Dynamic Flow Migration for Delay Constrained Traffic in Software-Defined Networks

Various industrial control applications have stringentend-to-end latency requirements in the order of a fewmilliseconds. Software-defined networking (SDN) is a promisingsolution in order to meet these stringent requirements undervarying traffic patterns, as it enables the flexible managementof flows across the network. Thus, SDN allows to ensure thattraffic flows use congestion-free paths, reducing the delay toforwarding and processing delays at the SDN nodes. However,accommodating new flows at runtime is under such a settingchallenging as it may require the migration of existing flows,without interrupting ongoing traffic. In this paper, we considerthe problem of dynamic flow migration and propose a polynomialtime algorithm that can find a solution if direct flow migrationis feasible. We furthermore propose an algorithm for computingboth direct and indirect flow migration and prove its correctness.Numerical results obtained on a FatTree network topology showthat flow migration is typically necessary for networks with amoderate number of flows, while direct flow migration is feasiblein around 60% of the cases