Our research and design proposal is a network engineering energy efficiency solution that reduces energy consumption and cost through SDN/NFV principles and applications, telemetry data, and big data analytics.
‘Green Power Forwarding (GPF)’, our software-defined networking (SDN) application, aims to promote energy efficiency in the broadband networks by enhancing inter-data center (DC) forwarding decisions by considering new routing factors such as the energy cost at the DC (electricity is less expensive in Wyoming than California) and whether the DC uses renewable energy or not, along with shortest-path/round-trip time. For non-critical traffic (user-defined), the algorithm will make the decision to route traffic via a ‘green’ DC, even if the path is longer than a ‘non-green’ DC route.
In addition to this, we propose to add network functions virtualization (NFV) to complement our GPF application within the DC. Increasing the number of virtual network functions (VNFs) allow reduction in the number of physical boxes at the customer edge as well as inside the DC, which implies reduced energy consumption.
Lastly, we propose to use the Energy Management (EMAN) SNMP MIB to gather telemetry data from network devices such as the power consumption, in addition to bandwidth and throughput at any given time. When we combine this data, we can analyze and visualize crucial parameters such as Mb/Watt/Time for any specific device. This formula will facilitate energy-conscious, informed decision-making (For example in an instance when deciding which vendor device would be more energy efficient).
The research results from this project can be used to provide a new paradigm for energy-aware networking which includes the SDN GPF algorithm/application, an NFV infrastructure, and EMAN for more informed decision making and network monitoring and visualization.