Energy Savings by Using Traffic Estimation for Dynamic Capacity Adaptation in Communication Network Operations.

Energy efficiency of telecommunication networks plays an essential role in the context of sustainability and climate change – as those networks are large power-consuming distributed infrastructures. Furthermore, also an economical network operation calls for low energy demand. A challenging and crucial task for energy-ef ficient and sustainable network operation is the load-adaptive operation of network elements such as routers, switches and access multiplexers. Since the traffic is temporally fluctuating, load-adaptive control of the network requires a robust traffic demand estimation. This is also of overwhelming importance, as a stable network operation is a central task of network operators – since it is expected by their customers as the service they pay for. Here, Wiener filtering has been identified as a robust solution for reliable traffic demand forecasting on relevant time scales. The results presented in this paper show that the capacity dimensioning based on the proposed Wiener filtering traffic forecasting leads to reliable outcomes in terms of predicted traffic enabling sustainable and efficient network operation.