The indoor wireless location technology research based on WiFi

摘要

The main research content of this article is based on fingerprint method of AP selection and location estimation algorithm. We introduce RANSAC algorithm used in image processing art to AP selection in the online stage for external detection. It can filter to remove the APs impacted by environmental variation, not only reduces the amount of calculation but also improves the positioning accuracy. Aiming at the disadvantages of traditional Bayesian algorithm and KNN algorithm, we improve the two kinds of algorithms. Based on traditional Bayesian algorithm, we adopt the concept of a regional division. Classification based on the traditional KNN algorithm is introduced into cluster and the cluster partition, allows a reference point to be assigned to multiple clusters, using different fingerprint in different clusters. Finally we adopt a new method of dynamic union combined with the above two kinds of improved algorithm. Based on the above research, the average error of our positioning system is 1.63 meters, the minimum error is 0.76 meters.