Network Calculus Results for TSN: An Introduction

TL;DR

This paper surveys the application of Network Calculus to Time-Sensitive Networking, providing a unified overview of analysis methods, results, and assumptions to facilitate understanding and future research in real-time Ethernet networks.

Contribution

It offers a comprehensive survey of Network Calculus approaches for TSN, introduces a common notation, and suggests an improved model for output analysis at sending devices.

Findings

Unified notation for NC results in TSN

Comparison of different NC analysis approaches

Identification of key assumptions and dependencies

Abstract

Time-Sensitive Networking (TSN) is a set of standards that enables the industry to provide real-time guarantees for time-critical communications with Ethernet hardware. TSN supports various queuing and scheduling mechanisms and allows the integration of multiple traffic types in a single network. Network Calculus (NC) can be used to calculate upper bounds for latencies and buffer sizes within these networks, for example, for safety or real-time traffic. We explain the relevance of NC for TSN-based computer communications and potential areas of application. Different NC analysis approaches have been published to examine different parts of TSN and this paper provides a survey of these publications and presents their main results, dependencies, and differences. We present a consistent presentation of the most important results and suggest an improvement to model the output of sending end-devices. To ease access to the current research status, we introduce a common notation to show how all results depend on each other and also identify common assumptions. Thus, we offer a comprehensive overview of NC for industrial networks and identify possible areas for future work.