Research And Realization Of Indoor Location Technology Based On RSSI Probability Distribution

With the rapid development of science and technology and the increasing number of large-scale buildings,people's time in the indoor environment has increased substantially,and the pursuit of convenient life has led to an increase in the demand for location services in the indoor environment.Indoor positioning technology is needed to provide location services in many fields such as smart home,mine,hospital,agricultural production,underground parking lot and so on.Influenced by the complex indoor environment,hardware and other factors,there is no recognized optimal indoor positioning system.ZigBee is suitable for wireless sensor networks because of its low cost,low energy consumption and high fault tolerance.It is selected as the location technology used in this paper.At present,the main problem in the indoor positioning process based on RSSI(Received Signal Strength Indication)is that the indoor environment and other factors have an impact on the positioning accuracy.Through the analysis of the research status at home and abroad,the location errors caused by environmental factors are mainly processed by improving matching algorithm,filtering RSSI or multi-feature fusion.In this paper,the RSSI characteristics are analyzed,and the time-varying,environmental characteristics,probability distribution and eigenvalue selection of RSSI are studied by real-world experimental data.Aiming at the time-varying and environmental characteristics of RSSI,this paper proposes an indoor positioning method based on RSSI probability distribution from the time point of view with the RSSI probability distribution as the fingerprint feature value for a period of time.The indoor location method based on RSSI probability distribution is based on location fingerprint location method.The probability distribution of RSSI is used as fingerprint characteristic value,the KL distance is used as fingerprint similarity,and the coefficient of variation is used to measure the degree of environmental impact of RSSI and to weigh the similarity.In the simulation environment,the influence of RSSI sampling number,number of signal access points,K value and noise intensity on traditional KNN and WKNN positioning methods are compared,and the KL and WCV_KL positioning methods proposed in this paper are also discussed.Finally,the positioning effects of several positioning methods are compared in real environment.The cumulative probability of the average error of KL and WCV_KL positioning methods is 40% and 60%,and the cumulative probability of the average error of less than 2 meters is 100%.The standard deviations of the positioning errors of KNN,WKNN,KL and WCV_KL are 1.77 m,1.76 m,1.00 m and 0.93 m,respectively,are 0.66,0.33 and 0.36.Experiments show that the proposed method has higher positioning accuracy,anti-noise and stability.