In this paper, we present our progress toward designing a ''smart'' hi
gh-peak power microwave (HPM) tube. We use iterative learning-control
(ILC) methodologies in order to control a repetitively pulsed high-pow
er backward-wave oscillator (BWO), The learning-control algorithm Is u
sed to drive the error between the actual output and its desired value
to zero, The desired output may be a given power level, a given frequ
ency, or a combination of both. The learning-control methodology is th
en verified in simulation, This methodology is applicable to a wide va
riety of HPM sources.