PCC: Re-architecting Congestion Control for Consistent High Performance
arXiv:1409.7092
Abstract
TCP and its variants have suffered from surprisingly poor performance for decades. We argue the TCP family has little hope to achieve consistent high performance due to a fundamental architectural deficiency: hardwiring packet-level events to control responses without understanding the real performance result of its actions. We propose Performance-oriented Congestion Control (PCC), a new congestion control architecture in which each sender continuously observes the connection between its actions and empirically experienced performance, enabling it to consistently adopt actions that result in high performance. We prove that PCC converges to a stable and fair equilibrium. Across many real-world and challenging environments, PCC shows consistent and often 10x performance improvement, with better fairness and stability than TCP. PCC requires no router hardware support or new packet format.
Cited by in corpus (8)
- Machine Learning for Networking: Workflow, Advances and Opportunities
- A Machine Learning Approach to Routing
- NG-Scope: Fine-Grained Telemetry for NextG Cellular Networks
- Fifty Shades of Congestion Control: A Performance and Interactions Evaluation
- Statistical Learning Based Congestion Control for Real-time Video Communication
- Adaptive ABAC Policy Learning: A Reinforcement Learning Approach
- The CoCo-Beholder: Enabling Comprehensive Evaluation of Congestion Control Algorithms
- Sayer: Using Implicit Feedback to Optimize System Policies