Computing systems are used in various critical domains including aerospace, energy, transportation, healthcare, and commerce. Failures of these systems may lead to catastrophic consequences such as injury, loss of life, damage to equipment, or financial loss.  This course focuses on techniques for designing and analyzing dependable computing systems that can continue to operate correctly in the presence of software and hardware problems. We will learn what can go wrong, how we can predict, prevent, and detect faults/errors, and how we can design systems that can tolerate faults and recover from failures.

Topics:

• Introduction to dependable computing
• Basic terminology, attributes, and evaluation techniques
• Combinatorial and state-space modeling
• Hardware fault tolerance
• Information redundancy
• Software fault tolerance
• Checkpointing and recovery
• Reliable networked systems
• Error detection techniques
• Dependability evaluation techniques
• Safety and Security

Time: Mon/Wed/Fri 9:00AM - 9:50AM
Location: Thornton Hall E304
Office Hours: Wed 10:00AM - 11:00AM - Thornton Hall E314

This is the tentative timeline for the class and subject to change.

The lectures and assignments are based on the following references:

• I. Koren and C. Mani Krishna, Fault-tolerant Systems, 1^st edition, 2007, Morgan Kaufmann. (Read online through UVA Library)
• J. Knight, Fundamentals of Dependable Computing for Software Engineers, 2012, CRC Press. (Read online through UVA Library)
• K. Trivedi, Probability and Statistics with Reliability, Queuing and Computer Science Applications, 2^nd edition, 2001, John Wiley & Sons.
• D. K. Pradhan, Fault Tolerant Computer System Design, 1^st edition, 1996, Prentice-Hall.

• An unpublished textbook by R. K. Iyer, Z. Kalbarczyk, and N. Nakka from the University of Illinois at Urbana-Champaign, who have agreed to let us use a pre-publication copy of the book. The book consists of multiple chapters, each contained in a separate pdf file which will be shared internally with you. Please do not redistribute.

* Each graduate student will select a paper on a special topic (of mutual interest) and will:

• Present the paper to the class (20 minutes)
• Prepare a short homework assignment based on the material in the lecture and the paper
• Grade the homework assignment and provide a solution to the class.
• The paper must be made available to the class at least a week prior to presentation and the homework assignment must be graded within a week of its submission.

** There will be a 10% penalty for late assignments (per school day).

*** The final projects will be performed by the teams consisting of both graduate and undergraduate students. Each team will propose a project on a related topic of interest, define measurable outcomes and deliveries for the project, and present the results as a short paper and a lecture to the class at the end of semester. All the students are required to actively participate in different aspects of the projects.