Elements that must be included in the initial project report (~8 pages)
and the final project report (12 pages + appendices):
-
Title
- Abstract:
Around half page (or less) outlining the work, its scope, what is the
problem
statement in brief, how is the approach different than main
previous approaches to this problem and what are the main
observations/results?
-
Motivation:
Why you think this topic is important? why is this a significant area of research?
What are the main problems in this area?
and why do you think your proposal is important for the topic?
-
Related Work:
What other researchers have proposed earlier to address the above problems? What were
their results and conclusions? Why you think you still need to do work in this area
(i.e., what are the shortcomings/pitfalls of earlier work?) What do you propose that
is different from related work? and what do you propose that is similar?
Clarify how your ideas address the shortcomings of earlier work.
Provide adequate and proper citations of related work.
-
Project Body:
-
Problem statement:
State very clearly and very specifically the question(s) that you attempt to answer
(Be as clear and as specific as you can). You can list the questions but add an
introduction sentence to link the questions and tell the whole story.
-
Evaluation metrics:
Mention very clearly and as specific as possible the evaluation criteria and metrics
you intend to use to evaluate your approach and ideas.
Examples of evaluation metric may include 'overhead' (need to clarify what overhead
is and how is it measured), response delay (how to measure it), etc.
-
Investigated Parameter Space:
Mention the design space investigated in the study. For example, a popularity
threshold of 5-50 with steps of 5 could be a design space investigated for the SDS
promotion scheme. Another example may be random selection of contacts, selection based
on stability measures (need to define what stability is and what are the limits
investigated).
-
Methodology:
Graph theory, finite state machines, queueing theory, network simulation, mathematical
modeling or experimentation are some examples of the methodology. Mention which of the
methodologies will be used and why. Elaborate on the tool used, if any. For example,
network simulator (NS-2) or GlomoSim may be used as network simulators. Also,
elaborate on the existing capabilities of the tools and what needs to be modified or
augmented to suit your analysis.
-
Scenarios:
Include topologies (network size, connectivity, etc.), traffic and multicast
membership patterns (if any), network failures, packet loss patterns, mobility
models/degrees that will be used to evaluate the protocol. Explain your choise of the
above parameters.
-
Expected results:
Show what change in performance/evaluation you would gain by using your approach (as
opposed to previous work).
- Results and Analysis: (*probably the most
important section*)
- Describe the specific settings of the evaluatoin scenarios and
outline the experiment(s).
- Show the result graphs and explain them very clearly (using
self-explanatory graph labeling, graph captions and explanation in the
text).
- Focus on the main points of the results and start extracting useful
observation that may help in drawing conclusions.
- important
Analyze these main points, explaining why they occured? was this
result expected? (if yes, was it the same reason initially thought or a
different one? if no, why not? was the flaw in the initial thinking?)
-
References: At least 6 references of high quality (i.e.,
IEEE/ACM conferences and journals) per project.