The Design Of Passive Rfid Location System Based On Nearest Neighbor Tags

There are two kinds of location systems: outdoor location system and indoor location system. The aim of indoor location is to locate objects in a room environment. Compared to the outdoor environment, indoor environment is generally more complex due to all kinds of factor, such as the limitation of space, the articles in the room and so on. Many existing location systems based on infrared, ultrasonic wave or bluetooth are not suitable for indoor location because the location accuracy cannot be achieved and the cost is high. Because of its non-touch, no light of sight, high accuracy and low cost advantage, the Radio Frequency Identification (RFID) technology has won more and more attention and become the most advantageous technology used for indoor location.In this paper, the technology of Radio Frequency Identification is introduced first, then the technology of the existing indoor location is disgusted, all kinds of algorithms used in indoor location are comprehensively researched, the advantage and disadvantage and applicability of the algorithms are compared, finally the nearest neighbor tag algorithm based on LANDMARC is selected as the location algorithm of the location system in this paper, and the algorithm is analyzed and improved. Then based on the R1000 radio development platform produced by Intel corporation, the antennas work in the way of TDM, simultaneously the location algorithm selected in this paper is coordinated with the hardware platform in the form of developing application program on the development platform.Finally the algorithm is tested on the software and hardware of the system platform. The data of the location result tested with all kinds of parameters is compared and analyzed, the histogram and the cumulative distribution curve of the location error probability distribution is provided. In the environment without obstacle, using 3 pair of antennas and 4 nearest neighbor tags, the average location error is 0.25m, the probability of the error which is less than 0.4m is 90%. At last the improved algorithm is actually tested and the navigation of the location system is simply extended.