Application reliability framework in wireless sensor networks

Wireless sensor networks (WSNs) came into reality after the advancements in miniaturization and low-cost low-power circuit design technologies. Now-a-days, sensor networks are being widely used to support a large variety of applications. As sensor nodes are generally power constrained, most of the research works put great emphasis on energy conservation of sensor nodes. The primary objective of this dissertation is to ensure data reliability of the networks, without which an extended network lifetime brings little benefit. The required level of data reliability depends on the criticality of the applications. However, data reliability in WSNs can be characterised in terms of node-level reliability and network-level reliability. Extended node lifetime ensures node-level data collection reliability whereas successful data transportation to the sink ensures network-level application reliability. In order to contribute to node-level data collection reliability, this study has developed an energy efficient MAC protocol that conserves node energy through intelligent sleep and wake-up scheduling of nodes. The proposed MAC protocol can be employed in both data gathering and event sensing networks. To ensure reliable transportation of sensed data from collection sites to the sink in data gathering networks, a routing protocol has been designed that reduces congestion during data transfer in the networks. Another congestion-and-delay aware routing protocol has been designed to support timely detection of events in event sensing networks. Finally, a complete architecture has been proposed that supports both periodic environmental data collection and time critical event detection within a single network. The architecture eliminates the necessity of deploying two separate networks for performing two different tasks. Extensive analysis and experimentation have proved the effectiveness of the proposed schemes over the existing works in literature.