Study On Indoor Positioning Model Of RSSI Based On ZigBee

GPS can not meet the needs of positioning in complex indoor environment.In the logistics system such as warehousing management,stevedoring,circulation processing and other indoor areas,indoor positioning navigation technology has a more extensive and effective application.Most of the current mainstream positioning systems have too high cost.ZigBee technology has the characteristics of low power consumption,low cost,high reliability,high security and high compatibility.It is very suitable for positioning and navigation in factories,exhibition halls,family and so on.So ZigBee is chosen as hardware carrier.RSSI based positioning has the advantages of low hardware cost,easy implementation and low power consumption.Therefore,RSSI is chosen as the basic principle of ranging and positioning.The aim of the research is to locate the target node and get the localization accuracy of the RSSI location model based on ZigBee.It can provide reference for others' research and application and try to improve the positioning accuracy of the RSSI positioning model based on ZigBee.Firstly,indoor location is studied by RSSI signal attenuation model based on ZigBee.On the basis of a large number of experiments in the field,the RSSI sequences are optimized by means of Average filtering and Gaussian filtering,estimating the parameters of the model according to the Least squares method,Kalman filtering algorithm is adopted to reduce the amplitude of RSSI sequence,the sequence is optimized further,and translating the RSSI value into distance base the signal attenuation model,computing the coordinates according to geometric positioning algorithm,and analysising error.After a series of filtering algorithms,especially the optimization of Kalman filter,the error of ranging and positioning is reduced effectively,and the error of positioning is reduced to about 7m.Then,the indoor location is studied by RSSI fingerprint matching model based on ZigBee.The fingerprint database of coordinates and RSSI sequence is constructed on the basis of experiments.Then,the similarity between mobile node fingerprint and in-store fingerprint is compared by the nearest neighbor algorithm,and the corresponding coordinates of the fingerprints are selected as the localization result.Analysing error,positioning accuracy is about 5m.Fingerprint matching model is better than signal attenuation model in location accuracy,while the advantages and disadvantages are opposite in workload,flexibility and cost.