Linear and Non-Linear Spatio-Temporal Input Selection In Wireless Traffic Networks Prediction using Recurrent Neural Networks

Abstract

For the optimization of computer networks with high bandwidth requirements, wireless network traffic prediction is necessary. Its goal is to reduce maintenance costs and enhance internet services. Feature selection is a major issue in the Multivariate Time Series (MTS) Spatio-temporal modeling. Another problem is the dependency between input features, time-lags, and spatial factor so that an appropriate model is needed. This study aims to provide solutions to two problems. The first is to improve a feature extraction and selection process in Spatio-temporal MTS data for relevant features using Detrended Partial Cross-Correlation Analysis (DPPCA) and non-redundant features associated with linear using Pearson's Correlation (PC) filters and non-linear associations using Symmetrical Uncertainty (SU) and combination of both PCSUF. The second is to develop a Spatio-temporal framework model using Recurrent Neural Networks (RNN) to get a better performance than traditional model. These methods are combined and tested using the dataset of cellular networks with one-hour intervals during November in three locations. Testing the effectiveness of the feature selection technique showed that 27.6% of the total extracted features. The forecasting model with the DPCCA–SU-RNN combination method gets the best performance by having RMSE = 380.7, R2 = 97%, and MAPE = 10%.