Predicting PON Networking Traffic Flow Based on LSTM Neural Network with Periodic Characteristic Data

摘要

PON(Passive Optical Network)traffic prediction can provide data base for port expansion and bandwidth dynamic adjustment, so as to simplify PON traffic operation and improve bandwidth utilization. In this paper, based on the LSTM(Long Short-Term Memory)neural network, the characteristic data is redesigned based on the periodic characteristics of the PON port traffic. Compared with ARIMA(Autoregressive Integrated Moving Average model)and the basic LSTM neural network, the prediction accuracy is significantly improved and the calculation time is reduced.