Scholarlink

<p>Network congestion control remains a critical challenge in modern networking, where traditional Active Queue Management (AQM) algorithms often struggle to balance low latency, fairness, and packet drops. The Controlled Delay (CoDel) algorithm was introduced to address bufferbloat but still faces limitations in dynamic traffic conditions, including unfair drop distribution and excessive packet loss during congestion. This paper presents Flow-CoDel, an enhanced version of CoDel that introduces coordinated per-flow packet drops to optimize congestion response. Through systematic evaluation, our algorithm demonstrates 31% lower average delay at near-congestion (ρ = 0.9) compared to standard CoDel, along with a drastic reduction in packet drops. Additionally, it ensures fairer packet drops across competing flows. These results suggest that Flow-CoDel offers a promising alternative for scenarios requiring consistent latency and efficient drop policies, such as real-time streaming and VoIP.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Flow-CoDel: per-flow active queue management for balanced delay and drops in bufferbloat mitigation

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年12月8日

查看全文
  • Morteza Seyyed Ashrafi,
  • Ali Ghiasian

摘要

Network congestion control remains a critical challenge in modern networking, where traditional Active Queue Management (AQM) algorithms often struggle to balance low latency, fairness, and packet drops. The Controlled Delay (CoDel) algorithm was introduced to address bufferbloat but still faces limitations in dynamic traffic conditions, including unfair drop distribution and excessive packet loss during congestion. This paper presents Flow-CoDel, an enhanced version of CoDel that introduces coordinated per-flow packet drops to optimize congestion response. Through systematic evaluation, our algorithm demonstrates 31% lower average delay at near-congestion (ρ = 0.9) compared to standard CoDel, along with a drastic reduction in packet drops. Additionally, it ensures fairer packet drops across competing flows. These results suggest that Flow-CoDel offers a promising alternative for scenarios requiring consistent latency and efficient drop policies, such as real-time streaming and VoIP.