AbstractsComputer Science

Lifetime and latency aware data collection in wireless sensor networks

by Jingjing Fei




Institution: University of New South Wales
Department: Computer Science & Engineering
Year: 2014
Record ID: 1048779
Full text PDF: http://handle.unsw.edu.au/1959.4/53362


Abstract

A Wireless Sensor Network (WSN) consists of a set of sensor nodes deployed in the environment where we intend to collect physical information such as temperatures. All the senor nodes are connected wirelessly, and work cooperatively to fulfill some specified tasks. Sensor nodes are typically battery powered. As a result, the network lifetime becomes a major optimization objective in the design of a WSN. Another important optimisation objective is to minimize the maximum latency of data collection for time-critical applications. In this thesis, we study the problem of lifetime and latency aware data collection in a static WSN with only one base station. We propose two novel routing structures, namely, k-tree and k-DAG, to balance the loads of the neighbouring sensor nodes of the base station to prolong the lifetime of the network while providing the maximum latency guarantee. Firstly, we investigate the lifetime aware data collection problem by using ktree. A k-tree is a spanning tree with the base station as the root such that the path from each sensor node to the base station is at most k hops longer than the shortest path from this sensor node to the base station. We propose a distributed algorithm for constructing a k-tree such that the loads of the base station���s children are balanced. Secondly, we study the lifetime aware data collection problem by using k-DAG. A k-DAG is a spanning Directed Acyclic Graph (DAG) with the base station as the only source node such that the path length of any path from each sensor node to the base station is not k hops longer than its shortest path length to the base station. We present a distributed algorithm for constructing a k-DAG such that the loads of the base station���s children are balanced. In addition, we propose an efficient distributed naming scheme to assign a unique ID to each sensor node for efficient point-to-point communication. We have implemented all of our algorithms by Cooja simulator. The simulation results show that our approaches significantly increase the network lifetime by up to 82%.