Computer simulation is the most common approach to studying wireless
ad-hoc routing algorithms. It offers a flexible means of
experimentation, in a repeatable and controllable fashion. Simulation
results, however, are only as good as the models the simulation uses.
One should not underestimate the importance of validation, as
inaccurate models can lead to wrong conclusions.  In this paper, we
use direct-execution simulation to validate radio models that drive
ad-hoc routing protocols, against real-world experiments.  This paper
documents a common testbed that supports direct execution of a set of
ad-hoc routing protocol implementations in a wireless network
simulator. The testbed read traces generated from real experiments,
and uses them to drive direct-execution implementations of several
ad-hoc routing protocols.  Doing so we reproduce the same network
conditions as in real experiments. By comparing routing behavior
measured in real experiments with behavior computed by the simulation,
we are able to validate the models of radio behavior upon which
protocol behavior depends.  Our experiments reveal two conclusions.
One is that it is possible to have fairly accurate results using a
simple radio model, but the second is that routing behavior is quite
sensitive to one of this model's parameters.  The implication is that
one either should (i) use a more complex radio model that explicitly
models point-to-point path loss, or (ii) use measurements from an
environment typical of the one of interest, or (iii) study behavior
over a range of environments to identify sensitivities.