Genetic algorithm based approach for RFID network planning

The design of a RFID network in a large-scale facility requires a placement of a large number of RFID readers to ensure the desired coverage. However, the placement of RFID readers is often done on a trial and error basis which is time consuming and results in less than optimal coverage. In this paper, a multi-objective function optimization model for the placement of RFID readers is proposed. Genetic algorithm (GA) is used to determine the optimal placement and number of RFID readers required in two simulations involving a 30m × 30m facility with 99 randomly placed RFID tags. The first simulation considered only 10 RFID readers and the results showed that 76 tags could be covered using the proposed model (compared to 72 tags covered in previous research). In the second simulation, the optimal number and location of RFID readers to cover all 99 tags in the facility was determined. The results showed that 21 RFID readers were necessary to cover all 99 tags, which is less than the 30 RFID readers derived from the concept of hexagonal packing.