Research On Indoor Position Fingerprint Location Algorithm Based On RSSI

With the development and popularization of Io T technology,services based on physical location show broad market prospects,such as emergency relief,health care,positioning and navigation,etc.However,the traditional indoor positioning technology such as Bluetooth,infrared,sensor and so on,because of the high hardware and software requirements,hinder the application of these technologies in the actual indoor positioning scene.With the rapid popularization of wireless network,it provides a new research idea for indoor positioning,and the indoor positioning technology of receiving signal strength indication(Received Signal Strength Indication,rssi)in wireless signal can realize accurate indoor positioning only with the help of wireless network resources.,Significantly reducing deployment costs,thus having a clear advantage in real-world applications.In this paper,an indoor position fingerprint technology location method based on RSSI is studied and analyzed,and a position fingerprint location algorithm based on Gaussian hybrid model and improved K near neighbor is proposed GMM-DWKNN,the main work is as follows:Firstly,according to the received signal strength indication in the wireless network,fingerprint information is collected in the offline location of indoor location fingerprint location to build a fingerprint database.In order to speed up the positioning speed and improve the timeliness,the fingerprint database is clustered to form a fingerprint cluster.However,the existing K-means clustering algorithm can only deal with data sets with convex shapes,it is difficult to converge or even cluster non-convex data.Aiming at this problem,this paper uses the clustering algorithm of Gaussian mixture model to cluster the fingerprint information of the reference point.The clustering process gives the probability that the fingerprint belongs to a certain cluster,and finally the reference point is classified into the cluster with higher probability.Compared with the clustering algorithm such as K-means,it is more general and can get better clustering effect.Secondly,in the indoor location fingerprint positioning online stage,the fingerprint information collected at the target to be located is matched with the fingerprint information in the fingerprint database,and the corresponding positioning algorithm is used to estimate the physical location of the target to be positioned.However,the K near-neighbor positioning algorithm,which is often used for indoor positioning,does not fully take into account the different degree of fingerprint information contribution of different reference points in the actual positioning process,and the DWKNN location algorithm proposed in this paper gives different weight information according to the different contribution degree of different reference points,thus realizing the improvement of positioning accuracy.Finally,the GMM-DWKNN algorithm is applied to the location fingerprint model for indoor positioning experiments.The experimental results show that compared with the original positioning algorithm of the location fingerprint model,it can improve in the real-time positioning stage.Positioning accuracy effectively reduces positioning errors.