Learning outcomes

Managing data in a cybersecurity context: networking, database, machine learning. Understanding the link between performance and security in various contexts.

 

Goals

The student is expected to

- Better understand the context and reasons that lead to existing attacks against service availability, and existing defenses.

- Better understand how to handle data. How to safely expose them, and what limit existing techniques have.

- Undestanding the difference between machine learning applied to computer security versus computer security applied to machine learning.

- Gaining knowledge on Pervasive Monitoring (Mass surveillance), and how to defend against it.

- Understanding the intrication between the performance and security of the Web.
 

Content

The course covers and studies the performance and security of systems in which
a lot of data is managed. It ranges from company networks expecting to
establish a secure architecture handling Gb/s of traffic, to safely exposing
data online while ensuring modern security definitions for privacy. The course
also contains alternative directions to the existing Internet architecture, and
compare benefits and issues. Among them, Onion services, decoy routing, GNUnet,
certificate transparency, and SCION will be covered in diverse degree of details.

We will also explore machine learning applied to detection intrusion, and cover
the pitfalls.

Finally, we will explore how to secure modern communications, develop how
End-to-End encryption works, what guarantees it offers, and cover various
messaging systems with subtle differences leading to major security
discrepencies.

Assessment method

The course evaluation will be based on the project, on participation in the class, and on the quality of the material given in the inverted classes.

Language of instruction

Anglais