Abstract
Wireless sensor network (WSN) technologies are increasingly employed in recent years for monitoring purposes in various fields ranging from the engineering industry to our immediate home environments due to their ability to intelligently monitor remote locations at low cost. While WSNs are continuously tasked to handle more complex functions, such as, data aggregation, frequent transmissions etc., in-network processing and intelligent computation still necessitate the available energy resources to be judiciously consumed in order to prolong their lifetime, since the nodes are often powered by batteries.
In this thesis, our original contribution to knowledge is by developing three key strategies that can efficiently optimize transmission protocols in relation to offer an improved network lifetime for WSN deployments. Our first strategy involves a design of a cluster-based network layer routing protocol. We propose a purely deterministic model that utilizes clustering to organize the WSN. Our design is referred to as DEC: a deterministic energy-efficient clustering protocol. It is dynamic, distributive, self-organizing and it offers better energy efficiency compared with the existing protocols in the same domain. In addition, it utilizes a simplified approach that minimizes computational overhead to self-organize the sensor network. Our simulation result shows a better performance with respect to the energy consumption, which is reflected in the network lifetime of up to 2X improvement in both homogeneous and heterogeneous settings when compared with the existing protocols. It is worth mentioning that our approach approximates an ideal solution to offer a balanced energy consumption in hierarchical wireless sensor networks.
Our second strategy to reduce energy consumption in WSN deployment involves designing a light-weight forecasting algorithm for prolonging the lifetime of sensor nodes. A detailed comparison with the state-of-the-art methodologies is conducted and they are evaluated using both a simulation study and a real-world deployment testbed implemented on a 32-bit mote class device. Overall, up to 96% transmission reduction is achieved using our Naive method, while still able to maintain a considerable level of accuracy at less than 2% error bound and it is comparable in performance to other complex models often used in the same domain.
Finally, our third strategy to further optimize energy-efficient transmission protocols in a real-world WSN deployment, involves combining the synergies between a data reduction method and a data acquisition method. Unfortunately, energy efficiency is often in contrast with network reactiveness for alarms and urgent messages. Thus, we propose SWIFTNET: a fast-reactive data acquisition protocol. SWIFTNET is built on the synergies between compressive sensing and prediction algorithms and it limits the energy consumption in environmental monitoring and surveillance networks. We show how this approach dramatically reduces the amount of communication required to monitor the sensor readings in a deployment. We use a wildfire monitoring application that requires high reactiveness with continuous data updates of the monitoring process. Experimental results show over 50% communication reduction can be further achieved without loss in accuracy over the common prediction approach. In addition, SWIFTNET is able to guarantee reactiveness by adjusting the sampling interval from 5 minutes up to 15 seconds in our application domain.
In general, extensive simulation study has been conducted to evaluate these new approaches together with real-world deployment testbeds using a commercial off-the-shelf (COTS) node platform that is currently available in the market. One good contribution of the overall design of our work is that it enables easy practical deployment and prototyping. To sum up, favorable results are obtained that support our design strategies.