Kartinah Zen

Date of Award

2010

Document Type

Thesis - ECU Access Only

Publisher

Edith Cowan University

Degree Name

Doctor of Philosophy

School

School of Engineering

First Supervisor

Professor Daryoush Habibi

Second Supervisor

Dr Iftekhar Ahmad

Abstract

Research in wireless sensor networks (WSNs) has evolved with advancements in low power devices technology during the last decade. The driving applications in WSNs are habitat and environmental monitoring. However, sensors can also be applied in smart homes, health monitoring, vehicles tracking, military, and fire detection. Since batterypowered sensors are intended to be left for months or years without human intervention, computational algorithms and communication protocols for WSNs focus more on energy conservation rather than throughput and delay as in other wireless networks.

Recently, WSN research has grown from static to mobile sensor networks (MSNs). However, there are some weaknesses in the standard medium access control (MAC) protocol for WSNs when applied to mobile sensor networks. Mobile nodes in MSN consume more energy due to more overhead messages, experiencing longer disconnectivity periods each time the mobile nodes enter a new coverage range, having difficulty maintaining synchronisation with neighbours and spending more time in connectivity set up rather than sensing and monitoring.

The aim of this thesis is to overcome the limitations in the standard WSN MAC protocol in supporting mobile nodes so that reasonable throughput and energy conservation can be achieved. The enhancements in MAC protocol that we implement for MSN are achieved using four different mechanisms. The first mechanism aims to minimise the waiting period between the instance when a mobile node moves out of the coverage range of the assigned coordinator, until a new coordinator is assigned to it and communication is established with the new coordinator. We propose a technique for mobile nodes to estimate their positions based on available or predicted link quality indicator (LQI) values. Simulation results show that this mechanism successfully reduces energy consumption and shortens disconnectivity periods.

The second mechanism resolves the issue of choosing the coordinator which possibly may offer the longest connectivity period to a mobile node which has just moved out of the coverage range of the last assigned coordinator. In this design, different ranges of LQI value within a coordinator’s coverage area are exploited to estimate the distance between the coordinator and the mobile node and to also predict the mobile node’s future direction. This mechanism increases the connectivity period of mobile nodes. Energy consumption is also reduced because a mobile node, when losing connectivity, deals with fewer overhead messages by optimising the choice of the new coordinator based on an estimate of its movement direction.

The third mechanism proposed in this thesis shortens disconnectivity periods for the case of high speed nodes when beacon frame intervals are also long, by encouraging the mobile nodes to switch coordinator before these nodes lose their connectivity with their current coordinator. This mechanism applies an adaptive switching time for a mobile node, based on beacon frame interval and the speed of the mobile node. Simulation results show that this mechanism successfully overcomes the delay in the case of long beacon frame interval, and can cope with higher speed of mobile nodes compared to the standard MAC protocol.

The last proposed mechanism in this thesis aims to avoid early energy exhaustion of a coordinator due to excessive number of nodes attached to it. This mechanism introduces load balancing among coordinators in a ZigBee tree topology. Load balancing can distribute energy consumption more uniformly across the network and reduce delays in data transmission. Overall, the work presented in this thesis improve the support for mobility of sensor nodes and achieve longer connectivity periods, shorter disconnectivity intervals, lower energy consumption, better capability to deal with higher speeds, and improved throughput and better utilisation of network capacity compared to the standard WSN MAC protocol.

LCSH Subject Headings

Edith Cowan University. Faculty of Computing, Health and Science -- Dissertations

Mobile communication systems -- Design

Wireless sensor networks -- Access control

Dissertations

Download

Share

 
COinS