Statistical Analysis of Link Scheduling on Long Paths

TL;DR

This paper analyzes how different packet scheduling algorithms affect end-to-end network performance on long paths using a network calculus approach, providing new bounds for delays and backlog.

Contribution

It introduces a sharpened method for computing statistical service bounds and derives closed-form expressions for key performance metrics.

Findings

Deterministic bounds are optimal for backlog and burstiness.

Statistical bounds are highly accurate for delay estimates.

New closed-form formulas enable better performance analysis.

Abstract

We study how the choice of packet scheduling algorithms influences end-to-end performance on long network paths. Taking a network calculus approach, we consider both deterministic and statistical performance metrics. A key enabling contribution for our analysis is a significantly sharpened method for computing a statistical bound for the service given to a flow by the network as a whole. For a suitably parsimonious traffic model we develop closed-form expressions for end-to-end delays, backlog, and output burstiness. The deterministic versions of our bounds yield optimal bounds on end-to-end backlog and output burstiness for some schedulers, and are highly accurate for end-to-end delay bounds.