Research On Algorithm Of Indoor Wireless Localization Based On Artificial Neural Network And Heron's Formula

Indoor positioning technology has been widely used in many fields and has become one of the important research objects.Accurately obtaining the location information of sensor nodes is the basis for completing other tasks.Optimized positioning algorithms can effectively reduce the impact of wireless channels and use less network resources to obtain higher positioning accuracy.This paper mainly studies the localization of indoor nodes based on artificial neural network(ANN)and Heron formula.The experimental environment is built by Python language.(1)Aiming at the problem of RSSI ranging affected by indoor environment,this paper proposes a distance measurement method based on ANN,which is combined with ANN's good data fitting ability,strong de-noising performance and supporting parallel computing.Firstly,the characteristics of wireless signal propagation and wireless channel are analyzed.Then,the parameter structure and training scheme of ANN network are introduced.And the influence of different network structures on the convergence speed and time overhead is compared to determine the network parameters.Finally,the ANN model is trained by anchor nodes deployed in WSN,and the model is used to complete the ranging function.Simulation results show that this method can effectively suppress the impact of noise on the ranging and has better fitting ability.(2)For the positioning method based on the Heron's formula,the method first uses Heron formula to test if the unknown node is within the triangle surrounded by the neighbor nodes,and then records the decision result into the grid array.Through the Heron test,the triangle formed by the neighbor nodes is traversed,and finally the maximum area of the centroid in grid is as the estimated coordinate.The algorithm has better positioning accuracy and better robustness.(3)Combining ANN Ranging with Heron Ranging,the indoor positioning algorithm of ANN-H is proposed.The feasibility of the algorithm is proved through simulation experiments,and the proposed algorithm overcomes the influence of the network topology and deployment environment existed in the traditional positioning algorithm,reduces the error brought by the traditional RSSI ranging.The proposed method has good robustness and self-adaptability,and at the same time improves the anti-interference capability.The localization accuracy is also better than the traditional centroid algorithm.