A transmit power control protocol for multipath wireless sensor networks

Abstract

In the recent years, we observed a rapid growing of Wireless Sensor Networks (WSNs) for research and commercial uses. Low-cost/low-power WSNs are utilised in various applications such as smart-home, industrial control, health care, agricultural fields, environmental purposes, biomedical systems, and scientific applications. The aim of this thesis is to develop a novel transmit power control protocol for multi-path non-uniform density single-channel WSNs. The developed protocol has two main purposes: (1) to reduce energy depletion and prolong the battery lifetime of sensor nodes by using transmit power control and, (2) to keep throughput and packet loss neutral by using multi-path routing. A limitation of most previous studies that minimise transmit power is that they fail to take into consideration the throughput reduction. Through a number of case studies, it was determined that trying to reduce the power by using multi-hopping also results in the reduction of end-to-end throughput. Hence, we propose using a multi-path routing protocol to maintain throughput. In this work, given our assumptions, we determined that the optimal number of hops must be between two and eight hops to save energy and the optimal number of paths is two paths to maintain throughput when the transmission rate is high. This is mainly due to the overhead of each packet as we as the receive power of the sensor nodes. It was also determined that there is no need to have more than 2 paths between source and destination in order to achieve throughout neutrality. QualNet 5.1 platform was used to develop "TPC for High Density WSNs" protocol that combines both TPC and some features of the MultiPath Optimised Link State Routing (MP-OLSR) protocols. The simulation results showed that using two, three, four and five hops scenarios can noticeably enhance the energy efficiency, and the optimal number of non-interfered paths must be two paths to enhance throughput neutrality and reduce overhead messages of IEEE 802.11g sensor nodes in a dense network.ii ACKNOWLEDGEMENTS