Research On The Indoor Location Algorithm Based On RFID Technology

In recent years,with the rapid development of sensor network technology and wireless communication technology,Internet of Things(IoT)has become a new generation of network technology.However,many applications of IoT are based on the positioning target to provide more intelligent and personalized service and bring a lot of convenience to human life and production,so positioning technology has attracted much attention of researchers.Positioning technology includes outdoor positioning and indoor positioning.Since indoor environment is related to human life,so indoor positioning technology becomes the research focus of many researchers.Traditional indoor positioning technology such as ultrasonic technology,Bluetooth technology and infrared technology are expensive,complex or having unsatisfactory location accuracy.While RFID technology has a great advantage for indoor positioning because of its non-line-of-sight,non-contact,low-cost and high location accuracy,etc.It becomes an excellent candidate for indoor positioning.RFID technology has been recognized as one of the most promising technology in the 21 st century.In this thesis,some recent and well known RFID positoning algorithms in the literature are reviewed.Due to the severe interference in the complex indoor environment,the probability of selecting the wrong reference tag of the classical indoor positioning algorithm LANDMARC increases.In addition,LANDMARC needs to calculate the Euclidean distance between a tag to be located and a reference tag,which results in a high computation complexity.To solve these two shortcomings,an improved double-tag LANDMARC positioning algorithm,namely DLANDMARC,is proposed in this thesis.The proposed DLANDMARC defines double tags,i.e.an active tag and a passive tag,to locate the targeting tag positioning model.Since the perception distance of a passive tag is limited,it can only be deteted by the nearby targeting tag,so the whole positoning area can be divided into many small regions,the probability of selecting wrong reference tags is decreased and the computational time is thus reduced.Experiment results show that DLANDMARC has a better performance in terms of the localization accuracy,the computational time and the stability compared with other existing algorithms in the literature.At the same time,in order to visually show the performance of the proposed algorithm,a positioning simulator is developed using C++ language to test the proposed positioning algorithm.For RFID systems that some characteristics cannot be obtained such as RSSI or TOA for positioning purpose,an RFID positioning algorithm based on power scanning of tag is proposed.The algorithm establishes the relational model of tag transmission power and current power induction range based on electromagnetic waves in the free space path loss model.Assuming reference tags can automatically adjust the transmission power according to certain step and the targeting tag receives the signal of reference tag.Then the targeting tag records the transmission power,and converts it to the corresponding induction distance,to identify the position of a target.Experiment results show that,compared with other algorithms,smaller step of power change has good positioning accuracy.Therefore,decreasing the step of the power change can significantly improve the positioning accuracy.Finally,the proposed positioning algorithm is tested on an actual RFID systems.