Hotspot Localization and Prediction for Broadband Multimedia Services in 5G Networks

Due to the generation of hotspots and the rapid change of hotspot location, 5G wireless network must have higher flexibility. How to accurately predict the future traffic load is a prominent problem. In this paper, we use Gaussian Random Field (GRF)-based to fit the distribution of spatial traffic density and then locate the hotspots for broadband multimedia services. At the same time, the Long Short-Term Memory (LSTM) model is trained to predict the next moment traffic value by using the actual traffic data collected from base stations. Finally, the model is used to predict the future traffic data of base stations and locate the hotspots. Through analysis and evaluation, numerical results show that the method can effectively locate hotspots.