well, when you're a researcher, the first thing you'll do is to search for what you'll search for.
i had a talk with the boss of my
research unit in networking when signing the contract and we came to the agreement that i'll work on
Active Networks ... I took monthes reading thesis, papers and the like about the existing technologies, looking for what could be interresting in the regionnal project that was "hiring" me ...
I picked up 'ANTS' because it looked both simple and smart to me ... Then i had to find a way to
simulate the behaviour of large ANTS networks (the usual
Network Simulator had little support for Active Networks), so i picked up a Java-based network simulator (SSFNet) and tried to integrate ANTS in it (which actually was a success).
Later on, partners were dropping SSFNet for J-Sim, so i had to understand that weird component-and-contracts based stuff, then refactor my code so that it now works in both environments.
The "RADAR" project actually popped up because i realized active nodes couldn't talk to each other in the simulator unless they were told where other nodes were. Since it already was a pain to sketch topologies of several tenths of nodes, i *didn't* want to give the "overlay" topology aswell and thus went for a way to build it automatically ...
Now, i'm no longer on the "regionnal project" so i slightly shifted my research program (no longer focussed on mobility) ... I'm about to submit a paper on my first research results in Atlanta, USA in may '05 so if you're interrested, watch for website changes by that time ...
(sorry, dehbi&keeper: i intentionnally not replied immediately to keeper since it wasn't the most appropriate place
