Dual power source aware algorithms for green optical network survivability

摘要

Survivability is an essential feature of high-speed networks. Transmission rates have risen to hundreds of gigabits per strand of optical fiber, thus disruptions can cause terabytes of data to be lost. Quick restoration is necessary to fulfill the expectations and requirements of users and customers, defined via service level agreements (SLA). Common practice is to establish a primary path on which the signal is transmitted and to reserve a secondary path that will be used in case of failure. Inherently this addition is not free; extra paths may lead to more labor cost, need for supplemental equipment and certainly higher energy requirements. This increased energy consumption, widely generated from fossil fuels, leads to more pollution. However, if there are renewable energy sources available at a number of nodal sites in the network, some specific paths cause fewer CO2 emissions than others. Availability and spatial density of these green sources is increasing worldwide. To lower a network's overall green house gas (GHG) emissions we develop different approaches to setting up these paths and study their respective impact on the environment. We show that if we optimize both the primary and secondary paths by individual, unique metrics, we can considerably lower GHG emissions while maintaining the desired survivability. Our high utilization of low carbon (HULC) sites heuristic aims to minimize the primary path's ecological impact but allows higher emission values for the secondary paths. Therefore, smooth network operations are guaranteed, yet the carbon footprint is significantly reduced.