Abstract
Deployments of Radio Frequency Identification (RFID) networks are anticipated to be dense and ad-hoc. These deployments usually involve redundant readers having overlapping interrogation zones and, hence, causing immense reader collisions. Elimination of redundant readers, from the network, is of utmost importance as otherwise they affect the lifetime and the operational capacity of the overall RFID network. In this paper, we propose a light-weight greedy algorithm that detects and eliminates redundant readers from the network. Our algorithm uses the ratio of tag counts to the number of neighboring readers of each reader to estimate the likelihood for that reader to be redundant. The proposed algorithm is highly scalable and poses a minimal communication overhead as compared with existing schemes in the literature.