ATL: A Link Failure Recovery Method with Fast Recovery Speed, Low Interruption Rate, and Small TCAM Consumption in SDN

Based on the architecture of the Software Defined Network (SDN), we argue that SDN together with Internet of Things (IoT) applications can enhance the control and management of IoT in terms of flexibility and intelligence. However, with the rapid development of SDN, link fault tolerance has become a crucial challenge in ensuring network reliability and stability. In this field, the current solutions mainly consist of proactive and reactive schemes, but there are still limitations and challenges. The proactive solutions often waste significant TCAM resources, while the reactive solutions face longer fault recovery time. In order to solve the shortcoming of both solutions, and to address the issues of traffic interruption caused by insufficient link bandwidth after link failures, this paper proposes a method called Adapt To Link (ATL) which combines the greedy algorithm with Pro-VLAN technology (a proactive method which calculates a backup path and assigns a VLAN ID for each link before link failures), aiming to achieve efficient fault-tolerant recovery for link failures in the SDN data plane. In this algorithm, we utilize the Pro-VLAN method and pre-compute multiple backup paths for each link before a failure occurs. When a link failure occurs, we can swiftly switch the affected traffic to the backup paths automatically without the controller, ensuring continuous flow of network traffic. Additionally, we introduce a greedy algorithm to satisfy the requirement of elephant flows, thereby improving network resource utilization and performance. By combining these techniques and methods, our algorithm exhibits the characteristics of low interruption rate, minimal resource consumption, and high efficiency. The experimental results demonstrate that under high load conditions, the ATL scheme reduces the interruption rate by 72.4% compared to the proactive approach and by 54.9% compared to the reactive approach, respectively. The number of additional flow rules in ATL is only 21.95% of the proactive scheme, and the recovery time after a failure is only 21.72% of the reactive scheme. As a result, the ATL algorithm enables rapid and reliable network recovery in the event of link failures, and optimizes the utilization of network resources simultaneously, validating its feasibility and effectiveness.