Doctor of Philosophy
School of Electrical, Computer and Telecommunications Engineering
Salarian, Hamidreza, Mitigating energy holes in wireless sensor actuator networks using a mobile sink and charger, Doctor of Philosophy thesis, School of Electrical, Computer and Telecommunications Engineering, University of Wollongong, 2013. http://ro.uow.edu.au/theses/4002
Wireless Sensor Networks (WSNs) are comprised of miniature, low-cost, low- powered, multi-functional sensor nodes with the ability to measure various parameters associated with their environment. In particular, they are able to monitor tremor, distance, direction, speed, load and pressure, temperature, humidity, light, vibration, motion and sound. Recently, researchers have extended the capability of WSNs to create Wireless Sensor Actuator Networks (WSANs), where nodes have the ability to manipulate their environment using actuators. Every node in both WSNs and WSANs has the ability to collect and process sensed data, and forward it to one or more sink/actuator nodes via its wireless transceiver in a multi-hop manner. Multi-hop communications, however, result in nodes that are close to a sink or an actuator to have a faster energy dissipation. This leads to non-uniform energy depletion, and the formation of energy holes around the sink/actuator.
This thesis studies and proposes solutions that mitigate the formation of energy holes. In particular, it explores the use of mobile sinks and proposes two innovative scheduling algorithms to generate the trajectory of a mobile sink. In the first method, a heuristic approach called Weighted Rendezvous Planning (WRP) is developed in which each sensor node is assigned a weight corresponding to its hop distance from the tour and the number of data packets that it forwards to the closest Rendezvous Point (RP). Experimental results indicate that WRP enables a mobile sink to retrieve all sensed data within a given deadline whilst conserving the energy consumption of sensor nodes. The results show that WRP can reduce energy consumption by 22% and increase network lifetime by 44% in comparison with existing algorithms.
The second method considers a mobile rover/robot with wireless recharging capability. The charging problem is formulated as an Integer Linear Program (ILP) with the objective of maximizing network lifetime. The obtained model is equivalent to the well known NP-hard, Capacitated Vehicle Routing Problem (CVRP). A heuristic method called Binary Search Wireless Charging (BSWC) is then developed in which the mobile charger preferentially visits sensor nodes with the shortest lifetime. BSWC uses the binary search algorithm to find the target lifetime that minimizes the residual energy of the rover's battery as well as using the shortest Hamiltonian path to reduce travelling cost. BSWC is validated mathematically and sufficient conditions are derived to ensure infinite network lifetime. Simulation results demonstrate that BSWC increases network lifetime by 400% as compared to Greedy-Plus, the current state-of-the-art algorithm.