Semester Thesis

Multiple Access Schemes for Tag Estimation and Identification in Dense RFID Networks

bit allocation

In radio frequency identification (RFID) systems, the proliferation of readers and tags introduce a number of technical problems. While traditional RFID applications deal with a single tag identification, one of the main challenges is the identification of multiple tags. If the number of tags is relatively large and two or more tags occupy the same RF channel simultaneously, we have collisions between tags, and the reader can not identify the data of the tags. The identification efficiency in RFID systems generally decreases for increasing the tag collisions. Advanced readers have the capability to cope with this problem by providing anti-collision techniques. So far, the most widely proposed schemes are variants of ALOHA and tree algorithms [1], where framed slotted ALOHA (FSA) is used in several RFID protocols [2,3] due to its simple implementation.

There are different variants of anti-collision techniques based on FSA [4], but most of them assume known the number of tags a priori, which is not practical. Since the number of tags present in the read range of a reader is usually not known, the performance of a transmission control algorithm depends on the estimation of the number of tags n, and the choice of the corresponding frame size N. In [5], a lower bound on the number of tags is proposed based on the fact that in every collision at least two tags are involved, and each tag chooses only one slot for sending its message.

In scenarios where n >> N, the probability of tag collision increases rapidly, and since the frame size N can not be increased indefinitely, we have to somehow restrict the number of responding tags. In [4], an enhanced DFSA (EDFSA) method is presented, where the problem has been solved by restricting the number of tag responses. The idea behind is to create groups of tags, and to allow only one group to respond at each read cycle. Although, the proposed scheme outperforms the conventional algorithms, it does not propose an optimum grouping strategy. Furthermore, the number of tags is assumed to be known in advance which is a non practical assumption.

In this project, we investigate a novel transmission control algorithm for RFID communication protocols based on FSA. The considered scheme combine the lower bound tag estimation with a new grouping strategy, and use a collision ratio to control the algorithm. The goal of proposed approach is to maximize the system efficiency while keeping moderate the total time for reading all tags, assuming that the number of tags is not known in advance.

[1] K. Finkenzeller, RFID Handbook, John Wiley and Sons, 2003.
[2] ISO 18000 Part 6, Information technology automatic identification and data capture - radio frequency identification for item management air interface, 2003.
[3] EPCglobal, Class 1 Generation 2 UHF Air interface Protocol Standard V1.0.9, 2005.
[4] Su-Ryun Lee and Chae-Woo Lee, An Enhanced Dynamic Framed Slotted ALOHA Anti-collision Algorithm, EUC Workshops, pp. 403-412, 2006.
[5] Harald Vogt, Multiple Object Identification with Passive RFID Tags, IEEE International Conference on Systems, Man and Cybernetics, 2002.

Type Semester Thesis
Subject area Multiple Access Schemes, RFID Networks
Type of work 60% Theory, 40% Software (Matlab)
Student Yinan Zhang
Supervisor Dr. Carlo Mutti
Professor Prof. Dr. Armin Wittneben