Manage Fault Tolerance and Load Balancing through Probability Distribution in SDN

Abstract

Software Defined Network (SDN) provides an agile paradigm that dissociates the control plane from the data plane. Still, a solitary controller has a higher impact of failure in a network. But multiple controllers can avoid a central point of failure in the network. Meanwhile, some challenges are counter like uneven traffic distribution between controllers which become an origin of cascaded failure of controllers; when a controller manages network traffic beyond its capacity. Then drop rate of packets is increasing exponentially in a network. This is happening due to the insufficiency of the load balance technique. So, it is necessary to distribute the appropriate workload among the controllers. Motivate of this paper is to propose a load balancing algorithm that gets rid of these challenges by evaluating an equilibrium state of distribution which describes the long-run probability of controllers by integrating Queuing Technique with Markov Continuous Chain, which aids to lessens the packet drop ratio in the network.