Ph.D. Preliminary Exam - Bin Tong

Date: 16 Feb, 2009
Time: 3:00 PM
Location: 217 Atanasoff Hall
Topic: Mobility-assisted energy management in sensor networks
Major Professor(s): Wensheng Zhang


Abstract:

Over the past decade, sensor networks have consistently been a focus of computer research community, and a large number of prototypes have been built for military and civilian applications. One of the fundamental research problems of sensor networks is the energy scarcity problem. Due to the nature of sensor networks, e.g., small-factor sensor node, large-scale deployment, and etc., sensor nodes cannot carry a large amount of energy or be conveniently recharged. The energy scarcity problem is defined as how to support sensor networks to work towards an expected period of time given scarce energy supply in sensor nodes. A large number of schemes have been proposed to address this problem. These schemes, however, have one or more the following drawbacks: i) energy cannot be replenished to the network, and thus the network lifetime is bounded by the amount of energy preloaded to sensor nodes; ii) the ways of replenishing energy to sensor nodes are not practical and reliable, and thus cannot support normal operations of the network; and iii) sensor nodes drained of energy are left in the sensing field, and thus cause pollution to the environment.

Mobility has been widely employed to solve a variety of research problems in sensor networks. In our preliminary work, we propose three novel mobility-assisted energy management approaches to solve the energy scarcity
problem: i) Overlay-based network: We build an overlay network composed of a small number of mobile resource-rich nodes to facilitate data forwarding in sensor networks. ii) Node reclamation and replacement: We dispatch a mobile robot or human technician periodically to traverse the network, reclaim nodes with low or no energy, replace them with fully-charged ones, and bring the reclaimed ones back to an energy station for recharging. iii) Wireless
recharging: We take advantage of the cutting-edge wireless charging technology, and employ a mobile recharger to periodically traverse the network and recharge nodes with low or no energy over radio. We have identified the fundamental challenges in realizing these approaches and proposed solutions to address them.

Extensive analysis and simulations have been conducted to verify the effectiveness and efficiency of our solutions. As the battery technology legs far behind MEMS, we believe energy replenishment is inevitable to long-lived surveillance sensor networks. In our future work, we plan to study practical issues in realizing our proposed approaches.