Research And Implementation Of Propagation Model And Correction Algorithm In Wireless Network Planning

In the actual construction process of base station,wireless network planning is one of the important links,which plays a crucial role in the service scope and service quality of the wireless communication network finally constructed.Among them,service quality is the top priority in wireless network planning.Signal quality is an important indicator reflecting the quality of wireless network service.It is mainly affected by the wireless propagation environment between transmitting antenna and receiving antenna.Path loss is a mathematical indicator to measure the degree of influence,and the most effective tool to predict path loss is wireless propagation model.In a word,wireless propagation model is the basis of network planning,and the accuracy of the propagation model is closely related to the rationality of wireless network layout scheme.Existing propagation model correction methods are mainly aimed at traditional wireless propagation models,such as using the classical least square method to correct the SPM model,while there is no definite correction method for the deterministic propagation model.This paper first describes the basic principle and process of wireless propagation model correction,and then proposes and implements three model correction methods suitable for ray tracing model.According to the actual road test data,the overall scheme of propagation model correction can be divided into two parts:data preprocessing and specific model correction.The data preprocessing part provides more accurate data preparation for the model correction part through pre-processing the original data.The core communication model correction part aims to design relevant algorithms to achieve the fitting of road test data,so as to ensure that the corrected communication model can be more consistent with the local actual communication environment,can be better applied to the actual network planning,and meet the requirements of network coverage.The key point of this module is the design and implementation of algorithm.In this paper,two kinds of propagation model correction methods are proposed,including linear optimization method:least square method,and nonlinear optimization method:Newton method.By studying the basic principles of the two correction methods,they are applied to the correction of ray tracing model.At the same time,a self-classification intelligent propagation model based on artificial neural network is proposed.Finally,based on a large number of road test data of a city,the three proposed methods are verified and analyzed respectively to prove that they can meet the requirements of model optimization.The least square method and Newton method need less time and can get the exact model parameters.As a linear optimization method,the least square method needs to ignore some data in the correction process,while the Newton method of nonlinear optimization method can effectively use all the test data and get more accurate correction results.Compared with the traditional model correction method,the self-classification intelligent propagation model requires a longer time,but the prediction results of the model are more concentrated,which can better solve the problems of the universality and accuracy of the propagation model.In engineering practice,different optimization methods can be selected for wireless network planning according to different calibration requirements and application scenarios.