert the watershed segmentation to an auxiliary graph which is then employed to determine the deployment quality. The surveillance quality is verified analytically. The temporal resilience of the surveillance quality is analyzed with a realistic discrete event simulator, and network lifetime definitions based on the deployment quality measures are proposed.
