RFID Gen-2 System Capacity
FAYETTEVILLE, Ark. — Engineering researchers at the University of Arkansas have developed a novel mathematical model that describes how radio-frequency-identification (RFID) readers capture tag data on a single inquiry. The researchers — Kazem Sohraby, professor of electrical engineering, and Chonggang Wang, post-doctoral fellow in the department of electrical engineering — also developed two critical performance metrics to measure capacity of a single RFID reader environment.
Radio frequency identification, also referred to as RFID, is a wireless technology that uses radio communication to identify objects with a unique electrical identity. The system includes a reader or multiple readers that emit a radio inquiry, a tag that receives inquiry from the reader and issues a response, and middleware that filters, cleans and organizes raw data. Like bar codes, RFID is used to identify items. Unlike bar codes, which must be brought close to the scanner for reading, RFID tags are read when they are within the proximity of a transmitted radio signal. The technology has applications for business and industry, including retailers, suppliers and transportation providers.
EPCglobal Inc., the organization that sets international RFID standards, recently ratified a new technical protocol, referred to as Gen-2, for RFID passive tags. The new protocol provides advanced features to improve performance and security. Major corporations, including Wal-Mart Stores Inc., have converted to the Gen-2 protocol.
The researchers’ mathematical model -- known as Discrete-Time Markov Chain -- helped them discover that the algorithms used to facilitate communication between the individual reader and tags can achieve a reading rate that is very close to the highest possible rate of successful tag identification for a single inquiry. Wang developed two critical performance metrics -- Successful Tag Identification Rate and Tag Identification Speed -- to measure RFID system capacity for a single inquiry. Using the model, researchers found a maximum possible successful tag identification rate of about 36 percent for each query command from a single reader. Based on the algorithm, the researchers achieved an actual successful tag identification rate of 34 percent.
Sohraby emphasized that their study explains how a reader captures tags, and the findings apply only to the speed of a single reader issuing one inquiry. They also give theoretical support for designing advanced protocols to improve RFID system capacity in environments that are prone to radio-signal interference.
Chonggang Wang |
For example, across campus, researchers in the University of Arkansas RFID Research Center, an EPCglobal Inc.-accredited research laboratory in the Sam M. Walton College of Business, produce successful identification rates that are rarely less than 100 percent.
“Readers start firing inquiries as soon as a case or pallet enters a read field,” said Bill Hardgrave, director of the RFID Research Center. “In a matter of seconds, you will have thousands of inquiries. The likelihood that a tag will be identified goes up each time a reader sends an inquiry, so chances that a tag will be 'counted’ are very good. Dr. Sohraby’s theoretical work on inquiry performance complements and reinforces the results of numerous applied tests we have conducted that demonstrate the superior performance of Gen-2 over Gen-1 and the ability to achieve 100-percent read rates on almost any product.”
Sohraby described the Discrete-Time Markov Chain model as a breakthrough in RFID quantitative analysis and predicted that the two performance metrics will become industry standards for measuring RFID system capacity and reliability.
“In the past, these metrics have been talked about in qualitative terms,” he said. “But they needed to be tested, and until now we didn’t have quantitative tools to do that. With the new model, we now have a basis for comparison.”
Sohraby and Wang introduced a novel framework to improve RFID performance of a single reader inquiry through a process called “reading-error prevention” and “reading-error correction.” These processes include methods to protect reader-tag communication through isolation, and tag/reader redundancy.
Wang presented these findings at the AT&T Research 2006 University Collaboration Symposium in New Jersey during the summer. The researchers will submit their study, which was funded by AT&T Research Laboratories, for publication at conferences and journals published by the Institute of Electrical and Electronics Engineers, Inc.
Contacts
Kazem Sohraby,
professor, electrical engineering
Department
of Electrical Engineering, College of Engineering
(479) 575-2635, sohraby@uark.edu
Chonggang
Wang, post-doctoral
fellow, electrical engineering
Department
of Electrical Engineering, College of Engineering
(479) 575-8655, cgwang@uark.edu
Matt
McGowan,
science and research communications officer
University
Relations
(479) 575-4246, dmcgowa@uark.edu
Headlines
Fay Jones School of Architecture and Design Presents Fall 2024 Lecture Series Lineup
Through these carefully selected presenters, the Fay Jones School continues to engage with the broad scope of issues, opportunities and challenges that society and the design disciplines confront today.
Honors College to Host Pulse Discussion on the History of Campus Protests
An expert panel moderated by the Democracy Fellows RSO will discuss the history, philosophy, emergency management and legality of campus protests on Sept. 12.
Baxter Joins Animal Science As Equine Instructor, Ranch Horse & Equine Judging Coach
Amanda Baxter, who has extensive experience in 4-H horse contests and clinics, is joining the Department of Animal Science as an equine instructor and head coach of the ranch horse and equine judging teams.
Wang Appointed Associate Editor for Biophysics Reviews
Yong Wang, associate professor of physics at the University of Arkansas, has been named an associate editor of Biophysics Reviews, the journal of the American Institute of Physics.
Data Science Program to Be Led by Previous Engineering Dean While Director Search Begins
Former Dean John English will serve as director while a search for the successor to University Professor Manuel Rosetti, the Data Science Program's inaugural director, is conducted.