Quality-of-service (QoS) performance is an important consideration for real-time and high-priority traffic on internet protocol (IP) networks. Service differentiation can provide a more efficient and customer-oriented internet. The “best-effort” internet models in use today cannot provide guarantees or service differentiation for end-to-end individual and aggregate data flows. Hardware-based models and software-based models do not completely address the total service-enabled solution. We propose a hybrid architecture that combines software and hardware features to handle network traffic with diverse QoS requirements. Since cloud providers leverage IP networks today, the model is based on a systems engineering approach that uses cloud computing technologies. The work describes the conceptual model and the reference model for the hybrid QoS system. Service levels can be defined in terms of absolute or relative guarantees on loss, delay, bandwidth, and burst size. End-to-end characteristics of individual flows are maintained within the aggregate flows of cloud network traffic.
K. R. Owens and S. E. Watkins, "Quality-of-Service Architecture for Cloud Computing Networking," IEEE-HKN The Bridge Magazine, vol. 117, no. 3, pp. 10 - 15, IEEE-HKN, Jan 2021.
Electrical and Computer Engineering
Article - Journal
© 2021 The Authors, All rights reserved.