Research On Measurement Properties And Application Of RSSI For Indoor Location

In recent years,the need for location information and location-based Internet of Things has increased greatly,and indoor positioning technology based on Wi-Fi RSSI has developed rapidly.Based on the basic principle of wireless signal attenuation,this paper makes a continuous study on the characteristics of RSSI signal propagation in indoor space.Firstly,according to the characteristics of RSSI ruler in one-dimensional space,the maximum RSSI observations have been used for signal denoising.Then,according to the characteristics of RSSI signal graph in two-dimensional space,the APs’ Virtual Positions based clustering algorithm has been used for fingerprint reference points clustering.Next,according to the uneven spatial resolution of RSSI,the geometric distance has been used.Finally,the indoor map has been used to optimize the navigation direction,and the indoor fusion positioning system based on APs’ Virtual Positions,geometric distance and Turning Detection has been proposed.The measurement properties of RSSI are studied.First,the measurement properties of RSSI ruler in one-dimensional space have been summarized.1)The bigger the RSSI value is,the better the theoretical spatial resolution is;2)The theoretical spatial resolution of Wi Fi signal intensity scaled by 0.1d B is better than that of integer d B.Then,the characteristics of RSSI signal graph in two-dimensional space have been summarized.1)RSSI signal feature points and RSSI signal feature lines are found based on the experience of a large number of RSSI data,and then RSSI signal graph is divided into regions and mapped separately;2)RSSI signal graph is outlined by imitating DEM contour lines,then the contour map is intersected by RSSI contours.3)Radio Map is segmented and mapped with wireless attenuation model as constraints on RSSI signal graph.In an indoor environment,due to walls and other structures,multipath propagation usually prevail,causing significant fluctuation in the RSSI observed by a mobile device.Meanwhile,shadowing such as due to the presence of a pedestrian between transmitter and receiver especially close to either of them will considerably reduce the RSSI.In addition,the Wi-Fi radio channel is generally shared by different systems or devices,and the interference may also decrease RSSI considerably.Therefore RSSI is significantly affected.Inspired by the measures taken in GPS multi-path data,a filtering algorithm based on the maximum RSSI has been proposed to abandon the poor quality RSSI data caused by environmental interference.The experimental results show that both the positioning accuracy and robustness of the proposed algorithm are better than those of the mean algorithm,Kalman filter algorithm and particle filter algorithm.For positioning in indoor area without linear constraints,such as a large office room or a large lecture theatre,it is difficult to make use of activity landmark corrections either because there are no such reliable landmarks.Therefore,the selection of RPs in indoor area without linear constraints should be carefully conducted for both good accuracy and simple implementation.Inspired by the phenomenon of virtual light source after reflection,refraction and diffraction,APs’ Virtual Positions based clustering algorithm has been proposed.The results demonstrate that reference points in indoor area without linear constraints can be clustered automatically by the proposed clustering algorithm,and positioning accuracy of the proposed clustering algorithm outperforms KNN,WKNN,RPLC and SDC.The most frequently used method in fingerprint positioning is the WKNN algorithm,which weights the reference points’ coordinates by the inverse of the RSSI distance.However,it will suffer from the exponential relationship between RSSI and physical distance.Moreover,both fusion methods and probabilistic methods [8-13] have not addressed the problem of the uneven spatial resolution of Wi-Fi RSSI.Inspired by the weighting index of leveling accuracy,physical distance based weighting algorithm has been proposed.Experimental results show that the proposed weighted algorithm considerably outperforms the KNN,Euclidean-WKNN,Manhattan-WKNN,EWKNN,Li FS and GPR in terms of positioning accuracy which is defined as the cumulative distribution function of position error.Corners are common between different indoor areas,and turns will occur in the walking process.The gyroscope sensor embedded in an ordinary Android mobile phone records the angle velocity,which would change dramatically when the pedestrian turns left or right.Inspired by the extensive use of outdoor maps in outdoor navigation,a directional optimization algorithm based on the indoor maps and landmarks has been proposed for navigation in indoor areas with linear constraints.The experimental results show that the positioning accuracy of WKNN algorithm has been improved obviously with the constraints of moving landmarks and indoor maps.Finally,an indoor fusion positioning system based on RSSI spatial resolution observation filtering,APs’ Virtual Positions,geometric distance and Turning Detection has been proposed.Firstly,all RSSI observations are preprocessed using RSSI filtering algorithm based on the maximum.Then,the indoor areas have been divided into indoor corridors and indoor rooms.For indoor corridors with linear constraints,the combination of the proposed directional optimization algorithm and the proposed physical distance based weighting algorithm has been adopted.For indoor areas without linear constraints,the combination of the proposed APs’ Virtual Positions based clustering algorithm and the proposed physical distance based weighting algorithm has been adopted.