Research On Indoor 3D Positioning Technology Based On WiFi

With the development of the IOT,indoor positioning technology is booming and it brought a lot of convenience to people.With the increasing popularity of WLAN,indoor positioning technology based on WiFi has become a hot spot.At the same time,the research on indoor positioning in two-dimensional plane has develop to maturation.However,in many scenarios(for example,in multi-storey underground parking lot),traditional two-dimensional positioning technology is difficult to meet the actual need.In this context,the three-dimensional indoor positioning technology based on WiFi network is discussed in this paper.The main work is as follows:(1)A three-dimensional location fingerprint database is built in a multi-storey building,the fingerprint database compression method was studied through many attempts and experiments,and the method for setting the reference points in the off-line phase for the indoor 3D positioning system based on the WiFi fingerprint database is given.By this method,the scale of the fingerprint database can be reduced by about 81.3%,which greatly reduces the workload required to collect fingerprint data in the off-line phase.(2)The noise reduction method of the fingerprint database based on the singular value decomposition of the Hankel matrix is studied.The simulation results show that with the number of the singular values that the Harkel matrix remained decreases,the average positioning error of the points to be located is keeping stable at first,but when the number of singular values decreases to about one-third of the total number,the average positioning error increases sharply.Before this deterioration occurs,the average positioning error using the denoised fingerprint database was reduced by 6.6% to 8.6% compared with the average positioning error that using the unprocessed fingerprint database,then divide the points to be located into two categories and use different fingerprint databases with different numbers of singular values to locate these two types of points.To a certain degree,the purpose of improving positioning results is achieved.In addition,the solution to reduce the repetitive calculations in the matrix recovery stage of traditional Hankel matrix singular value decomposition is introduced which greatly reduces the computational overhead required.In general,this paper provides theoretical and practical support for the method of setting reference points in the off-line phase in the indoor three-dimensional positioning system based on the WiFi fingerprint database and the method of denoising the fingerprint database based on thesingular value decomposition of the Hankel matrix.