Scholarlink

<p>This paper presents a&#xa0;high-performance framework for flexible network load and DDoS attack pattern generation utilizing a&#xa0;P4-programmable Intel Tofino switch. Central to the concept of Sovereign IT, this approach grants operators complete control over both hardware and code. By providing a&#xa0;system open for inspection and adaptation, it offers a&#xa0;transparent and verifiable alternative to proprietary “black-box” solutions.</p><p>We demonstrate the ability to generate line-rate traffic at 100 Gbit/s to evaluate network and end-system resilience against volumetric ICMP, UDP, and TCP SYN floods. Building upon existing research, this work provides a&#xa0;scalable methodology for identifying performance bottlenecks and hardening infrastructure against evolving threats. By offering a&#xa0;re-programmable platform for autonomous stress-testing, this framework enables operators of critical infrastructure to assess network robustness in high-bandwidth environments without vendor dependency. The complete implementation is provided as open-source to support reproducible advances in network security.</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

Using P4 to build a DDoS stresser platform

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年6月15日

查看全文
  • Pascal Bast,
  • Thomas Scheffler

摘要

This paper presents a high-performance framework for flexible network load and DDoS attack pattern generation utilizing a P4-programmable Intel Tofino switch. Central to the concept of Sovereign IT, this approach grants operators complete control over both hardware and code. By providing a system open for inspection and adaptation, it offers a transparent and verifiable alternative to proprietary “black-box” solutions.

We demonstrate the ability to generate line-rate traffic at 100 Gbit/s to evaluate network and end-system resilience against volumetric ICMP, UDP, and TCP SYN floods. Building upon existing research, this work provides a scalable methodology for identifying performance bottlenecks and hardening infrastructure against evolving threats. By offering a re-programmable platform for autonomous stress-testing, this framework enables operators of critical infrastructure to assess network robustness in high-bandwidth environments without vendor dependency. The complete implementation is provided as open-source to support reproducible advances in network security.