Divide and Conquer: Detecting and Tracking Passive RFID Tags in Retail Spaces

Detecting and tracking items marked with passive RFID tags are essential in modern retail stores, in which the RFID tags are used for inventory control and checkout-free shopping. In many retail stores, passive RFID tags replace traditional bar codes for payment processing and inventory control. Since the typical operational distance of RFID readers for most practical cases is very short (about several feet), to cover the entire store, there is a need for several RFID readers for large retail stores. The issue addressed in this paper is the optimal placement of RFID readers, which provides full coverage of a given entire space, subject to cost, delay, and load balance constraints, using a minimal number of RFID readers. Using the "divide and conquer" strategy, several coverage schemes are introduced and analyzed.

I. Introduction

The usage of Radio Frequency Identification (RFID) systems in retail stores is rapidly increasing. Typically, an RFID system consists of two types of devices: RFID tags which are used to mark the items in the store, and one or several RFID readers which are used to track and identify the RFID tags. Each RFID tag has a unique identification number (ID). This ID identifies the physical object to which the RFID tag is attached. An RFID tag can be either active (i.e., has a battery), or passive (i.e., battery-less). Due to cost-saving reasoning, passive RFID tags are preferred for marking items in retail areas, for obvious reasons. The RFID reader can identify and track the location of the RFID tag. In practice, the effective range of identifying passive RFID tags is typically limited to several feet. Hence, to have full coverage of a large store there is a need for several RFID readers. The issue addressed in this study is how to use a minimal number of RFID readers to provide full coverage to all the RFID tags located in a given space, subject to delay and load balance constraints.