Abstract:
Middleboxes play a key role in network function virtualization and the backups are necessary to ensure the reliable operation. A shared backup model based on queuing theo...Show MoreMetadata
Abstract:
Middleboxes play a key role in network function virtualization and the backups are necessary to ensure the reliable operation. A shared backup model based on queuing theory and unavailability takes into account various factors such as failure, repair, recovery, and workload. This model uses a simulated annealing (SA) algorithm to find the backup assignment that minimizes the maximum unavailability among considered middleboxes by multiple iterations. The SA algorithm that computes the exact maximum unavailability (EU) in each iteration, which is called SA-EU, can find an assignment with lower maximum unavailability but has a longer computation time; while the SA algorithm adopting performance bounds (PB) instead of EU at each iteration, which is called SA-PB, has a shorter computation time but eventually finds an assignment with a higher maximum unavailability. This paper proposes SA-MIX, an SA algorithm that uses both EU and PB methods to evaluate the performance of assignments in an exchangeable manner in different stages. This allows the algorithm to take the advantages of both evaluation. Numerical results show that the proposed SA-MIX provides 87% shorter computation time than SA-EU, while its maximum unavailability is comparable to that of SA-EU. In addition, compared with SA-PB, the maximum unavailability of SA-MIX decreases by 37%; the computation time is increased by 3.41 [s].
Published in: 2022 13th International Conference on Information and Communication Technology Convergence (ICTC)
Date of Conference: 19-21 October 2022
Date Added to IEEE Xplore: 25 November 2022
ISBN Information: