Year

2009

Degree Name

Doctor of Philosophy

Department

School of Electrical, Computer and Telecommunications Engineering - Faculty of Informatics

Abstract

Radio Frequency Identification (RFID) has revolutionized the asset tracking industry, with applications ranging from automated checkout to monitoring the medication intakes of elderlies. Its wide acceptance and applicability has spurred researchers to create novel RFID applications. One promising development is equipping nodes in a Wireless Sensor Network (WSN) with a RFID reader to create a distributed, self-configuring, ad hoc wireless network for tracking objects with an RFID tag.

A key problem in RFID-enhanced WSNs is the limited battery life of sensor nodes, which imposes a severe energy constraint on communication protocols. To put this into perspective, this thesis shows that the energy consumed by an RFID reader to read a single 96-bit tag is higher than a sensor node transmitting and receiving 96-bits of data. Moreover, in practice, an RFID reader has to read multiple tags in its interrogation zone, all of which may reply simultaneously. As a result, the RFID reader experiences collisions and unnecessary energy wastage that ultimately shortens a WSN's lifetime. For these reasons, it is imperative that a comprehensive study on the energy efficiency of existing RFID tag reading or anti-collision protocols be conducted in order to determine their suitability for use in RFID-enhanced WSNs.

This thesis, therefore, investigates the energy efficiency of Aloha based RFID anti-collision protocols. These protocols have low memory and bandwidth requirements, adaptive to changing tag population, and a small number of reader to tag commands; thereby, making them easy to implement on sensor nodes. Using analytical and simulation studies, this thesis shows that collisions and idle listening to be the key causes of energy consumption. Idle listening consumes a significant amount of energy, especially when the number of tags is low, but as the number of tags increases, collisions become the main cause of energy expenditure.

Another major finding is that existing anti-collision protocols are unable to monitor tags in an energy efficient manner. Specifically, in order to monitor tags, these protocols must undergo the collision resolution process repeatedly. This problem is particularly acute when tag population changes frequently. Hence, there is a clear need for energy efficient protocols that can determine new and old tags quickly. To this end, this thesis is the first to propose ResMon, an anticollision protocol that is designed to be energy efficient during identification and monitoring. Extensive simulation studies show ResMon's energy consumption to be significantly lower than state of the art framed Aloha variants; thus, making it ideal for use in RFID-enhanced WSNs.

02Whole.pdf (8252 kB)

Share

COinS