A wireless sensor network (WSN) is a collection of sensors which are equipped with wireless radios and deployed in a given network domain (i.e., an environmental monitoring region). Recently, a number of efforts, both from industry and academia have been made for the purpose of effectively deploying WSN for a variety of applications (e.g., forest fire, chemical contamination of water or soil). The power constraints of sensor devices pose
many fundamental design issues in WSNs, such as coverage, connectivity and lifetime. For example, design of an environmental monitoring WSN
needs to target maintaining a satisfactory sensing coverage and transmit the required sensor data (e.g., temperature) to the base station (BS) or sink node, with period ranging from months to years. However, multihop transmission makes the sensors close to the BS consume more energy than other far away nodes because of packets to be relayed to the BS. Once closeby nodes use up their energy, data from far away nodes can not be transmitted to the BS directly and the network performance is heavily affected.
In this dissertation, we study different deployment strategies for WSNs. We introduce the use of some redundant nodes so that the lifetime of WSNs could be prolonged. Nonuniform deployment of WSNs is also discussed. The main idea is to deploy more nodes close to the BS so as to balance the energy consumption among all sensors. Three kinds of deployment strategies (deterministic deployment, partially controlled deployment and distribution controlled deployment) are studied. We also provide a data collection protocol design to support the nonuniform deployment architecture.
We also explore the future research directions related with WSN deployment strategies.