Wireless Sensor Networks (WSNs) consist of a large number of very small networked nodes that are widely distributed. The nodes incorporate communication, processing and sensing capability, and are resource-constrained due to limitations in size. The most scarce resource is the available energy budget and thus the energy consumption of WSNs is of paramout importance. Having an extremely large application space, WSNs fundamentally enhance our capability to control and monitor the surrounding physical world. Distributed Source Coding (DSC) is a solution to realise energy-efficient WSNs. This compression scheme aims at reducing the amount of sensed data to be communicated and the number of required data transmissions to achieve energy savings. In this thesis we develop and evaluate a practical and adaptive DSC scheme applied in WSNs under realistic technical requirements and limits. The scheme is adaptive to environmental variations and can exploit arbitrary spatial correlations in the sensed phenomena. We implement this DSC scheme on real hardware platforms and demonstrate its feasibility through an experimental testbed. In addition, we analyse realistic network deployment models and derive enhanced deployment models that directly lead to optimised deployment strategies. Furthermore, we derive and evaluate a performance prediction metric that is applicable to rapid and lightweight evaluation of WSNs. Following a realistic modelling methodology we take into account the energy consumption of the DSC-related signal processing algorithms and practical deployment strategies. We develop the comprehensive and detailed evaluation framework to exactly and quantitatively determine achievable gains. The framework includes an energy consumption model extracted based on accurate measurement data gathered from our experimental testbed. Using the framework, we evaluate the practical and adaptive DSC scheme as well as topological effects on the performance of realisticly deployed WSNs. The applied scheme shows substantially improved energy savings and in addition strong operational lifetime extensions of WSNs under realistic conditions. Overall the energy-efficiency of WSNs is significantly improved through efficient communication and optimised network deployment as proposed in this work.
展开▼
