Research On Improved Algorithm Of Indoor Location Based On Fingerprint

With the rapid development of mobile communication technology,people’s demand for Location Based Services(LBS)is increasing.Compared with the mature outdoor positioning technology,indoor positioning has higher requirements in terms of accuracy,cost,and anti-interference.At present,the most common indoor positioning algorithm based on Received Signal Strength Indication(RSSI)is susceptible to indoor obstacles,signal multipath effects,etc,making the distance error calculated based on the RSSI path loss formula larger.Aiming at the problems that RSSI is vulnerable to external interference in indoor environments and the online matching accuracy of fingerprint-based indoor positioning algorithms is low,and positioning accuracy is not high,this paper has done the following two aspects of research work:1.Propose an optimization algorithm based on RSSI path loss.By analyzing the change characteristics of RSSI,select representative reference node data to quantitatively calculate the path loss factor and distance deviation in the real environment.Calculate the confidence interval of the distance between the node to be located and the wireless access point(AP)according to the measured value of RSSI and the distance deviation.The intersection of the confidence intervals is obtained as the estimated range of the position of the node to be located.2.Propose an improved algorithm for fingerprint-based indoor positioning.Firstly,the RSSI is preprocessed to reduce the influence of unstable measurement values on the positioning results;then the support vector machine algorithm(SVM)is used to classify the area to be positioned to obtain its corresponding area number;traverse the area in the fingerprint database The reference node data with the same number combines the three distance formulas of Euclidean,Manhattan,and Chebyshev to calculate the RSSI similarity between the two nodes to obtain the position estimate;then judge whether the position estimate is within the intersection range of the confidence interval,if so The position estimate is updated to the center point of the intersection;finally,the Pedestrian Dead Reckoning(PDR)algorithm is combined to perform Particle Filtering(PF)to output the positioning result.Experimental results show that the improved algorithm effectively improves the accuracy of indoor positioning.