Research On Walkslam Based Radio MAP Crowdsourcing Construction Algorithm

In the last decades,both military and civil have seen increasing need for efficient and accurate location-based service.Location based service,as its name indicates,control features and provide functions regarding location data of the users.Thus,the basis of location-based service is to provide a location estimate of the user according to the measurements of devices distributed in the environment taken by the user's device,which requires mapping measurements into exact locations.In this scenario,a sheer coordinate with respect to a basis is insufficient to indicate the users' situation due to a lack of information about landmarks distributed in the environments.To extract landmarks information manually,however,is inefficient and thus vulnerable to changes of the environments.Simultaneous localization and mapping,or SLAM system,can solve the localization and landmarks information extracting problems,and although originally designed for autonomous robots,some recent SLAM solutions are altered to run on a smartphone for economy and accessibility.This paper presents Walk SLAM,a SLAM solution that estimates both the path taken by the user and the locations of Wi Fi devices in the indoor space,using a smartphone.This solution extends the previous work in the following ways:(1)To solve the reliablility of environment for Radio Map construction,a Radio Map construction algorithm using Walk SLAM is introduced,which features independency on knowledge of the target environment.(2)To give a SLAM solution,this paper presents Walk SLAM,a SLAM solution regarding the walking pattern of human as a factor that can supervise the path estimation.In the live test,results show an noticeable improvement both on the user's poses and landmarks from Walk SLAM in comparison to the traditional Fast SLAM algorithm.(3)To achieve a crowdsourcing construction algorithm with the ability to stitch several Radio Maps together,a leveled landmark registration algorithm is presented.This algorithm is based on ICP algorithm.Evaluation both on simulated and live data shows an improvement on the process' s efficiency and accuracy.