Passive UHF RFID Network Planning for Accurate 3-D Location via Restricted Genetic Algorithm

Abstract

Indoor localization via radio-frequency identification (RFID) carries critical importance due to its high accuracy and low hardware requirement. The positions of reader antennas can affect the positioning accuracy and coverage in RFID network. In this letter, we present a novel RFID network planning (RNP) approach to optimize the deployment of reader antennas for accurate 3-D location. First, the 3-D antenna radiation mode of a passive UHF RFID system is set up. Then, the received signal strength (RSS) and path loss characteristics (PLC) are analyzed and a restricted genetic algorithm (RGA) is developed to obtain the optimal solution of the RNP via maximizing the total reward function with constraints. Finally, the convolutional neural network (CNN) and weighted <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$K$ </tex-math></inline-formula> -nearest neighbor (WKNN) algorithms are respectively used to verify the localization effect. For comparison, the Cramér-Rao lower bound (CRLB) is also derived. The experimental results show that the proposed approach can improve the positioning accuracy as well as the coverage, and enhance the anti-noise capability of the location system.