Performance of Networked Systems

Course Objective

Students will acquire basic knowledge of:

• Performance issues in communication systems, a complex interplay between communication networks (TCP/IP, UDP), application servers (HTTP, QUIC), services and clients
• Performance models for wireless and mobile network protocols (WiFi, GSM, GPRS, UMTS, HSDPA, 4G, 5G and mobile ad-hoc networks)
• Performance engineering principles and management of traffic in IP networks
• Performance modeling of streaming and elastic traffic
• Quantitative models for planning capacity in communication networks

Students will gain experience in modelling, engineering and planning performance of networked systems, and will learn how to tackle performance problems arising in today's information and communication systems.

Course Content

Over the past few decades, information and communication technology (ICT) has become ubiquitous and globally interconnected. As a consequence, our ICT systems are expected to process huge amounts of digital information, which puts a tremendous burden on our ICT

infrastructure. At the same time, our modern society has become largely dependent on the well-functioning of our ICT systems: large-scale system failures and perceivable Quality of Service (QoS) degradation may completely disrupt our daily lives and have huge impact on our economy. These developments raise the need for understanding the nature and source of performance problems, and for finding effective ways to make our networks robust and reliable.

Motivated by this, the course will address performance-related challenges and solution methods

in communication networks. To this end, we focus on performance modeling, traffic management, QoS provisioning and capacity planning for both wired and wireless/mobile communication networks (such as WiFi, 2G, 3G, 4G, 5G networks, and mobile ad-hoc networks). We address questions like:

• How do wired and wireless communication network technologies work?
• How can we model their performance?
• How can we manage and control traffic flows and provision QoS in IP-based networks?
• How can we deal with the omnipresence of uncertainty in virtually all aspects of communication networks?
• What techniques are available to model, analyze and optimize performance?

Additional Information Teaching Methods

Classroom lectures and one or two practical homework assignments.

Method of Assessment

The assessment will be based on both homework assignments (total 40%) and a written exam (60%). Homework assignments can be made in groups of one or two students.

Literature

Textbook, supplemented with a reader on Stochastic Performance Modelling.

High Performance Browser Networking, Ilya Grigorik, O-Reilly, 2013.

Additional Information Target Audience

Explanation Canvas

Recommended background knowledge

The students should have basic knowledge of computer networks.