Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

QFAM: Mitigating QUIC Handshake Flooding Attacks Through Crypto Challenges

Abdollah Jabbari, Y A Joarder, Benjamin Teyssier, Carol Fung

TL;DR

QUIC handshake flooding can exhaust server CPU due to high amplification during cryptographic handshakes. The paper introduces QFAM, a proactive mitigation that embeds cryptographic challenges in an Enhanced RETRY token and activates a mitigation mode under attack, shifting cost to attackers. The design includes a detailed token structure, server/client procedures, and complexity analysis, showing $O(2^n)$ client work versus $O(1)$ server work, which helps balance resource usage under flood conditions. Experimental results on a modified aioquic setup demonstrate reduced server CPU load and attacker cost with acceptable overhead for legitimate clients, highlighting practical resilience gains for QUIC deployments.

Abstract

QUIC protocol is primarily designed to optimize web performance and security. However, previous research has pointed out that it is vulnerable to handshake flooding attacks. Attackers can send excessive volume of handshaking requests to exhaust the CPU resource of the server, through utilizing the large CPU amplification factor occurred during the handshake process under attack. In this paper, we introduce a novel defense mechanism by introducing the concept of crypto challenges into the handshake protocol. This enhancement involves a proposal of modifying the RETRY token to integrate a cryptographic challenge into it. The client must solve crypto challenges during the handshake process in order to receive a high priority on the server side. By properly choosing the difficulty level of the challenges, the CPU amplification can be reduced, thus the DDoS vulnerability is naturalized. We evaluated the effectiveness of our proposed solution by integrating the crypto challenges into the clients and server of \textit{aioquic}. Our experimental results demonstrate that our solution can effectively balance the resource usage between the attacker and the server during of handshake flooding attacks while maintaining a low overhead for legitimate clients.

QFAM: Mitigating QUIC Handshake Flooding Attacks Through Crypto Challenges

TL;DR

QUIC handshake flooding can exhaust server CPU due to high amplification during cryptographic handshakes. The paper introduces QFAM, a proactive mitigation that embeds cryptographic challenges in an Enhanced RETRY token and activates a mitigation mode under attack, shifting cost to attackers. The design includes a detailed token structure, server/client procedures, and complexity analysis, showing client work versus server work, which helps balance resource usage under flood conditions. Experimental results on a modified aioquic setup demonstrate reduced server CPU load and attacker cost with acceptable overhead for legitimate clients, highlighting practical resilience gains for QUIC deployments.

Abstract

QUIC protocol is primarily designed to optimize web performance and security. However, previous research has pointed out that it is vulnerable to handshake flooding attacks. Attackers can send excessive volume of handshaking requests to exhaust the CPU resource of the server, through utilizing the large CPU amplification factor occurred during the handshake process under attack. In this paper, we introduce a novel defense mechanism by introducing the concept of crypto challenges into the handshake protocol. This enhancement involves a proposal of modifying the RETRY token to integrate a cryptographic challenge into it. The client must solve crypto challenges during the handshake process in order to receive a high priority on the server side. By properly choosing the difficulty level of the challenges, the CPU amplification can be reduced, thus the DDoS vulnerability is naturalized. We evaluated the effectiveness of our proposed solution by integrating the crypto challenges into the clients and server of \textit{aioquic}. Our experimental results demonstrate that our solution can effectively balance the resource usage between the attacker and the server during of handshake flooding attacks while maintaining a low overhead for legitimate clients.

Paper Structure

This paper contains 31 sections, 4 equations, 12 figures, 2 tables, 1 algorithm.

Table of Contents

  1. Introduction
  2. Related Works
  3. Preliminaries
  4. Handshake Mechanism of QUIC
  5. Basics of QUIC Address Validation Mechanism
  6. Attack Strategies of QUIC Handshake Flooding Attacks
  7. QUIC Handshake Flooding Attack in Absence of Address Validation
  8. QUIC Handshake Flooding Attack with Address Validation Mechanism
  9. Amplifying the CPU Amplification Effect Through Computationally Intensive Cryptographic Algorithms
  10. Proposed Mitigation Scheme
  11. Enhanced RETRY Packet Structure with Cryptographic Challenge
  12. Original RFC 9000 RETRY packet structure
  13. Enhanced RETRY Packet Structure
  14. Key Modifications and Enhanced Retry Token
  15. Server's Process for Generating and Sending Enhanced RETRY Packet
  16. ...and 16 more sections

Figures (12)

  • Figure 1: 1-RTT and 0-RTT Handshake Mechanisms in details; (a) 1-RTT handshake without address validation mechanism; (b) 1-RTT with address validation mechanism; (c) 0-RTT facilitating early data transmission
  • Figure 2: Handshake flooding attack without QUIC Address Validation
  • Figure 3: Handshake Flooding Attack With QUIC Address Validation
  • Figure 4: Illustration of QFAM
  • Figure 5: Structure of the QUIC RETRY Packet
  • ...and 7 more figures