7.5 ECTS credits, period 2
Course responsibles: Prof. Rolf Stadler, KTH/EES, Prof. Seif Haridi, KTH/ICT

Level: The course will be offered to students on two different levels:

· Advanced undergraduate course for Master students. The target audience includes third and fourth-year students from the D, E, I, IT programs and the KTH International Master programs.

· PhD-level course for PhD Students. The target audience includes doctoral students in EE, CS, ICT, and the ACCESS Graduate School.

· While the lectures that introduce the topic domains for the projects are the same for both levels, the projects themselves are different. The projects for the PhD-level version are more demanding on the analytical level, while they are comparable to the undergraduate version regarding the required programming skills and effort.

Topic area: The course focuses on examples of distributed algorithms that enable key functions in emerging technologies, such as p2p services, networked control systems and network management for next-generation networks.

The algorithms covered in the course are usually not taught in introductory classes on algorithms and data structures or distributed algorithms.

Example of topics, which can vary from year-to-year:

· algorithms for data aggregation in networks and distributed systems

· algorithms for distributed search

· algorithms for publish-subscribe systems

· algorithms for distributed trust and reputation schemes

Course format: The course is organized around two project assignments, each of them focusing on a specific topic. The project modules are independent of each other. Each topic will be introduced based on papers from the research literature. The project assignment has a programming part (Java), and students must deliver a report on the project results.

Prerequisites: Students must have taken introductory courses in networking, programming and algorithms. Programming experience in Java is essential. Knowledge in distributed algorithms is helpful, but not mandatory.

The course language is English.

Projects and Material for 2011

Project 1: Distributed Resource Allocation
Gossip-based algorithms for resource allocation in a data-center setting. The project will include a simulation study.


Lectures: Dr. Fetahi Wuhib (slides) 


Project:   Dr. Fetahi Wuhib, Rerngvit Yanggratoke (project description ) (additional materials)


Project 2: Leader Selection and Broadcast in Overlay Networks

Exploiting overlay network topology to build efficient leader selection and broadcast algorithms. Preferential selection algorithms for gossip-based overlay contstruction.   The project will include a simulation study.

Lectures: Dr. Jim Dowling (slides)

Project:    Dr. Jim Dowling (project description) (project slides)

Course organization

Week 1	Project 1	Topic introduction and project assignment
Week 2	Project 1	Project work
Week 3	Project 1	Project Work
Week 4	Project 1	Report submission and grading
Week 5	Project 2	Topic introduction and project assignment
Week 6	Project 2	Project work
Week 7	Project 2	Project Work
Week 8	Project 2	Report submission and grading

Schedule of Lectures and Project Deadlines for 2011

1	Project 1:  Distributed Resource Allocation	Mon Oct 24 10:15-12:00	Q26 KTH Stockholm Campus
2	Project 1:  Distributed Resource Allocation	Wed Oct 26 10:15-12:00	Q21 KTH Stockholm Campus
3	Project 1:  Distributed Resource Allocation	Tue Nov 1   10:15-12:00	Q26 KTH Stockholm Campus
4	Project 2: Leader Selection and Broadcast in Overlay Networks	Mon Nov 21 10:15-12:00	Q26 KTH Stockholm Campus
5	Project 2: Leader Selection and Broadcast in Overlay Networks	Thu Nov 24 10:15-12:00	Q26 KTH Stockholm Campus
6	Project 2: Leader Selection and Broadcast in Overlay Networks	Thu Dec 1   10:15-12:00	Q26 KTH Stockholm Campus
7	Project Review	2nd week Jan; date to be determined	Room to be determined

Submission Deadline for Project 1: Fri Nov 11
Submission Deadline for Project 2: Fri Dec 9



Comments

· For both projects, lectures are scheduled for topic introduction and project assignment.

· Students work on projects either individually or in teams of two.

· During the project work phase, students work on their own. They are supported by a bulletin board (Bilda).

· Software and documentation needed for the project will be available through the course web site.

· After handing in the report, there is a short interview with each student or team, in which the project results and the report are discussed.

· Grading

· To pass the course, students must pass both projects.

· For each project, a student receives a grade on the scale F, FX, E, D, C, B, A.
Students who receive an F for the project must repeat the project.

· The final grade is computed as the average of the two project grades, rounded if needed.

Last modified: Nov 17, 2011