|
Journal
Week 1: I met with Prof Peh and her grad students. They
explained the project to me and gave me some reading to do. I spent
the rest of the week reading the important chapters in
Interconnection Networks by Duato, Yalamanchili, and Ni.
Week 2: I met with one of the grad students and received
more reading I spent the rest of the week reading important chapters
from a book that is in the process of being written. On Friday, my
group went to a talk at Rutgers where a post-doc presented results
from his research on power efficiency in hard disks.
Week 3: I continued reading chapters from the same book,
and at the end of the week, I finished my assigned reading. Then I met with
my grad student to sum up everything I had learned and to figure out
what I should do next.
Week 4: I downloaded the code for a network simulator that Prof
Peh had written in C and C++. My grad student asked me to look at the code,
and figure out how we would use it for our research. The server was down, so we
could not compile or run the simulator. At the end of the week, I met with my grad
student to discuss how we would use the simulator.
Week 5: I continued looking at the code for the simulator, so I could
understand how it worked. I compiled the simulator and tested it with data to try to
understand it better. Once I had a good understanding of the simulator, I tested some
permutations that my grad student gave me. We used the results from these permutations
to test his theory about power in networks.
Week 6: I tested some more permutations with the simulator. I also
added some routing algorithms to the router. My grad student and I used the other
routing algorithms to test his theory about how the number of links used affects
power. I ran the simulator with the new routing algorithms and a few different
permutations.
Week 7: I tested some more permutations and some more routing
algorithms. I used a specific permutation for a 5x5 network and tested it with
different routing algorithms. Each routing algorithm used a different
number of links, so I then compared the power consumption of the network for
the different routing algorithms. I did the same for a permutation for a
8x8 network.
Week 8: I gathered all of the results from the tests I had run on
the simulator into an excel file. I used graphs and tables to compare the power and
delay for each test run. Then, I used the parameters from the simulator to do some
calculations and make sure our results were correct. I also began working on my
final report.
Week 9: I finished up my analysis and my final report. On Friday,
my group went to a talk to Rutgers where a grad student gave a talk about
heterogenous networks.
|