Doctoral Defense

Mobile Data Collection and Energy Replenishment in Wireless Sensor Networks Theoretical and Experimental Approaches

Ji Li

May 8, 2017
11:30AM
Light Engineering room 250
Advisor: Prof. Yuanyuan Yang

Abstract: Energy constraint has been a major problem in traditional battery-powered wireless sensor networks which adopt static data gathering. Since sensory data is uploaded to data sinks through long-range multi-hop relay, quick depletion of battery energy is observed in sensors around data sinks, which results in the formation of energy holes and terminates the operation of the networks. Mobile data gathering provides a reasonable approach to alleviate this problem by employing mobile data collectors which work as data sinks roaming over the sensing field. Dynamically changing of locations mobile data collectors effectively balances the energy consumption in the network, and the virtually increased number of data sinks shortens the length of relays and decreases the amount of data being transmitted, thus energy can be significantly saved in sensors. In the meanwhile, energy replenishable technologies, such as solar harvesting and wireless recharge, have been considered as promising solutions to relieve the energy constraint exerted by the small-capacity batteries.

This dissertation focuses on the design of mobile data gathering and energy replenishment in WSNs, and experimental approaches for performance evaluation in such networks. In the first section, we focus on theoretical formulation and analysis of mobile data gathering, energy harvesting and wireless energy replenishment. We develop an efficient NDN-based protocol to collect real-time battery information from the network. We present a joint design of mobile data gathering and energy replenishment to maximize the network utility in wireless sensor networks. We propose a mobility assisted data gathering scheme with solar irradiance awareness which utilizes data sinks with limited mobility in solar-powered sensor networks to maximize the amount of sensory data that can be collected from the network. In the second section, we design a versatile testbed for mobile data gathering in wireless sensor networks, which includes both wireless sensor node and mobile data collector. We implement the functionalities for mobile data gathering for geese detection on the testbed, and conducted experiment to validate the networks performance. The experiment captures the impact of many real factors which are usually omitted in theoretical analysis and software simulation due to the limitation of the models.