OPEX cost savings and CO2 reduction are becoming major objectives for operators. To maintain a competitive edge, low energy consumption may become a key differentiator. Increased network complexity is both a risk as well as an opportunity for energy saving. Energy Efficiency and related functionalities will typically need to be distributed across management-, RAN- and BTS domains, thereby impacting network architecture as well as hardware realizations of various nodes.
Network energy consumption scales badly with traffic today; improved scalability and modularity are key for reduction, both at network and equipment level. This is especially important for the introduction of 5G that by careful design can take advantage of improved sleep modes. In this block we will address energy saving features – primarily employing machine learning techniques – to gain RAN level energy savings.