Towards resources optimization in deploying service function chains with shared protection

Abstract

Emerging artificial intelligence techniques enable the Internet-connected devices automatically generate massive SFC requests in scenarios with ultra-reliability demands such as unmanned vehicle systems, smart factories. In such scenarios, it is imperative to deploy these huge amounts of SFCs in a fault-tolerant manner. The existing works on fault-tolerant SFC deployment employs the dedicated SFC backup protection with the resources redundancy of at least 100%, which addresses the ratio between the backup and primary resource consumptions. Due to this, the dedicated SFC protection likely prevents these reliable SFCs from being massively deployed. To tackle this challenge, for the first time, this work studies how to enhance the resource utilization by deploying SFCs with shared protection. First, we introduce new concepts of SFC protection set and sharing-risk SFC graph (SCG). Next, we define a new problem called SFCs embedding with SCG-based shared protection. Then, we propose a heuristic algorithm called SCG-based SFC embedding and protecting (SCG-SEP), which is proved to be logarithm-approximate. Extensive simulations show that SCG-SEP outperforms the benchmarks by an average of 12.25% and 43.67%.