Reliable middleware framework for RFID system

The reliability of RFID systems depends on a number of factors including: RF interference, deployment environment, configuration of the readers, and placement of readers and tags. While RFID technology is improving rapidly, a reliable deployment of this technology is still a significant challenge impeding wide-spread adoption. This research investigates system software solutions for achieving a highly reliable deployment that mitigates inherent unreliability in RFID technology.;We have considered two different problem domains for large scale RFID deployment. One is item tracking and the other is guidance-monitoring.;The basic contribution of our work is providing novel middleware solution that is able to serve the application taking into account the inherent unreliability of RFID technology. Our path abstraction that uses the physical flow of data as an ally to generate a logical system level flow enhances the performance in many ways. The contributions of this dissertation are summarized below:;Defining novel system architecture for item tracking applications: We have defined a system architecture referred to as Reliable Framework for RFID (RF2ID) that takes into account the unreliability of RFID devices and provides a scalable, reliable system architecture for item tracking applications. It uses a distributed system abstraction named Virtual Reader (VR) that handles RFID data in different geographic locations. Virtual Path (VPath) is the abstraction that creates channels among the VRs and facilitates a data flow oriented data management in the system.;Implementation of RF2ID: We have implemented RF2ID that is able to incorporate physical RFID devices as well as emulated devices for scalability study taking into account various real world challenges of large scale RFID deployment.;Load Shedding Based Resource Management: RF2ID requires a mechanism to handle unexpected system load in the presence of asynchronous arrival of data items. Space based load shedding and time based load shedding techniques are used in RF2ID. The basic idea is to exploit the VR and Vpath abstraction to intelligently share the load among the VRs in the presence of high system load, and yet provide some guaranteed Quality of Service (QoS).;Architecture for GuardianAngel: We define an architecture for an indoor pervasive environment which provides novel system abstraction and communication framework. The layered architecture has distributed computational elements known as the virtual station (VS) that are in charge of serving different regions of the environment. The Mobile Objects (MO) are the physical and logical entities that use sensing device and traverse the environment. The environment itself is tagged with RFID. The MO uses its sensing device to make guidance decisions locally. The VS keeps status information of MOs and keeps coarse grained information of the MO over time and space providing a virtual location for each MO.;Implementation of GuardianAngel: We have implemented the GuardianAngle system as defined by the architecture. We have used a testbed that uses real RFID readers and tags in the pervasive environment in a limited laboratory setup. We have also developed a distributed system setup using emulated tags for a scalability study of the proposed architecture. We have also implemented a prototype application, to test its feasibility in the real world.;Evaluation of the system: We have conducted extensive evaluation using the real RFID testbed as well as scalability study using emulated readers and tags. The evaluation using the real RFID tags and readers gives us the credibility of the system under various environmental considerations. The large scale experimentations provide us with scalability and feasibility study to strengthen our limited resource study using real RFID testbed. (Abstract shortened by UMI.)...