AI-Optimized Adaptive Routing for High-Performance Data Centers: A Predictive Framework for Dynamic Network Optimization
Vishwanath Hiremath
Abstract
The growing need for dependable, scalable, and lightning-fast data transport in today's data centres necessitates effective network management solutions. This paper presents an AI-optimized adaptive routing framework, aiming for dynamically steer network traffic in high-performance data centers towards efficient utilization. The framework uses machine learning algorithms, particularly predictive models, to analyze real-time traffic data, predict potential congestion points, and dynamically adjust routing paths to optimize data transfer efficiency. Their proposed implementation integrates with the already deployed data center network infrastructures, facilitating real-time, automated decision-making that reduces latency, improves throughput, and ensures resource optimization. The proposed AI-based method shows significant improvements over the traditional static routing algorithm in terms of network performance, adaptability against fluctuations and efficiency, according to the experimental results. The paper also addresses the environmental impact, scalability, and sustainability of AI-based routing solutions for large-scale data center environments. Lastly, the paper summarizes future directions in research, specifically with the use of deep learning techniques to achieve improved accuracy for routing prediction and lower operational costs.