Nodes Research Methodologies

582701
1-2
Networking and Services
Advanced studies
Year Semester Date Period Language In charge
2013 spring 11.03-26.04. 4-4 English Lea Kutvonen

Lectures

Time Room Lecturer Date
Thu 10-12 D122 Lea Kutvonen 11.03.2013-26.04.2013

Information for international students

Overview

The NODES research group coordinates the various research groups at the Department of Computer Science on the area of Networking and services. The NODES group interestes range from the effective organisation of Internet protocols and wireless communication to the new challenges of globally interoperating business services. The focus areas are networked systems and their enablers: interoperability (e.g., service and software platforms, interoperability management, contracting, trust management, privacy), mobility (technology and location independence, wireless computing), information networks, service networks, context awareness and ubiquitous computing. The group combines the departmental tradition of wireless and mobile computing to new research themes. The focus of the group has been extending for the last years from platform protocols to solving application layer challenges.

The NODES group also coordinates the BSc, MSc and PhD level education on this area. At this front, the NODES group mission is to educate experts and strategic leaders  for the design and realisation of new,  global platforms or infrastructures for example in collaboration and interoperability, mobility,   information networks / service networks, and  context awareness.
 
The professional skills of graduating students from any university program include those of research methodologies in their selected topic area. Students who become researchers must be able to conduct their work according to accepted norms on their field, and potentially, also further develop these methodologies. Students who seek into positions in companies or governmental institutions, or create their own companies, must understand the quality of the research results they are applying in their work.
 
The common thing in scientific methods is the constant questioning of previous hypothesis and improvement of already reached solutions. Beyond that, the research methods differ in terms of the artefacts produced and the techniques applied in testing their qualities and adoptability. 
 
This course is structured as a series of independent talks from the principal investigators or their representatives from the NODES group member groups. Each talk will introduce a specific research methodology and exemplify how that specific group has applied the methodology in their research work. In addition to that, some classical texts about research methodologies with their strengths and weaknesses will be read, and as a concluding exercise, some group discussions on pontential research challenges and methodology selection are conducted.
 
 

 

Schedule (subject to changes!)

 
  • Thu 14.3. 2013 Opening and overview
  • Thu 21.3. 2013  
    • We complete the overviews of the remaining research groups​
    • N. Asokan: Methodologies for designing secure systems
    • Exercise of the week (learning diary entry, to be sent to Lea by email or through this weeks google form here) is expressed as a quiz on slide 24 in Asokan's slides above. Exercise responses / cumulative learning diary entries are to be submitted each week separately, by the next meeting start time (as some solutions may be opened at the next session). 
    • Hint: El Gamal is not CCA-secure.  Try to construct an adversary A who modifies c and c' to get a c'' in such a way that when he gets back  its decryption m'' he can figure out if c was an encryption of m_0 or m_1
  • Thu 28.3. 2013  - Easter break
    • .
  • Thu 4.4. 2013  
    • ​Toni Ruokolainen: Use of formal methods in distributed systems construction and analysis
    • Teemu Kerola: Queuing networks
    • Assignment for the cumulative learning diary entries: 
    • a) For what kind of research questions each of these techniques is applicable?
    • b) What prequisite skills you need to obtain to successfully utilise these techniques? (e.g., parts of statistics ...)
    • Please send any pending learning diaries by email to Lea as most of you have already done so.
  • Thu 11.4. 2013
  • Thu 18.4. 2013 
    • Ilkka Kosunen, Giulio Jacucci: User and Interaction Research Methods (Ethnographic Field trials, qualitative and quantitative methods, Experiment Design in HCI)
    • Introduction of the discussion group topics for the course roundup and group formation
      • To do preferrably today:  Go to the shared, openly accessible google documents page, a) sign up with one of the groups for the next session b) add your suggestions for additional discussion topics, or detailed questions for the presenters on the course
    • Learning diary entry question for this week:
      • ​What is the difference between qualitative and quantitative research? Why this differentiation is relevant in this context? What category the presented research belongs to why?
      • will sqeeze the slides here after the lecture ...
    • ​Opened the group discussions (groups 1,2,4 were active)
  • Thu 25.4. 2013 Discussion groups
    •  Liang Wang: Emulation
      • Discussion topics and groups 
        • ​To do days before the session: Go to the shared, openly accessible google documents page, a) sign up with one of the groups for the next session b) add your suggestions for additional discussion topics, or detailed questions for the presenters on the course.
        • Results from the group work:
          • group presents (a few minutes) the findings and opinions from the group discussion to the class, and we have a few minutes discussion at these reports;
          • joint learning diary entry from each of the groups (use the shared areas provided); deadline 7 days after the last session)
            • Discussion 1: Processes around publications, journals and conferences, scientific organisations 
            • Discussion 2: Challenges in applying standard research methods in adversarial settings
              • ​Standard research methods for evaluating a system implicitly assume benign settings: i.e., failures or faults occur by accident and not intentionally. These may not work equally well in an adversarial setting, where an attacker is deliberately trying to circumvent the system. Think about some evaluation method you are familiar with (e.g., something you used or plan to use in your thesis) and see if it would perform differently if you consider an adversarial setting. (Further hints available on request)
            • ​Discussion 3: On network performance evaluation
              • ​ Consider the aspects that you need to take into account if your task is to compare the network performance of two Web browsers for typical web browsing. Try to list everything that may affect the performance and consider also tradeoffs with different alternatives you need to select in between. That is, do an initial planning for such a performance evaluation.
            • ​Discussion 4: On formal methods
              • ​Based on the paper
                • ​Critical view: Ralf Kneuper, Limits of Formal Methods. Formal aspects of computing (1997) 3:1-000,
              • discuss the question settings that are not appropriate for being studied on formal methods alone. What kind of methods would you combine with formal methods, or replace formal methods with, in such cases.
            • Discussion 5: Massive scalability for Internet services
              • ​Consider strategies for addressing Internet service scalability. A service needs to cope with flash crowds, sudden arrival of many clients, and incremental growth. In the former, the service temporarily requires more servers and network capacity. In the latter, the number of servers and network capacity need to be increased incrementally to meet with the demand. Important points include the runtime addition and removal of servers, server placement, and network utilization. How would you design and evaluate such a system to ensure that the system works with the anticipated workload (number of clients, usage patterns).

 

 

Materials and links (under construction)

 
Reading, writing and presenting scientific results
 

Simulation Studies

 

Wireless Experiments

 

Secure systems

 

Formal methods

Basic intro
  • paper suggested on the slides
Critical view:
  • Ralf Kneuper, Limits of Formal Methods. Formal aspects of computing (1997) 3:1-000.
 

User and Interaction Research Methods

Text book:

Rogers, Y., Sharp, H., & Preece, J. (2011). Interaction Design: Beyond Human Computer Interaction.

 

Queuing networks

Easy-reader

Books:

  • Edward D. Lazowska, John Zahorjan, G. Scott Graham , Kenneth C. Sevcik Quantitative System Performance: Computer System Analysis Using Queueing Network Models (available at http://homes.cs.washington.edu/~lazowska/qsp/)
  • R. Jain, The Art of Computer System Performance: Analysis, Techniques for Experimental Design, Measurement, Simulation and Modeling, Wiley, New York, 1991. Lawrence W. Dowdy, Virgilio A.F. Almeida: Performance by Design: Computer Capacity Planning By Example, Prentice Hall, 1994
  • Daniel A. Menasce,Lawrence W. Dowdy, Virgilio A.F. Almeida: Performance by Design: Computer Capacity Planning By Example, Prentice Hall, 1994