The WiFi Indoor Location Method Based On LMBP Neural Network

In recent years,the number of mobile smart phone users has increased dramatically,In order that users can easily use the network for communication,large cities have laid a large number of WiFi hot spots in public places,Therefore,many indoor positioning researchers take the WiFi hot spot as the access point of the location area,and put forward the indoor location technology based on WiFi.The indoor positioning technology based on WiFi uses the shared hot spot coverage area in the public place as the location area,thus reduces the cost of the system development,and has more market competitiveness compared with other indoor positioning technology.This paper expounds the research significance of indoor positioning technology in the present situation,analyzes the research status of indoor positioning at home and abroad,and conducts intense research on several common indoor positioning technologies.Different indoor positioning techniques have their specific application scenarios,and their positioning accuracy is also different.The improvement of positioning accuracy also depends on choosing the appropriate location algorithm,this paper selects several common indoor positioning algorithms for research and analysis.There are more network nodes to be laid in large indoor stadiums,which leads to the large dimension of sample data collected,and increases the localization time and algorithm complexity.Therefore,it is necessary to preprocess the original data in the off-line stage.In this paper,the feature extraction algorithm is used to process the original data,which effectively reduces the data dimension and removes the redundant information,even in the noisy environment,we can get a better positioning effect.The wireless signal has the characteristics of reflection,diffraction,shadow effect and multipath fading,The localization methods based on propagation model,such as Taylor series method,three side location method,hyperbola method,etc.The location accuracy is not high because of the complex indoor environment.In order to avoid the impact of the propagation model parameters,many researchers have proposed an indoor location algorithm based on BP neural network,which effectively solve the inaccuracy problem of parameters in the signal model.However,the convergence speed of the standard BP neural network is slow and the positioning error is large.In this paper,the indoor localization algorithm based on LMBP neuralnetwork is adopted,and uses the Matlab software to write the neural network code to simulate and analyze the performance of the algorithm.The simulation results show that the LMBP algorithm can effectively solve the problem of slow convergence and improve the positioning accuracy.This paper also analyzes the architecture of the indoor positioning system,writes the client software of the mobile phone and builds the positioning system server on the laptop,and tests the performance of the developed positioning system and records the positioning results.