| Wireless Sensor Networks (WSNs) have emerged as a new information-gathering paradigm based on the collaborative efforts of large number of sensors. For real-time WSN applications such as monitoring where decision, control and update processes are based on the received data, reliable packet delivery is an important issue. An elegant reliability solution should benefit by constructing an energy-efficient topology, in order to be effective within this resource-constraint networking domain. In this thesis, the problem of energy-efficient reliable data transport is addressed targeting wide range of WSN domains that matches the unique characteristics of sensor networks. The proposed protocols fit into different class of wireless sensor networks supporting both centralized and distributed solutions. We first present the design of an asymmetric and reliable transport (ART) mechanism, and evaluated the scheme by simulating and implementing it using realistic scenarios based on a reference home WSN application. Next, an energy efficient two-tier self-scheduling (TTS) paradigm is proposed. Specifically, TTS enables sensor nodes to construct a scalable topology under stringent energy, coverage and reliability constraints. Sensor nodes aim to preserve sensing coverage, while scheduling themselves into sleep in phases for energy conservation. By incorporating ART reliability mechanism, we show that high precision event detection at the collector node can be achieved with guaranteed event and query delivery performance. TTS has been studied to fit both centralized and distributed WSN domains. In distributed version, sensors are self-organized targeting to generate the scalable topology by self-discovery and self-calculation of their sensing coverage. In addition, self-organization is extended for wireless multimedia sensor networks having directional sensing views. A distributed scheme is designed and simulated for multimedia sensor nodes to compute their directional coverage, through which orientations are calculated for efficient self-organization. Finally, a case study of questioning and improving the reliability of home WSNs is presented and performance of ART in home wireless sensor networks is investigated. The results are promising and provide a basis for future investigations of home WSN applications that requires reliable communication services. |
