RSCAT: Towards zero touch congestion control based on actor–critic reinforcement learning and software-defined networking

Diel, Gustavo; Miers, Charles Christian; Pillon, Maurı́cio A.; Koslovski, Guilherme

doi:10.1016/j.jnca.2023.103639

Cited by 6 publications

(2 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Diel et al [10] used a supervised data classification algorithm to detect real-time direct current congestion and categorize it into four main classes. It then used actor-critic reinforcement learning (RL) to find better Transmission Control Protocol (TCP) parameters.…”

Section: Traffic Classificationmentioning

confidence: 99%

Review of Load Balancing Mechanisms in SDN-Based Data Centers

Du,

Cui,

Tang

et al. 2024

JCC

View full text Add to dashboard Cite

With the continuous expansion of the data center network scale, changing network requirements, and increasing pressure on network bandwidth, the traditional network architecture can no longer meet people's needs. The development of software defined networks has brought new opportunities and challenges to future networks. The data and control separation characteristics of SDN improve the performance of the entire network. Researchers have integrated SDN architecture into data centers to improve network resource utilization and performance. This paper first introduces the basic concepts of SDN and data center networks. Then it discusses SDN-based load balancing mechanisms for data centers from different perspectives. Finally, it summarizes and looks forward to the study on SDN-based load balancing mechanisms and its development trend.

show abstract

Section: Traffic Classificationmentioning

confidence: 99%

Review of Load Balancing Mechanisms in SDN-Based Data Centers

Du,

Cui,

Tang

et al. 2024

JCC

View full text Add to dashboard Cite

show abstract

“…Load balancing in the context of data centers has been the focal point of extensive research over the past decade, primarily due to its capacity to enhance the performance and reliability of server operations [4]. Traditional load balancing approaches, though effective during their inception, are becoming increasingly obsolete in the face of the modern internet's explosive growth and the subsequent surge in data traffic [5]. These methods often fall short in predicting and managing traffic demands dynamically, resulting in suboptimal performance [6].…”

Section: Introductionmentioning

confidence: 99%

Load Balancing in DCN Servers Through Software Defined Network Machine Learning

Beissenova,

Zhidebayeva,

Kopzhassarova

et al. 2024

IJACSA

View full text Add to dashboard Cite

In this research paper, we delve into the innovative realm of optimizing load balancing in Data Center Networks (DCNs) by leveraging the capabilities of Software-Defined Networking (SDN) and machine learning algorithms. Traditional DCN architectures face significant challenges in handling unpredictable traffic patterns, leading to bottlenecks, network congestion, and suboptimal utilization of resources. Our study proposes a novel framework that integrates the flexibility and programmability of SDN with the predictive and analytical prowess of machine learning. We employed a multi-layered methodology, initially constructing a virtualized environment to simulate real-world DCN traffic scenarios, followed by the implementation of SDN controllers to instill adaptiveness and programmability. Subsequently, we integrated machine learning models, training them on a substantial dataset encompassing diverse traffic patterns and network conditions. The crux of our approach was the application of these trained models to anticipate network congestion and dynamically adjust traffic flows, ensuring efficient load distribution among servers. A comparative analysis was conducted against prevailing load balancing methods, revealing our model's superiority in terms of latency reduction, enhanced throughput, and improved resource allocation. Furthermore, our research illuminates the potential for machine learning's self-learning mechanism to foresee and adapt to future network states or exigencies, marking a significant advancement from reactive to proactive network management. This convergence of SDN and machine learning, as demonstrated, ushers in a new era of intelligent, scalable, and highly reliable DCNs, demanding further exploration and investment for future-ready data centers.

show abstract