|
|
This week, I ran a number of different simulations and plotted the data looking for trends. I began by breaking last week's data into the three energy components: receive, transmit and idle energy. I had to make some minor changes in the NS-2 source code in order to do this. The results showed a constant and large portion of idle energy. This left the trends found in the receive and transmit areas. The next step was to start correlating these two energy components with the timekeeping metrics from Yong's previous paper. These include Single Path Duration(SPD), Multiple Path Duration(MPD) and Multiple Path Interval(MPI). SPD measures the minimum duration of a single path, MPD measures the time during which at least a path exists from any source to sestination, and MPI measures the average time in which no path exists between all pairs of nodes. We expected to see some correlation; the duration of a route would directly affect the necessity to compute a new route after a given time period. However, the data I collected did not show any trends. We hypothesized that this could be because if a route does exist, no control packets will be sent out for route computation, but data packets would be sent out more frequently. Therefore the energy consumption of the energy and data packets might have been contradicting one another, giving us meaningless results. I gathered some more databy post-processing the trace file using a gawk script, and separated the control and data packet energy consumption values. I still could not see any correlation. After more discussions with both Professor Peh and Yong, it seems there are still more factors that can be affecting the energy consumption. Perhaps there is still a correlation, but there is more to the model than just these metrics. Some ideas include a dependency on the number of hops in a route, and the number of failed packet delivery attempts. Both can be tested, but I do not have time to incorporate them into this project. During my meeting with Professor Peh, she gave me some guidelines of how to structure my final report. I started my introduction on Friday.
|
|