Design And Implementation Of Indoor Location System Based On Radio Frequency Sensor Network

In recent years, the internet of things development rapidly, and it has gone deep into our lives. So it is very importment to achieve the person location information. In this article we set the main project which is achieve the person location in internet of things, and design the target tracking system in the room. The system based Zig Bee protocol, use the radio frequency tomography to achieve the location in real-time. So that we solve the problem which the track accuracy is not enough.This article mainly aimed the target tracking problem, study the tracking system based radio frequecy tomography. The radio frequency tomography is a solution which receive the RSSI from some sensors. When the target are near the line-of-sight path between two nodes, the target are easily cause the receive signal strength substantial attenuation or amplification. In this article, we talk about some measurements mode for measure the target distance. include the exponential model, magnitude mode and skew laplacian model. This system successfully tracking two targets in the room, and we compare some tracking algorithm, and exam the performance when target number is known and unknown. and successfully deploy the sensors node in two rooms, and find the problems when tracking the targets. We compare the practicle filter and multi-particle filter and MCMC filter algorithms when the target’s number is known. and find the best algrithm to achieve the tracking errors is 0.3m(single target) and 0.7m(two targets).In our target tracking system, using our propose improvement observation model, have the more accuracy, and the test result shows that this system can successfully track the target, and has the good accuracy, achieve the desire purpose.