Research On Improvement Of Indoor Location Algorithm Based On WiFi

For the modern Internet era,location information services have brought a lot of convenience to people,such services as outdoor driving path planning,dynamic driving position of online ride-hailing vehicles,indoor underground parking space navigation,store navigation in large shopping malls,and so on.It allows people to know where they are at any time and arrive at their destination quickly and accurately.Outdoor GPS(Global Positioning System)navigation technology has been quite mature,but it is not suitable for indoor positioning,and indoor positioning is exactly what people urgently need.Therefore,reliable and efficient indoor positioning research becomes particularly urgent.After collecting and reading a large number of relevant materials,this paper carries out detailed analysis and research on the indoor positioning technology,which is commonly used nowadays.Indoor positioning based on WiFi is accomplished by virtue of the transmission characteristics of RSSI(Received Signal Strength Indication)attenuation of WiFi signal with the increase of distance.According to different positioning principles,indoor positioning can be divided into positioning based on propagation models and positioning based on location fingerprints.In this paper,indoor positioning research will be carried out based on location fingerprints.Firstly,in terms of signal noise reduction processing,on the basis of theoretical and experimental analysis,the disadvantages of WiFi signal being easily interfered by all kinds of noises in the real environment are analyzed.Thus,a four-way dual filter is proposed to process WiFi signal.When collecting WiFi signals,the WiFi signals in the positioning environment are collected from four directions for many times,and then the collected signals are processed by gaussian filtering method first,finally processed by the average filtering method.To some extent,the influence of noise on positioning is reduced,and the accuracy of indoor positioning is improved.Secondly,in the aspect of WiFi location algorithm,the nearest-neighbor method is widely used due to its strong applicability and ease of use without too much consideration of parameter Settings.However,this method has the disadvantages of low positioning accuracy and low computing efficiency.On the basis of WKNN,this paper proposes a learning vector quantization clustering algorithm based on region division and the processing of clustering boundary points.In addition,it introduces the steps and process of the improved algorithm indetail.Finally,the effectiveness of the improved algorithm is verified and analyzed by experiments.In the experimental environment of this paper,the average matching efficiency of the improved algorithm combined with learning vector quantization clustering is 85%higher than that of the original algorithm,and the average positioning accuracy is improved from 2.27 meters to 2.14 meters.