Replace the standard random number generator with a Distributuion object

## What shortcoming will this fix cover
Currently, Jetedge uses a plain random number generator for all of the object generation. This should be replaced with a Distribution class to provide a level of abstraction.

## Why its important to make this change
By having a more generic interface for random number generation, it will be possible to overload the standard flat value generation with a distribution: Gaussian, Poison, Exponential, etc...

## How much time I would invest in making this patch (if I had the free time to do it)
2 hours

No files or tests are identified; start by tracing the plain random number generator used throughout Jetedge's object generation. Define the Distribution abstraction and confirm that object generation can use it while supporting future Gaussian, Poisson, and Exponential distributions.

Written by the indexing model from the issue text.