Routing, deployment and in-network aggregation strategies in wireless sensor networks.

摘要

A wireless sensor network consists of a set of mostly battery-powered sensors and is used for environmental monitoring, event tracking, security, and system control. Each sensor network can be treated as a distributed database system that provides various types of query services to end users. One important factor, the constrained energy resource of sensors, limits the lifetime of a sensor network. For a sensor network with uniformly distributed homogenous sensors and a stationary sink, one observation is that the energy of the sensors near the sink is depleted much faster than that of the sensors that are distant from the sink. When the neighbor sensors of the sink exhaust their energy, the sink is disconnected from the network, and the lifetime of the sensor network is over, even though there is still a large amount of unused energy in the remaining sensors. Hence, how to efficiently utilize the unused energy to prolong the lifetime of the energy constrained sensor networks presents design challenges of all the levels of the sensor network protocol stack.; In this dissertation, three important factors: routing techniques, sensor deployment, and in-network data aggregation, are taken into account to extend the lifetime of sensor networks. The designs related to these three factors are not independent of each other. By considering these factors, four sensor deployment strategies and their related routing protocols are proposed to support various types of queries. In the first strategy, mobile sinks are used in sensor networks with uniform sensor distribution. The second and third strategies employ non-uniform energy distribution models to extend the lifetime of sensor networks. Based on the non-uniform energy distribution models, their associated routing protocols are also presented. Finally, for sensor networks with uniformly distributed sensors and a stationary sink, a broadcasting-based routing scheme is suggested to slow down the energy depletion rate of the sensors near the sink and to efficiently support aggregative queries. Compared with existing approaches, the four strategies have a significant performance gain in terms of the total number of queries handled by the network.