Research On Indoor Location Fingerprint Location Technology Based On WiFi

In this paper, off line, a Radio Map optimization algorithm based on the normal test is proposed to establish high precision Radio Map. First, the algorithm determines whether the measured signal samples are in the normal distribution;secondly, the probability density of normal samples is estimated by normal distribution function, otherwise the kernel function is used instead; finally, the average value of the signal which has high probability is selected into Radio Map.The experimental results show, that the method accurately estimates the distribution of WiFi signal sampling in indoor environment, effectively eliminates the singular value of the signal samples, establishes the Radio Map with high precision, to improve the positioning accuracy of the system. Compared with the mean model method and the normal model method, the average positioning error is respectively reduced by 15.62%, 8.46%.On line, a hybrid localization algorithm is proposed. It introduces the PSO algorithm into the traditional K weighted nearest neighbor algorithm (WKNN). This algorithm is divided into two steps, the first step: estimating position by WKNN algorithm; the second step: using PSO algorithm to optimize the estimated location.The experimental results show that the positioning accuracy is improved by 19.89% compared with the K weighted nearest neighbor algorithm, The probability of positioning error less than 2m is 75%, which is greater than that of WKNN algorithm by 66.67%. Therefore, the estimation results of the hybrid location algorithm are closer to the actual location, but also have higher stability.The "fingerprint data collector" is designed and developed in this paper. The application is based on Android and had developed in Eclipse+ADT+Android SDK environment with Java language. It has fingerprint data acquisition and storage functions.