Description:
The talk will focus on two parts. The first part will be on how to design a test-bed of real ad hoc nodes moving around in an office environment. The main challenge is how to create experimental repeatability when nodes are carried around by humans under varying radio conditions. The second part of the talk will be on the value of combining simulation, emulation and real world test-beds for the same testing scenario. Results will be presented that are not visible in simulations. Protocol constructs that improve performance in simulations are shown to degrade performance in the real world.
Speaker(s):
Per Gunningberg, Ph. D., Professor, Computer Communication, Uppsala University
|
