Research On Fingerprint Database Construction Algorithm For Indoor Localization Based On Crowdsourcing

With the rapid development of situational awareness services and wireless communication technology,Location Based Service(LBS)has become a hot research topic in contemporary society.The core task of LBS is location technology.At present,indoor localization is mainly based on the positioning technology of radio signals.With the advent of the era of big data,the perception of crowdsensing is attracting more and more attention.Received Signal Strength(RSS)based location method in location fingerprinting can easily use existing wireless networks and intelligent devices to locate,which is the most popular method to study.The unsolved problems for the construction of fingerprint database are as follow:requiring a large number of professional's data acquisition workload to achieve high-precision positioning purposes,fingerprint database is seriously affected by the environment,which need to update regularly.It is very time-consuming and laborious.So in the modern society with the popularity of intelligent devices,the collection method for fingerprint based on crowdsensing can use the data from ordinary users to construct fingerprint database efficiently,but this method also has some challenges: fingerprint redundancy and fingerprint annotation.Considering the positioning accuracy and the workload of data acquisition,this thesis proposes a new method to construct fingerprint database by combining crowdsourcing with unsupervised learning.Firstly,the received signal strength was acquired from Aps by crowdsensing,which is the original fingerprint data set.Then,to deal with the problem of data redundancy and fingerprint annotation of the original fingerprint data,a novel algorithm-FLM based on the combination of learning vector quantization(LVQ)and multidimensional scaling(MDS)was proposed.Finally,the construction of indoor positioning fingerprint database was effectively achieved.Based on Ray-tracing,a simulation scenarios of our proposed algorithm was established.Simulation results show that the efficiency of the fingerprint database construction is improved greatly.80% of localization error is lower than 2.6m by appling to the basic localization algorithm,and the amount of localization calculation of the single position is reduced by 63%.At the same time,this thesis deeply studies the indoor wireless signal propagation characteristics and the positioning technology of inertial navigation.The Friis formula and the ray tracing model are combined to design the indoor RSS simulation scene to obtain the RSS signal in real time,and the wireless access point can be flexibly set.In addition,based on the pedestrian dead reckoning,the RSS is used to simulate the location of the personnel in the active location group to collect the RSS,which is used as the data source for the subsequent fingerprint database construction.The results show that the simulation platform has high practicability and effectiveness.Finally,in this thesis,the existing intelligent terminal is used in the actual environment to carry out verification experiments of FLM algorithm,which is used to verify the effectiveness and robustness of the proposed algorithm.