One of the core methods to verify understanding is the use of Diary after each lecture. Throughout the T-110.6130 course every student has own unique topic, which was assigned and available in Noppa's section “Assigned topics”.

Format diaries according to latex IEEE proceedings format. 

The Diary plays the role of studying step of your own assigned topic in context of the studied methods step-by-step. Do not confuse, the Diary is not the notes from the given lecture, it is notes about your topic in context of the given lecture.

If during the lecture you were studying some research method – RM (computer simulation, data analysis, modeling, etc), the diaries should answer the following kinds of questions:

• How your topic was studied in the literature using given RM (good references are welcome)?

• If for my topic I find essentially different use of given RM: what are the main differences? What are limitations and what are benefits of those?

• (rarely) If for your topic You could not find good set of RM use (or found only one example of RM), answer additional questions explicitly: 

  How else can You study your topic using this RM? 

  Compare your own suggestions about using RM with existing ones? 

  Again, what are limitations and what are benefits of your way of study the topic (good criticism is needed)?

   

If my topic is “Internet topology” and the lecture on RM – Computer Simulation. Then first I will say that Internet topology was extensively studied and resulted in set of models, which were implemented for simulators.

An example of these is the BRITE network generator, which is able to generate different network models for NS-2, OMNeT++ and so on. These topologies are fully based on mathematical model; their usage in simulation process may reveal some general problems of protocols and architectures, while still does not show some practically existing problems in real Internet topology. Another example is the CAIDA work on reconstructing real architecture of today's Internet using some basic networking tools (ICMP, BGP tables). Using these real topologies it is possible to apply real topologies to many network simulators. An example of such applications is scientific papers, for example LIPSIN paper [1] (at SIGCOMM 2009).Although it was mentioned in [2] that CAIDA topologies are still unable to show many peering connections and MPLS communications.

It is still important to mention that simulators in mass are able to reproduce static topologies (measured by traceroute, BGP tables, or WHOIS methods), while topologies with realistic network failures are still challenging and inaccurate.

Upload the diaries to optima.aalto.fi as mentioned in the first lecture (attach own pdf and tex files) before the next lecture.