QUIC Flow Latency Measurement Using Handshake and Mid-Flow ACKs

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

The shift to IETF QUIC protocol in internet traffic has rendered conventional methods of measuring and managing latency ineffective, as TCP-based sequence numbers and acknowledgments are not available, making it difficult for network providers to assess network quality and Quality of Experience (QoE).

Innovation Solution

A system and method for measuring and managing latency on computer networks using IETF QUIC by identifying QUIC traffic flows, determining packet parameters, calculating handshake and mid-flow latency, and providing corrective actions when thresholds are exceeded, utilizing a data processing engine, data collection module, packet behavior identification, and state machine to label and correlate packets.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If IETF QUIC protocol is used for transport layer, then network performance and security are improved, but latency measurement capability deteriorates because TCP sequence numbers and acknowledgments are not available

Engineering Contradiction:

Improvenetwork securityVSAvoidlatency measurement

Core Design Contradiction:

ReliabilityVSMeasurement precision

Solution Approach 1:

The patent introduces an intermediary latency measurement mechanism that operates at the application layer within QUIC. Instead of relying on TCP-level acknowledgments, the system uses QUIC's application-layer ACK packets and sequence numbers as mediators to measure latency. The measurement system correlates sent data packets with received ACK packets to calculate round-trip latency, effectively bridging the gap between QUIC's UDP-based transport and latency measurement requirements.

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

The patent shifts the measurement dimension from transport layer (TCP) to application layer (QUIC). By moving latency measurement to the application layer where QUIC operates, the system can utilize QUIC-specific packet structures including application-layer sequence numbers and ACK packets. This dimensional shift allows latency measurement to function effectively within QUIC's architecture despite the absence of traditional TCP measurement mechanisms.

Inventive Principle:

Principle #17Another dimension (Dimensionality change)

2Productivity

If applications migrate from TCP to IETF QUIC, then modern network performance is improved, but conventional latency measurement methods become ineffective

Engineering Contradiction:

Improvenetwork performanceVSAvoidlatency detection

Core Design Contradiction:

ProductivityVSDifficulty of detecting and measuring

Solution Approach 1:

The patent enables QUIC traffic to self-identify and self-measure latency through intrinsic packet characteristics. The system automatically identifies QUIC traffic flows by examining packet structures, extracts relevant packet parameters including sequence numbers and ACK packets, and performs latency measurement without external assistance. This self-service approach allows latency measurement to adapt automatically to QUIC's protocol structure.

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The patent changes the measurement parameters from TCP-specific fields (transport layer sequence numbers and acknowledgments) to QUIC-compatible fields (application layer sequence numbers and ACK packets). By adjusting the measurement parameters to match QUIC's protocol structure, the system maintains latency measurement capability while accommodating the protocol migration from TCP to QUIC.

Inventive Principle:

Principle #35Parameter changes

3Speed

If QUIC protocol is implemented without TCP-style acknowledgments, then protocol efficiency is improved, but latency monitoring capability deteriorates

Engineering Contradiction:

Improveprotocol efficiencyVSAvoidlatency monitoring

Core Design Contradiction:

SpeedVSMeasurement precision

Solution Approach 1:

The patent extracts the essential latency measurement function from TCP's acknowledgment mechanism and reimplements it within QUIC's application layer. Instead of relying on TCP's built-in acknowledgment structure, the system extracts and utilizes QUIC's application-layer ACK packets for latency measurement. This extraction allows latency monitoring to function independently of the transport protocol's specific implementation details.

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The patent implements a dynamic latency measurement system that adapts to QUIC's flexible packet structure. The measurement mechanism dynamically identifies and correlates data packets with their corresponding ACK packets based on sequence numbers and packet characteristics. This dynamic approach allows the system to maintain accurate latency monitoring despite QUIC's variable packet formats and flexible acknowledgment timing.

Inventive Principle:

Principle #15Dynamics

Data Source

PatentUS12506673B2System and method for measuring and managing latency on a computer network

Publication Date: 2025.12.23 SANDVINE CORP

AI summary

A system and a method for measuring latency on a computer network. The method including: identifying a QUIC traffic flow; determining packet parameters associated with the QUIC traffic flow; determining a handshake latency associated with the packet parameters; and monitoring mid-flow latency for any changes in the latency of the QUIC traffic flow. The system for measuring and managing latency on a computer network including: a data processing engine configured to determine a QUIC traffic flow; a data collection module configured to determine packet parameters associated with the QUIC traffic flow; a state machine configured to determine a handshake latency associated with the packet parameters and monitor mid-flow latency for any changes in the latency of the QUIC traffic flow.