Reliable and Energy-Efficient Data Collection in Wireless Sensor Networks

Feng Wang, Assistant Professor, University of Mississippi

Tuesday, June 4th, 2013 at 10:15.

Polacksbacken, room 1145

Wireless sensor networks (WSNs) have attracted significant attentions in the past decade, and the advances in computer and communication systems have made multi-hop data forwarding possible among distributed and collaborative sensor nodes. Data collection from these nodes to a central data sink for further processing is no doubt one of the most important WSN services. Unfortunately, today´s wireless sensor nodes are still severely constrained by their battery reservoir, computation power, and memory space. Reliable and energy-efficient data collection thus remains a daunting task. It is further complicated given the inherent unreliability of wireless channels, particularly with increased collision in fields of densely deployed sensor nodes.

In this talk, I will present our recent studies that systematically examine a series of issues across all stages of data collection, namely, node deployment, message dissemination, and sensor data delivery. I will introduce an optimal traffic-aware relay node placement scheme for multi-hop data forwarding. I will then identify potential failures of existing broadcast services for message dissemination in energy-efficient low-duty-cycle WSNs, and present effective solutions for remedy. For data delivery, we focus on an important but challenging application scenario: high-rise structure monitoring. I will show a number of unique challenges and also potentials therein, together with an elevator-assisted solution. It will be further elaborated through case studies from our prototype deployment on the Guangzhou New TV Tower, one of world´s tallest buildings.