A dynamic heterogeneous network traffic prediction method based on deep spatiotemporal neural network

Abstract

The invention belongs to the technical field of wireless communication, and in particular relates to a dynamic heterogeneous network flow prediction method based on a deep spatio-temporal neural network. Aiming at the problems of small coverage area, low prediction accuracy and short prediction time of existing mobile data traffic forecasting methods, a dynamic heterogeneous network traffic forecasting method based on deep spatiotemporal neural network is studied. Considering the characteristics of user mobility and traffic data spatio-temporal correlation, in-depth research on the long-term mobile data traffic prediction mathematical model description method with wide coverage in dynamic heterogeneous networks; To predict the long-term trend of mobile traffic in dynamic heterogeneous networks; study the spatial-temporal correlation of three-dimensional convolutional neural network models to capture the small fluctuations of mobile traffic sequences in dynamic heterogeneous networks; integrate the long-term trend prediction model and short-term change model of mobile traffic above, In this way, long-term mobile traffic prediction with wide coverage and high precision in dynamic heterogeneous networks can be realized.

Description

technical field [0001] The invention belongs to the technical field of wireless communication, and in particular relates to a dynamic heterogeneous network flow prediction method based on a deep spatio-temporal neural network. Background technique [0002] In order to meet the increasing transmission rate and mobile data volume requirements of mobile users, heterogeneous networks deploy different types of low-power small cells on the basis of traditional macro cells, and the system capacity and network coverage capabilities are significantly improved. In order to meet the demand for on-demand traffic offloading in hotspot areas, small cells need to optimize communication parameters in advance to meet instant communication requests, and UAV base stations must be deployed in hotspot areas in advance by adjusting their locations. Therefore, with the help of powerful data analysis methods of machine learning, traffic congestion events can be predicted and the gap between supply ...

AI Technical Summary

Problems solved by technology

[0005] In order to overcome the problems of small coverage area, low prediction accuracy, and short prediction time of existing mobile data traffic forecasting methods, the purpose of the present invention is to propose a dynamic heterogeneous network traffic forecasting method based on deep spatio-temporal neural network

Images