Networks and Distributed Systems
Course: Networks and Distributed Systems (320422)
Instructor: Jürgen Schönwälder
Tuesday 11:15 - 12:30 Seminar-Room R1
Thursday 09:45 - 11:00 Seminar-Room R1
Start: February 6th, 2007

This graduate seminar course is project oriented. Students choose a topic for a project and work on that project throughout the semester. In addition, the state of the art in the choosen topics will be explored in a series of paper presentations. Each student will present one or several research or survey papers from the literature.

Course Materials:
  • See below...
Date Speaker Topic
2007-02-06 Introduction
2007-02-13 Assignment of Projects
2007-02-20 Todor Stoyanov Ensemble Case-Based Reasoning
Mattia Parigiani Matrices, Vector Spaces, and Information Retrieval
2007-02-27 Vladislav Marinov Promise Theory
Georgi Chulkov Google and PageRank
2007-03-06 No Meeting
2007-03-13 Ioana Varsandan Routing indices for peer-to-peer systems
Status Reports
2007-03-20 No Meeting
2007-03-29 Nevatia Yashodhan AODV-RSS Routing in Ad Hoc Networks
Todor Stoyanov Learning Collaboration Strategies for Committees
Status Reports
2007-04-17 Vladislav Marinov Improving Anomaly Detection Event Analysis Using the EventRank Algorithm
Status Reports
2007-04-19 Ha Tran PhD Proposal
Mattia Parigiani Improving Case Representation and Case Base Maintenance in Recommender Agents
2007-04-24 Georgi Chulkov Wrapper Induction
Ioana Varsandan Semantic Search in Peer-to-peer Networks
2007-04-26 Yashodhan Nevatia Interference-Aware Fair Rate Control
Status Reports
2007-05-08 Status Reports
2007-05-15 Vladislav Marinov Integration of cfengine and scli
Todor Stoyanov Distributed Case-Based Reasoning
Yashodhan Nevatia Ad-Hoc Routing for USARSim
Ioana Varsandan A Gnutella Peer to Peer Network
Mattia Parigiani CBR: Representation and Retrieval in IR Systems
Georgi Chulkov Buglook: A Semantic Search Engine for Bug Reports

While most of the time will be spend on project work, I expect that every student presents two papers during the seminar and I also expect active participation during the discussion of presentations. In concrete terms, I plan the following grading scheme:

  • Paper presentations: 30% (2 papers, each 15%)
  • Project presentation: 10%
  • Project report: 40%
  • Project implementation: 20%

Quite some emphasis is on the project (60%). So choosing the right project is of course crucial. I will assign papers that are related to the projects students will be working on.

The project report should be structured like a normal research paper with an abstract, an introduction and motivation and problem statement, the description of the work carried out, a section on related work, and conclusions. The length should be about 10 pages in the IEEE two-column journal paper format. (Note that this format is pretty space efficient.)

The deadline for projects reports is May 18th. All presentations (15 minutes each) will be done on May 15th. On May 15th, we will also schedule individual appointments for demonstrations.


Project topics (but other proposals are possible):

  1. cfengine / scli integration (Vladislav Marinov)

    The cfengine program is used to configure Unix machines. It is a policy engine which reads a file describing a policy how a system should be configured and tries to enforce this policy. The scli program is an SNMP management program which has among other features the ability to configure VLANs. The goal of the integration project is to configure VLANs from a cfengine policy configuration.

    The cfengine program in addition has a daemon which collects data and does timeseries analysis on it. Another goal is to integrate scli to access data from SNMP agents.


    • M. Burgess, S. Fagernes: Pervasive computing management: A model of network policy with local autonomy. IEEE Transactions on Networking, (submitted).
    • M. Burgess, S. Fagernes: Promise theory - a model of autonomous objects for pervasive computing and swarms. Networking and Services, ICNS '06, (submitted), 2006.
    • M. Burgess: An approach to policy based on autonomy and voluntary cooperation. Lecture Notes on Computer Science, Proc. DSOM 2005, LNCS 3775, 2005.
    • M. Burgess: Probabilistic Anomaly Detection in Distributed Computer Networks, Science of Computer Prgramming, 60(1), 2006
    • K. Begnum, M. Burgess: Improving Anomaly Detection Event Analysis Using the EventRank Algorithm, Proc. AIMS 2007, LNCS, 2007.

  2. trouble ticket search engine (Georgi Chulkov)

    Trouble ticket systems or bug trackers are used throughout the open source world to track bug reports or operational problems. Such systems support a formalized format and free floating text.

    Many people use google these days to resolve problems with software systems and often helpful information is found in trouble ticket systems and bug trackers. However, a generic search engine like google does not really take advantage of the structure information present. The goal of this project is to design and build a search engine which crawls bug trackers, taking advantage of the structured format to create a "semantically enhanced" search frontend.


    • S. Brin, L. Page: The Anatomy of a Large-Scale Hypertextual Web Search Engine. In Proceedings of the Seventh International World Wide Web Conference, pp. 107-117, 1998.
    • L. Page, S. Brin, R. Motwani, T. Winograd: The PageRank Citation Ranking: Bringing Order to the Web, 1999.
    • N. Kushmerick: Wrapper Induction: Efficiency and Expressiveness. Artificial Intelligence 118(2):15-68, Elsevier, 2000.

  3. peer-to-peer network for distributed case-based reasoning (Ioana Varsandan)

    The implementation of P2P systems is an important step to investigate further problems on the distributed environment. The goal of this work is to develop two P2P systems: unstructured and structured from scratch or by using available packages. The system is supposed to run on the PlanetLab infrastructure.


    • A. Crespo, H. Garcia-Molina: Routing indices for peer-to-peer systems. In Proc. of the 28th Conference on Distributed Computing Systems, July 2002.
    • C. Tang and Z. Xu and S. Dwarkadas: Peer-to-peer information retrieval using self-organizing semantic overlay network. Proc. Conference on Applications, Technologies, Architectures, and Protocols for Computer Communications (SIGCOMM '03), 2003
    • M. Li and W. C. Lee and A. Sivasubramaniam: Semantic Small World: An Overlay Network for Peer-to-Peer Search. Proc. 12th IEEE International Conference on Network Protocols (ICNP '04)

  4. case representation and retrieval (Mattia Parigiani)

    Case-based reasoning is a problem-solving methodology using experience. CBR systems are used for problem classicification and problem resolution. A short explanation of how a CBR system works: it takes a problem (input), compares to similar problems and solutions (cases) solved previously, adapts retrieved solutions to the new circumstance of the problem, confirms the new solutions and updates changes to a storage (case base). Three important components in a CBR system: case base, case retrieval, case reasoning.

    This work studies the case retrieval related to case representation and evaluation. There has been a representation method which employs feature-value and semantic vectors to present a case. The corresponding evaluation functions are also available in research papers. The goal of this work is to implement the case retrieval component and to measure its performance based on the bibliography data set.


    • M. W. Berry, Z. Drmac, E. R. Jessup: Matrices, Vector Spaces, and Information Retrieval, SIAM Rev. 41(2): 335--362, 1999
    • M. Montaner and B. Lopez and J. Lluis de la Rosa: Improving Case Representation and Case Base Maintenance in Recommender Agents, Proc. 6th European Conference on Advances in Case-Based Reasoning (ECCBR '02), 234--248, Springer-Verlag, 2002

  5. distributed case-based reasoning (Todor Stoyanov)

    Distributed case-based reasoning brings case-based reasoning method into distributed environment. Not only many CBR systems operate simultaneously, but multiple case bases involve in distributed CBR. There are roughly two trends on how CBR systems work: 1. CBR systems work indenpendently on their case bases to propose solutions; 2. CBR systems work indenpendently on their case bases, but collaboratively propose solutions. The former puts more weights on case retrieval, while the latter exploits collaborative reasoning capabilities.

    This study is about the collaborative capability of CBR systems to propose 'ensemble' solutions. Considering the scenario that a P2P peer want to rate (validate) its solutions among solutions from other peers. How does it work? The task is to find out a simple framework for this (with a demo).


    • E. Plaza and S. Ontanon: Ensemble Case-Based Reasoning: Collaboration Policies for Multiagent Cooperative CBR, ICCBR, 437-451, Springer Verlag, 2001
    • E. Plaza and S. Ontanon: Learning collaboration strategies for committees of learning agent, Autonomous Agents and Multi-Agent Systems, Springer, 2006

  6. ad-hoc routing for the usam-sim simulator (Yashodhan Nevatia)

    The usar-sim robot simulator features a wireless propagation model that calculates, given the position of two robots, whether two robots share a wireless link and can communicate. The goal of this project is to add wireless ad-hoc routing to this simulator and in particular to select an ad-hoc routing algorithm that performs well in the virtual environment. Note that the usar-sim simulator is able to take obstacles into account when computing signal strength and thus many existing simulation that are typically obtained using models without any obstacles can't be applied directly to select a good ad-hoc routing algorithm.


    • R.S. Chang, S.J. Leu. Long-lived Path Routing With Received Signal Strength for Ad Hoc Networks, 1st International Symposium on Wireless Pervasive Computing, 2006.
    • S. Rangwala, R. Gummadi, R. Govindan, K. Psounis: Interference-Aware Fair Rate Control in Wireless Sensor Networks, SIGCOMM 2006, Pisa, 2006.