A Novel Time-Based Switch Migration Method (TSSM) for Load Balancing in Distributed Software-Defined Networking

Abstract

Switch migration is predominantly used for workload balancing in distributed Software Defined Network (SDN) controllers. The Time-Sharing Switch Migration (TSSM) technique allowed the switch's responsibility to be divided based on the time duration condition of overload, which resulted in fewer overload events, fewer ping-pong controller issues, and more efficient controllers. However, it has higher controller resource utilization during TSSM working hours, higher migration expenses, and requires more controllers to complete. Therefore, based on flow characteristics, we proposed a shortest-feasible open flow path algorithm that optimizes controller selection during the TSSM phase. The new TSSM approach maintains the benefits of TSSM while lowering controller resource consumption and migration expenses. An open network operating system is used to accomplish the suggested method for practical reasons. The testing results show that the proposed method lowers migration costs and controller resource utilization by about 33% compared with the existing system.