Statistical properties of random transport models defined on discrete
space-time are investigated both numerically and analytically. As an e
xtreme limit we first consider aggregation limit of massive particles.
With the presence of permanent injection we have a nontrivial steady
state where the mass distribution follows a power law. It is shown tha
t the steady state is universal and very robust. Next, we analyze the
cases of imperfect aggregation that a finite portion is transported at
a time. We have a Gaussian fluctuation governed by the ordinary diffu
sion equation in the nonaggregation limit, while the system converges
to the power law steady state in the aggregation limit even without in
jection. In the intermediate cases the fluctuations are always between
Gaussian and the power law. Underlying relations to the exponential-l
ike distributions in fluid turbulence are also discussed.