Cocoa: Congestion Control Aware Queuing

TL;DR

Cocoa is a new AQM scheduler that dynamically adjusts buffer sizes based on flow needs, improving throughput and reducing overbuffering without endpoint participation, inspired by model-based congestion control principles.

Contribution

This paper introduces Cocoa, an AQM scheduler that adapts buffer sizes per flow without endpoint input, enhancing throughput and avoiding overbuffering.

Findings

Cocoa interacts well with various congestion control algorithms.

Cocoa significantly increases throughput compared to fair CoDel.

Cocoa avoids overbuffering while maintaining high throughput.

Abstract

Recent model-based congestion control algorithms such as BBR use repeated measurements at the endpoint to build a model of the network connection and use it to achieve optimal throughput with low queuing delay. Conversely, applying this model-based approach to Active Queue Management (AQM) has so far received less attention. We propose the new AQM scheduler cocoa based on fair queuing, which adapts the buffer size depending on the needs of each flow without requiring active participation from the endpoint. We implement this scheduler for the Linux kernel and show that it interacts well with the most common congestion control algorithms and can significantly increase throughput compared to fair CoDel while avoiding overbuffering.