Attaining High Bandwidth In Cloud Computing Through SDN-enabled Multi-tree Multicast

TL;DR

This paper presents a novel SDN-enabled multi-tree multicast scheme that significantly improves bandwidth utilization in cloud computing environments by efficiently distributing traffic across multiple routing trees.

Contribution

It introduces a new multicast protocol leveraging SDN for high bandwidth in cloud computing, addressing complex routing challenges with a multi-tree approach.

Findings

Enhanced bandwidth utilization through multi-tree multicast

Prototype demonstrates performance improvements over traditional methods

Effective traffic load distribution across cloud resources

Abstract

Achieving high bandwidth utilization in cloud computing is essential for better network performance. However, it is difficult to attain high bandwidth utilization in cloud computing due to the complex and distributed natures of cloud computing resources. Recently, a growing demand for multicast transmission is perceived in cloud computing, due to the explosive growth of multi-point communication applications, such as video conferencing, online gaming, etc. Nonetheless, the inherent complexity in multicast routing in cloud computing, existing multicast plans failed to produce effective and efficient protocol schemes, which limits the application of multicast communication on the Internet. In this paper, a technique is proposed in how the newly developed network architecture, Software Defined Network (SDN), can promote the design of the multicast protocol and improve the performance of the multicast transmission in the cloud computing. The approach is to use the SDN-cloud Computing-enabled multicast communication scheme with ultra-high bandwidth utilization. The bandwidth utilization is enhanced by measuring various routing trees for each multicast transmission session and distributing the traffic load over all available routes in the cloud computing resources. The SDN is utilized to tackle with various design hurdles in the cloud computing, including both the current ones with the conventional multicast pattern and the newly emerged ones with multi-tree multicast. The prototype implementation and experiments demonstrate the performance enhancement of the proposed approach in the cloud computing in compared to conventional single-tree multicast designs.