Hi, This is quite surprising, because speed is dealt with in the same manner for epfl and uniwaypoint model (they both inherit from the virtual class 'mobility' in mob_base.ml; the speed_mps instance variable in that class is then used to determine the time delta to get a new position, in the 'move' private method).
Are you setting everything up in the same manner (granularity, traffic model, warmup time, etc) between both simulations, and only switching between mobility models? If you want to send me your script file (you can send directly to me if you wish) i will take a look and try to understand what's going on. regards, henri ps: about the random seed: it is incremented each time a new mobility model object is created, so that two different mobility models generate a different sequence of random waypoints, otherwise they would all be the same. Incrementing by 1 will give a different random number stream for each mobility process, so i don't think that would be the cause of your problem. > hi. > i am graduate student in USC.i astonished the marvelous NAB so far. > however, i found some strange result for epfl mobility model ( u know the campus map > model) > > our research group under Prof. Helmy increases the node speed in the simulation. > The uniway point model is affected by the increasing node speed. > However, the epfl model shows us a same result(cost reslut) as we increase them from > 1.0 to 3.0. > - we use the hook function in the lesson3.ml to calculate the costs. > > Here is my observation,in general cases, to generate the random number we need the > random seed, right? > but when i checked the mobs.ml the random seed is initialized by zero from the > start then it is increased by 1. > i think this will cause same sequence pattern of generating random numbers. As I > know, in general, the random seed grabs the time information when it is run. ( > randomseed = time(); ). > > Does anybody have an idear about my oppinion?? > > thanks for ur reading this. > >
