A Simulated Annealing Strategy for Reliable Controller Placement in Software Defined

Abstract

Software defined networking (SDN) empower the service providers and enterprisers to respond quickly to changing business requirements. Due to the logically centralized but physically distributed control architecture the resilience of the control plane is a major challenging issue. Failure of a controller needs reassignment of switches to other working controllers with enough capacity. Due to this, the propagation latency may increase which will lead to the increase cost of the network. In this paper, a heuristic technique called simulated annealing strategy for reliable controller placement (SASRCP) is proposed which reduces the propagation latency and minimizes the routing cost of the network. This algorithm is compared with a greedy (GRE) and a random algorithm (RA). The evaluation results confirms that SASRCP performs better than GRE and RA.