Distribution Model for OpenFlow-Based Networks

OpenFlow, a popular software-defined networking (SDN) standard, is a promising approach to computer networks that adapts to current developments in information technology. Although the OpenFlow architecture facilitates the effective dynamic control of substantial networks from a centralized logical framework, it is vulnerable to single-point of failures when dealing with a substantial increase in traffic. Studies show that task distribution has potential to address single-point of failures. In this paper, a distribution model is introduced to reduce the load on the SDN controller by moving the distribution task of the data from the controller into one or more of its associated switches. This approach utilizes any available and accessible network devices to enable task distribution proficiently without requiring any additional resources. A simulation program using Python code is developed to evaluate the impact of using distributor switches on performance of the controller. Experimental results demonstrate that the measure of data transmitted by the controller is decreased by up to 76%, while the total transmission delay of entries is reduced by up to 29%.