دانلود رایگان مقاله الگوریتم پوشش RFID غیر متمرکز با استفاده از تگ قابل نوشتن

Abstract

A Radio Frequency IDentification (RFID) reader network is as a collaboration of RFID readers that aim to cover (i.e., identify, monitor, and track) every RFID tag in a given area. The RFID coverage (RFC) problem is defined as follows. Given a reader network, assign to each tag t a specific reader v in its proximity such that v is responsible for covering t (called its owner), while minimizing the number of owner readers. The problem has applications in energy conservation and in eliminating readers and data redundancy from the reader networks. We introduce a number of decentralized algorithms for the RFID coverage problem: 1) algorithms RANDOM, RANDOM+, and MAX-MIN which are randomized algorithms that run in O(1) write/read rounds, 2) algorithm GDE which is an efficient decentralized implementation of the greedy set cover algorithm, and 3) an improvement of GDE which is called . Our algorithms assume that the RFID tags are writeable, where a writeable tag is a passive RFID tag with writeable memory. We show using simulation experiments that our algorithms outperform major RFID coverage algorithms in various scenarios with respect to a number of performance metrics.

8. Summary and Conclusions

A summary of the algorithms introduced in this paper is given in Table 3. We introduced in this paper two sets of algorithms that solve the RFID coverage problem using writeable tags. The first set consists of randomized algorithms, called RANDOM and RANDOM+. It also consists of a variant of these algorithms, called MAX-MIN. The second set of algorithms consists of deterministic algorithms, called GDE and LIMITED-GDE. Our algorithms were shown to outperform major existing algorithms in different scenarios. There are still more variations of this problem (e.g., adding load balancing constraints, or others). A promising direction is fault tolerance. That is, how to solve the RFID coverage problem if it is assumed that the links between readers and tags may fail (e.g., a reader is not able to write in a tag in its proximity). Another direction is the problem of optimizing write/read rounds. Lastly, we see that the reader-tag RFID coverage problem still requires future research for two main reasons; the novelty of the reader-tag communication model, and the various and promising applications of the problem.