基于通信导频数据环境识别研究OA

Based on the real communication equipment,the pilot data for channel equalization is obtained by modifying the physical layer protocol,and the delay power spectrum of the channel is obtained through the mathematical processing of the pilot data.Then,this paper conducts field measurements of the office environment with dense buildings,the environment of frontage roads and the suburban environment with sparse buildings,and based on the measured delay power spectrum data,the delay power spectrum image data is obtained through data processing for the learning and verification of the three environmental identifications.Finally,the ResNet50 deep network is constructed by deep learning,and the data are trained and verified,and the simulation results show that the recognition accuracy of the three environments can reach 90.31%.In this paper,the research on environment recognition based on pilot data is applicable to the integrated upgrade of real communication equipment,which can realize the recognition of the environment by transplanting the training model without the need for communication protocol reconstruction.