Research On User Location Area Detection Technology In Indoor Scene

With the rapid development of wireless networks,wireless devices that provide network access services are gradually entering the home.People are also looking for a solution that can detect the area where the user is located and limit the network requests of users outside the designated area.However,the current detection scheme is difficult to promote due to the difficulties of deployment,the accuracy is easily affected by the environment,and the cost is high.Therefore,this thesis proposes a location area detection scheme based on Channel State Information(CSI),which can quickly determine the location area of a user using only one device.This solution does not require much preparation,and can be directly integrated and deployed with commercial Wi-Fi equipment.In this thesis,the proposed scheme is theoretically analyzed and verified in practical scenarios.The research contents of this thesis are as follows:First,an area detection scheme based on the convex hull of reflection points is proposed.The scheme uses the two-dimensional Matrix Pencil algorithm to estimate the two-dimensional parameters of the Wi-Fi signal,and after analyzing the errors in the CSI,a differential time-of-flight model is derived,then the position model of the reflection point is constructed with this model.By constructing the convex hull of reflection points,combined with spatial constraints and adaptive clustering algorithm,the area where the target is located relative to the receiver is obtained.Then,in order to realize the location area of the target,it is also necessary to perform distance estimation on the detected target located in the same area as the receiver.This thesis first proposes an adaptive model,which only needs to collect several calibration points in the to-be-located area in advance to fit a linear relationship.The data packets collected in real time can obtain the corresponding loss coefficient after obtaining the environmental factor,and then substitute the loss coefficient into the classic "path-loss" model to achieve distance estimation.After accumulating multiple data packets and eliminating outliers,the location area of the target is obtained.Finally,in order to verify the scheme,this thesis builds a set of experimental platforms to verify the effectiveness of the area detection and localization algorithms in three typical indoor scenarios.The experimental results show that the CSI-based indoor scene location area detection technology can achieve 96.38%,94.19% and 90.45% area detection accuracy in three scenarios.The positioning accuracy of 93.16%,95.51% and92.52% can be achieved respectively under the grid division of 1.5 m ×1.5 m,and it also has good performance of 91.24%,94.44% and 90.38% under the grid division of 1.0 m× 1.0 m.