In this era of massive cyber traffic, there is a corresponding number of Internet devices consuming billions of watts, independent of traffic loads. Proposed solutions to this energy problem include aggregating traffic to improve routing efficiency and switching underutilized devices to sleep. Consequently, routing convergence problems can arise (for example, loops and black holes) resulting in severe packet loss that further amplifies energy consumption.

This paper presents GreenFRR, a scalable and efficient fast-rerouting-based algorithm (FRR). The design attempts to save energy by addressing the difficult problem of guaranteed loop-free routing while maximizing the number of sleeping links (MNSL). It proceeds to prove that MNSL is NP-hard, but includes heuristic algorithms that relax the solution’s complexity via the link-utilization ratio and path-stretch techniques. The GreenFRR base model is then enhanced with advanced packet forwarding (GreenFRR-Adv), which lessens rerouting path overlap and path stretch, yet maintains high energy savings and reduced rerouting convergence time.

The main contribution of this paper is the design of the very energy-efficient GreenFRR. To achieve its design goals, it smartly alleviates problems associated with MNSL. Additionally, the authors carry out a thorough analysis of their work via well-crafted proofs to validate their model. They also compare GreenFRR performance to other methods using a simulation of important performance factors, which proves its superiority over the competitive distributed least flow model. GreenFRR is shown to manage Internet router/link states well enough to sustain solid energy conservation and efficient rerouting.