[Ns-developers] [ns-3] proposed RNG changes

[Michele Weigle]

Hi all,

After discussions with Tom and Mathieu, here are some proposed changes  
to the way the RNG and RandomVariable classes work in ns-3.  Please  
feel free to comment.

* no time-based seed initialization

We decided to remove the default behavior of setting the global RNG  
seed by timeofday.  Instead, the default will be to have the seed set  
to some constant value.  This will result in ns-3 producing the same  
results each time it is run, unless the user changes the global seed  
or changes the run number.

The reason for this change is to enforce good simulation practices.   
If the global seed is set by timeofday, the user needs to carefully  
record the exact seed used in order to re-run the same experiment.   
More importantly, if the user plans on running several experiments to  
obtain statistics like an average with confidence intervals, using  
timeofday to set the seed does not guarantee that the multiple runs  
will be independent.

The L'Ecuyer RNG default in ns-3 provides a very easy way to ensure  
that multiple runs are guaranteed to be independent. This is by  
setting a global seed once and changing only the run number for each  
replication of the experiment.  The resulting numbers are guaranteed  
to be random and independent.

In addition, with the L'Ecuyer RNG, there is no notion of 'bad' seeds  
or 'good' seeds.  Any number used as a seed is valid. The range of  
possible seeds is 1 to MAXINT.

The RNG also provides a large stream of random numbers so that for  
each experiment, up to 2^51 (or, 2.3 x 10^15) independent replications  
are possible.  This means that the maximum run number allowed is 2^51  
(or less, depending on the datatype used to store the run number).

* change the interface of setting the RNG seed to 1 value instead of 6  
values

This is just a simplification of the interface.  The actual seed in  
the L'Ecuyer RNG is a 6-tuple, which will be set all to the same value  
provided by the user.

* remove the GetSingleValue function from RandomVariable classes

The GetSingleValue function allows for potential non-independence as  
it draws values from a single RNG stream for different random  
variables.  Each random variable should have its own RNG stream.

* provide Stream API

We will provide a Stream API to allow the user to set the seed and run  
number of the global RNG.  This Stream API will also allow access to  
the first RNG stream (not associated with any random variable) and  
allow the user to draw numbers from a uniform distribution with min  
and max as parameters.
Ex: double PickUniform (double min, double max);
This is different from the random variable Uniform because once the  
RandomVariable object is created, min and max cannot be changed.

-Michele

--
Michele Weigle
Assistant Professor
Department of Computer Science
Old Dominion University
Norfolk, VA 23539
mweigle at cs.odu.edu
http://www.cs.odu.edu/~mweigle
(757) 683-6001 ext. 5050