Design and evaluation of a dual dynamic adaptive reservation approach in media production networks

详细信息    查看全文

• 作者：Maryam Barshan ; ^{maryam.barshan@intec.ugent.be} ; Hendrik Moens ; Bruno Volckaert ; Filip De Turck
• 关键词：Advance bandwidth reservation ; Media production network ; Runtime adaptation ; Network failure handling
• 刊名：Journal of Network and Computer Applications
• 出版年：2017
• 出版时间：15 February 2017
• 年：2017
• 卷：80
• 期：Complete
• 页码：109-122
• 全文大小：6691 K
• 卷排序：80

文摘

Media production networks require an efficient collaboration between geographically distributed actors and offer predictable workloads, making it possible to exploit this predictability and use advance bandwidth reservation services to achieve greater bandwidth utilization and service guarantees. To offer reliable reservations, the incorporation of fault-tolerance related features in bandwidth reservation strategies is a necessity, although this imposes a waste of capacity and extra performance overhead. As a first provisional stage to offer a robust reservation system, deploying protection mechanisms ensures that the reservations remain valid when the system is in operation. To maximally utilize the network, and to ensure there is a quick response in a dynamic network environment, constant monitoring and optimization is needed. In this article, we propose an efficient dual optimization approach consisting of two consecutive processes. First, a schedule is produced by a resilient advance reservation algorithm. Then, the generated schedule is continually updated over time using a runtime adaptation approach in order to be capable of dynamically adapting the network to changing conditions and mitigating the side effects of provisioned reliability. This step uses the interconnecting network links’ leftover capacity, resulting in an increased performance both in steady and unsteady network conditions. Our evaluations show that in failure-prone environments, the proposed approach leads to significant increase in the success rate of admitted requests, up to 6.77 times, compared to the resilient advance reservation algorithms.