Rapid restoration techniques for software-defined networks

The rapidly increasing number of critical business applications and secured data transmission delivered over modern day communication networks gives rise to the demand for a high degree of network robustness and reliability. In recent years, the paradigm of software-defined networking (SDN) has emerged as a promising approach for defining network architectures that are capable of being adaptable and robust. The separation of data and control planes means that the network controller is responsible for the network configuration. In the case of network failure events, the network may need to be reconfigured quickly rather than optimally especially when the network operates in a harsh environment where the failure rate is high. We propose SDN recovery methods that use information about the network topology to expedite network restoration when a link failure has occurred. Our SDN recovery methods find alternative paths which have low operation and installation costs by detecting the underlying community structure of the network topology. We conduct extensive simulations to test the SDN recovery methods using well-known simulation tools. Our simulation results provide evidence that the proposed methods improve the performance of finding quick alternative paths amongst a set of possible ones.