Synthesizing Economic Incentives and ML for Multi-Tenant Fairness in Data Centers

Papers like Han et al. (2022) and Wang et al. (2025) focus on fairness and congestion control in multi-flow environments, but largely through algorithmic tweaks or active queue management. Meanwhile, ML-based traffic prediction (Xu, 2024) offers fine-grained forecasts of traffic surges. This idea proposes an architecture where tenants bid for bandwidth or low-latency “slots” based on their predicted future needs (as forecasted by ML). The network allocates resources dynamically via auctions or spot pricing, ensuring that fairness is priced and enforced even under contention. This is a radical shift from “fairness by algorithm” to “fairness by market,” enhanced with predictive analytics. Such a system could be especially powerful for cloud providers or multi-tenant edge clouds, where traditional fairness mechanisms struggle with bursty or unpredictable workloads.

References:

1. A QoS-Based Fairness-Aware BBR Congestion Control Algorithm Using QUIC. Yi Han, Mengjie Zuo, Huijun Yuan, Yi Zhong, Zhenhui Yuan, Ting Bi (2022). Wireless Communications and Mobile Computing.
2. DCTCP-FQ: Enhancing Fairness and Convergence Time in Data Center Congestion Control. Haoyu Wang, Xiaoqian Zhang, Allen Yang, Bo Sheng (2025). International Conference on Computing, Networking and Communications.
3. Machine Learning Based Traffic Prediction and Congestion Control Algorithms in Software Defined Networks. Yanying Xu (2024). 2024 International Conference on Interactive Intelligent Systems and Techniques (IIST).