Center for Embedded Network Sensing

Efficient deployment of a distributed wireless systems is a fundamental and difficult task. When considering the problem from the perspective of a sensing system, the challenge of efficient deployment becomes two-fold. First, the main purpose of the deployment is the gathering and quality of the spatial phenomena collected from the environment. Secondly, the selected positions should communicate efficiently.Due to the presence of unknowns in the deployment environment and lossy communication links, prediction of efficient deployment positions for communication is difficult. We propose the use of deployed sensor nodes for analysis of the communication qualities in the deployment environment. Using a small collected set of communication quality measurements, we propose an optimization approach for determining the deployment positions of additional nodes into the network in order to improve the communication quality and therefore the lifetime of the network. In order to solve the optimization problem, we map the communication qualities into an equivalent corresponding instance of non-linear programming in a new communication and lifetime space. We analyze the new deployment approach using both actually deployed networks, as well as extensive simulation studies.