A Partitioning Algorithm to Solve the Controller Placement Problem in Software‐Defined Networks

ABSTRACT

A Centre of Mass‐Based Sorting algorithm (CMBS) is proposed that divides a network into multiple sub‐networks and deploys one controller in each sub‐network. CMBS has been derived by minimizing the average distance of the network nodes. The performance of the CMBS method has been evaluated on real‐world network topologies and compared with two state‐of‐the‐art approaches, such as Particle Swarm Optimization (PSO) and Density‐Based Controller Placement (DBCP). Experimental results show that CMBS requires fewer controllers and the lowest average Switch‐to‐Controller (S‐C) latency for a given network compared to others. It is shown that CMBS is not only able to minimize the (S‐C) latency but is also beneficial to reduce the system cost. The proposed method is a cost‐effective solution with the lowest average (S‐C) latency for the controller placement problem in Software‐Defined Networks compared to (PSO and DBCP).