Energy Efficient GPON using Neural Network Traffic Prediction

Since the fourth industrial revolution has geared up, the demand for high data sharing and data storage has increased, which has led to the need for incremental bandwidth demand. Also, it has led to a massive increase in energy consumption. Although the power requirement of the optical network is low, gradually the trend is shifting to all-optical networks, and very soon the high demand for optical networks everywhere will lead to book significant overall consumption. The proposed architecture not only supports the growing demand for bandwidth but also makes it energy efficient by using artificial neural networks. ANN algorithms are used to predict the traffic flow of a network and hence smartly switch the transmitting module ON and OFF in order to save power consumption. Traffic prediction is access network dependent and therefore can adapt to the changing pattern of different environmental conditions.