Intelligent RFID-Based Localization and Tracking Model Design Using Unscented Kalman Filter and Convolution Neural Network

The Radio Frequency Identification (RFID) is solitary of the widely used wireless technologies in the field of surveillance of non-stationary objects in the grassland of communication, privacy, security, internet of things, data management system, logistics, traffic control, etc. For precise localization and tracking of mobile objects, proper scheduling of RFID readers is needed, to fulfill this requirement, an intelligent RFID-based tracking system is designed based on the most popular and advanced algorithm. In this proposed work, the RFID readers and tags are used for the reception of RSSI (Received Signal Strength Indication) values and then identification, localization, and tracking of single object with the help of multi-input RFID antennas has experimented. The input to CNN (Convolution Neural Network) are samples that are generated from the RSSI values received from RFID readers and then exact identification, localization, and tracking of moving objects are done with the help of UKF (Unscented Kalman Filters). This research work has experimented on the basis of detailed literature review of different authors and experimental results show that this intelligent RFID tracking model is efficiently implemented in the field of indoor surveillance systems i.e. airports, railway stations, shopping malls, etc.