Research On Indoor Fingerprint Location Technology Based On RSSI Data Modified

The 21 st century is in the era of rapid development of informatization.All walks of life are rapidly advancing toward informatization and digitization.The rapid development of information technology has brought new opportunities and opportunities for indoor positioning technology based on Location Based Service(LBS)challenge.At present,the research on indoor fingerprint positioning technology is more,but the existing research work has the following two main problems:(1)In the actual positioning process,the received signal strength indication(RSSI)is only Simple filtering is performed to reduce its volatility,and it is impossible to obtain RSSI data with a small deviation from the real RSSI value,which affects the positioning accuracy;(2)In the actual positioning process,the existing Wi-Fi signal is used for position sensing.Algorithms,such as Support Vector Machines(SVM)classification algorithms,Bayesian classification algorithms,etc.,have low recognition accuracy and poor stability in large indoor locations.In order to solve the above-mentioned problems,this paper chooses indoor Wi-Fi fingerprint positioning technology as a research topic and launches related research work.The 21 st century is in the era of rapid development of informatization.All walks of life are making rapid progress towards informatization and digitization.The rapid development of information technology has brought new opportunities and opportunities to indoor positioning technology based on Location Based Service(LBS).challenge.At present,the research on indoor fingerprint positioning technology is more,but the existing research work has the following two main problems:(1)In the actual positioning process,the received signal strength indication value(RSSI)is only Simple filtering is performed to reduce its volatility,and it is impossible to obtain RSSI data with a small deviation from the real RSSI value,which affects the positioning accuracy;(2)In the actual positioning process,the existing Wi-Fi signal is used for position sensing.Algorithms,such as Support Vector Machines(SVM)classification algorithms,Bayesian classification algorithms,etc.,have poor recognition accuracy and poor stability in large indoor locations.In order to solve the above-mentioned problems,this paper chooses indoor Wi-Fi fingerprint positioning technology as a research topic,and launches related research work.By expounding the research status of indoor positioning technology at home and abroad,gain insight into the research dynamics of scientific researchers,and carry out related research work in conjunction with the research topics of this article.This paper proposes two improved indoor fingerprint positioning methods:(1)an indoor fingerprint positioning method based on GF-KF modified RSSI;(2)an improved indoor location sensing method based on a combination optimization algorithm.This article focuses on the above two methods to carry out research work,the main research work is as follows:(1)By collecting a large amount of Wi-Fi data in a complex indoor environment,the RSSI data received by the mobile terminal is analyzed and studied.This article mainly conducts experimental verification and analysis on the three characteristics of RSSI signals:attenuation,time-varying,and spatial difference.At the same time,it also conducts theoretical research and experimental comparative analysis on common RSSI data processing methods.(2)Aiming at the problem that Wi-Fi signals are easily affected by external uncertain factors such as noise and the RSSI value received by the mobile terminal is deviated from the real RSSI value,which results in low indoor positioning accuracy,a modified RSSI based on GF-KF is proposed.Indoor fingerprint positioning method.This method uses the characteristics of RSSI-like Gaussian distribution to perform Gaussian Fitting(GF)on RSSI data to obtain a more definite RSSI value.The error correction of the fitted RSSI data is carried out by introducing the Kalman Filtering(KF)algorithm,and then combined with the Weighted K-Nearest Neighbor(WKNN)algorithm for online matching and positioning.Experimental results show that the average positioning error of this method is 1.5m,and the distribution probability of positioning error within 2.0m is90.06%.(3)Aiming at the problem of low recognition accuracy of using Wi-Fi signals to locate moving targets such as people in an indoor environment,an indoor position sensing method based on a combined optimization algorithm is proposed.Use mobile terminals such as smartphones to collect RSSI to build a user's location fingerprint database while walking indoors,and combine particle swarm optimization(PSO)and gravitational search algorithm(GSA)to obtain training neural network models The weight coefficient is used for model training,and the trained model is used to test the effect of indoor location perception.Experimental results show that the indoor location perception recognition accuracy of this method is 98.35%.Compared with neural networks,Bayes,support vector machines and similar methods based on neural network improvements,the recognition accuracy is at least 4.79% higher.