Research And Implementation For Key Algorithm Of Wi-Fi Indoor Location Based On RSSI

In the era of digital consumption and wisdom experience,fast,real-time and accurate indoor positioning capabilities can combine the physical objects of real world with the virtual space data information,which dramatically change the mode of operation for retail,manufacturing,logistics,first aid and other industries.The connection between human beings and things becomes closer together,which is the basis for truly realizing the interconnection of all things.With more and more complex interior space and diverse indoor activities,the demand for positioning and guidance of parking lots,shopping malls,airports and other places is becoming increasingly stronger.At the same time,industries such as precision marketing,intelligent robot manufacturing,and unmanned medical care also require computers to identify specific objects in the room.Future smart manufacturing,smart city,and smart building applications will all rely on high-precision indoor positioning capabilities.All these needs bring great opportunities for indoor positioning technology,but also face many unique challenges.In this paper,the Wi-Fi indoor positioning technology and the RSSI-based Wi-Fi indoor positioning algorithm are researched in depth.According to the shortcomings of the existing Wi-Fi indoor positioning method based on RSSI,the corresponding optimization and improvement schemes are proposed.Then,an indoor positioning method based on local linear regression algorithm and unknown AP prediction is completed.This paper mainly did the following research:(1)Aiming at the difference of environment around each access point AP in the real indoor environment,the local linear regression algorithm is used to fit the mapping relationship between the signal strength and the distance of each AP,which avoids the possible under-fitting phenomenon with the same linear model for all access points;(2)This paper analyzes the advantages and limitations of the RSSI-based indoor positioning methods.Aiming at one situation that with noise interference in the signal propagation process,the signal coverage area of AP does not exactly meet at one point,this paper researches some regional positioning methods and then proposes the probabilistic weighted location method,which effectively makes use of the original data and improves the positioning accuracy,also.(3)Considering the fact that in addition to the deployed known APs in the actual environment,many APs with stable signal propagation but unknown locations exist.In order to make full use of these devices,based on the basic positioning solution,the method to convert the unknown AP to known AP is studied correspondingly.To a certain extent,the scalability of indoor positioning method in this paper is improved.(4)Aiming at the storage and maintenance of original collected data in indoor location system,this paper introduces the Hadoop cloud platform technology and studies the Hadoop basic components and MapReduce programming model.