Research On Node Localization Algorithm On RFID Sensor Network For Environment Monitoring

In recent years, with the rapid development of interdisciplinary and multiple technologies in parallel, application of technology integration is used to more fields. Driven by the internet of things for the existing localization services, the integration of Radio Frequency Identification (RFID) and wireless sensor networks (WSN) will become the future trend of localization. The RFID sensor network, which is realized by the technology integration, has wide application prospects in smart home, intelligent transport, warehouse logistics, medical care and other environmental monitoring fields. On the basis of the RFID and the WSN, the paper gives an environmental monitoring-oriented RFID sensor network model according to advantages of the RFID and the WSN. The model meets requirements of strong anti-interference ability, low power consumption, low cost and strong information security. On the basis of the model, according to the characteristics of mobile nodes, two localization algorithms based on RFID sensor network are given.Aiming at lack of the existing environmental monitoring in real time and supervision, an environmental monitoring-oriented RFID sensor network model is given, in which a network is created by RFID reader nodes as ZigBee coordinator. In the process, an adaptive network communication structure is given. The structure introduces RF tags into sensor nodes so that RFID sensor tag nodes can get electromagnetic energy and identify specific targets. Therefore, the model realizes intelligent environmental monitoring, and achieves global monitoring to specific targets and distributed environmental parameters monitoring, and meets requirements of low power environmental monitoring.Aiming at the real-time requirements of monitoring system for RFID reader node, a mixed localization algorithm based on RF fingerprint is given. In order to reach a high location accuracy level for RFID reader node, the algorithm combines estimation of posterior probability of Bayesian filtering for RFID reader node with the nearest neighbor centroid algorithm. The simulation results show that the localization algorithm does not require additional hardware support, it uses owner characteristic value of RF signal to achieve high localization accuracy, and it is especially suitable for low power environmental monitoring needs.Aiming at the lack of the larger sampling region and the prior information that is utilized insufficiently in Monte Carlo localization algorithm, a localization algorithm based on genetic Monte Carlo localization algorithm is given. The algorithm describes posteriori probability distribution of the RFID reader node by frame of Bayesian filtering, and introduces genetic crossover and mutation operation into the Monte Carlo process in order to diversify and optimize sampling. The simulation results show that the algorithm has better localization accuracy and higher robustness, and restricts the noise of NLOS to a certain extent, and reduces the sampling number of localization and storagespace as well as computing resource.