Optimization Research Of Outdoor Fingerprint Database Location Technology Based On Fusion Clustering

In nowaday booming mobile communication network,location-based services have added low-latency and high-reliability positioning scenarios,and the demand for this has become higher standards.Location services bring convenience to us,and they attract our high attention.People are looking forward to enjoying location services which have higher accuracy and faster speed,so positioning technology is still a research hot spot.Among the passive positioning technologies that do not require additional power consumption,the location fingerprint has gained many favors due to its high positioning accuracy.Whether in indoor and outdoor positioning scenarios,location fingerprint technology is currently receiving the most attention.However,the location fingerprint is also caught in the dilemma of slow positioning speed because of the huge amount of data in the database,that cannot meet the higher speed requirements of location services.At the same time,in the indoor positioning scenario,the location fingerprint has been developed to the stage of selecting a suitable clustering algorithm to optimize the fingerprint database,which can improve the speed.The current clustering algorithms which used to positioning scene still have a lot of room for improvement.The outdoor location fingerprint technology is still at the stage of constructing the fingerprint database and not considering the improvement of the speed due to the changeable environment and the complex construction of the fingerprint database.But outdoor positioning scenes are more extensive,the larger the fingerprint database,the slower the positioning speed.So the positioning speed of fingerprints in outdoor locations needs to be improved urgently.Therefore,this paper proposes a fusion clustering algorithm based on the problem of location fingerprint’s slow speed in outdoor scene and poor performance of traditional clustering algorithms.Based on this,the major work is as follows:Firstly,an idea to improve the positioning speed in the outdoor scene is proposed that using clustering algorithm to optimize the processing of the location fingerprint database.After analyzing the research work and achievements of location fingerprint for indoor and outdoor scenes worldwide,this paper analyzes the bottleneck of location fingerprinting technology for outdoor scenes.Aiming at the bottleneck of slow positioning speed of outdoor fingerprint database,a clustering algorithm is proposed to optimize the outdoor fingerprint database.The simulation proves that the positioning speed of the outdoor fingerprint database optimized by the clustering algorithm is at least 40.65%higher than that of the traditional outdoor fingerprint database.Secondly,an improved fusion clustering algorithm is put forward,which enhances the effect of the traditional K-means clustering algorithm.Several clustering algorithms which applied to indoor fingerprint positioning are analyzed by this paper,including K-means,CFSFDP,hierarchical clustering,DBSCAN,and I-CFSFDP.An improved fusion clustering algorithm that combines two clustering algorithms is proposed to further enhance the clustering effect of the traditional K-means algorithm.In the simulation process,the improved fusion clustering algorithm and the traditional K-means algorithm are respectively applied to the outdoor fingerprint database positioning.The simulation results show that the positioning accuracy of the former is improved by 42.69%compared with the latter.Then,in view of the accuracy of the online-stage matching algorithm of the fingerprint positioning,WKNN localization algorithm is optimized and improved by the adaptive clustering algorithm.The simulation results show that the improved WKNN algorithm is 18.11%better than the traditional WKNN algorithm.In addition,The SPM propagation model to simulate the form of RSRP signal is adopted by this paper,and the universal Kriging interpolation algorithm is used to establish the fingerprint database of the real base station location in the outdoor locating area,and an overall simulation platform is built.When verifying the algorithm in the simulation stage,the fingerprint database processed by different clustering algorithms and the traditional fingerprint database not processed by clustering were selected.Combined with the traditional WKNN algorithm and the improved WKNN algorithm as comparative cases,the experimental results were obtained.Finally,we analyze the future research orientation in view of above content.