BioWarp: An SDN Failure Recovery Scheme Based on Bio-Mimetic Optimization and Weighted-Cost Multi-Path Routing

The escalating complexity of networks due to advancements in cloud computing, IoT, and data centers necessitates a shift towards more adaptable and scalable network architectures. Software Defined Networking (SDN) has risen as a prominent solution, offering a separation of control and data planes to bolster flexibility and manageability. However, network failures remain an inherent risk, undermining the reliability of both traditional and SDN environments. This paper presents BioWarp, a novel SDN failure recovery scheme that integrates bio-mimetic optimization with a Weighted-Cost Multi-Path (WCMP) routing approach. BioWarp confronts the unique challenges of SDN, such as TCAM overflow and controller computation stress, by proposing a hybrid strategy that merges the strengths of proactive and reactive recovery methods. We introduce a flow label based aggregation method to alleviate the burden on flow tables and controller resources. The Coati Optimization Algorithm (COA), inspired by the behaviors of coatis, is adapted to optimize weight update in real-time, facilitating rapid and efficient recovery from network faults. Simulation experiments are conducted to substantiate the scheme’s efficacy, demonstrating improved recovery times and resource management without compromising network stability.