Supporting Future Dynamic Traffic Environments Through Concept Drift Adaptation
Xiangyu Yu, Lihua Ruan, Jamie Evans, Elaine Wong · 2024
This paper introduces an enhanced concept drift detection and adaptation (CDDA) framework to accommodate diverse application traffic patterns within heterogeneous network environments. Results show that the enhanced CDDA can achieve the predictive dynamic bandwidth allocation scheme yields ∼ 50% latency improvement in dynamic traffic environments compared with baseline techniques.