A Multi-Constrained Routing Algorithm for Software Defined Network Based on Nonlinear Annealing

Multi-constrained Quality of Service (QoS) routing algorithm is always a difficult problem in routing research area, which is a NP problem. Software Defined Network (SDN) is a new network architecture, in which there's few research on QoS routing. This paper generates the whole network virtual topology according to the characteristics of SDN, and based on the principle of simulated annealing, proposes a nonlinear annealing algorithm, which adapts to the SDN. Firstly we simplify the network topology by Dijkstra-like algorithm, and then introduce the nonlinear energy function, and then iterate the initial solution according to the simulated temperature, until find a feasible path from the source node to the destination node which satisfies the condition. Experimental results show that this algorithm has a higher success rate, better expansibility of network size, and better transplantation for the SDN than traditional QoS routing algorithm.