Single and Multi-Type Controllers with Soft Computing Methods and Routing in Software Defined Network

Abstract

The software defined networking topology is enabled for efficient configuration for monitoring and enhance the network performance. The multi-controller placement becomes very critical, and it can be rectified by the proposed approach. The comparison of single type and the multi controllers are discussed in this paper.The first paper is compared with the dragonfly algorithm, in this we have considered only multi POX controllers. So the latency difference depends on the iteration of the algorithm. Thus the latency is reduced in proposed algorithm. Similarly the second paper has compared with the existing algorithm of firefly, which states the multi type of controllers such as POX, NOX, BEACON and FLOODLIGHT, among that latency is calculated and compared. The selection criteria are varied according to its type. Hence the routing is carried out on the two controllers alike POX, and NOX, through ad hoc on demand distance vector routing protocol. Subsequently, the selected controllers specifically are placed and simulated in Pareto-based Optimal Controller placement (POCO) tool for the Internet 2 topology and network simulator(NS2). Finally the comparison results makes the proposed system is better than previous methods.