The research develops an approach to preserving energy in Wireless Sensor Networks (WSNs). Energy is a constrained resource, and thus increasing the lifetime of nodes to extend the flexibility of network deployments and ease maintenance is a core challenge; the network can operate unattended, autonomously and deliver applications for longer periods of time without human intervention. Energy is an especially limited resource for WSNs since invariably, nodes are battery powered and any scheme that extends the viability of the limited energy resource is much sought after. Inherent to the principles of WSNs is that each node is designed through a restricted set of resources and is equipped with the ability to gather, store, process and communicate data. In comparison to processing, transmitting/receiving data is very costly in terms of power consumption. Thus a viable strategy to conserve energy is to limit the amount energy owing to receiving and transmitting data through embedding intelligence within the protocol stack. The dynamic adjustment of the transmission power according to an application requested success rate, implemented through an extension of data embedded in the message packets is proposed and evaluated. Although a reduction in the transmission success rate results through lost packets, the level of energy savings is not compromised by an excessive increase in packet loss owing to a non-linear trade-off between the level of transmission power and transmission success. The dissertation introduces the motivation and background to the solution, defines the mathematical framework with which the approach to energy saving is founded and presents the emulation environment in which the performance of the solution has been evaluated. The expected energy savings owing to the utilisation of the scheme is presented and compared with techniques reported in the literature. Results show that notable energy saving is achievable with the proposed scheme.
|Date of Award||10 Sep 2015|
- University Of Strathclyde
|Supervisor||Ivan Andonovic (Supervisor) & Craig Michie (Supervisor)|