This paper develops control analysis and synthesis techniques for systems w
ith finite signal-to-noise-ratios (called the finite-signal-to-noise (FSN)
models). The system is characterized by an easily computed measure phi. The
control synthesis finds output feedback control laws to tolerate finite si
gnal-to-noise ratio uncertainty on an interval. An algorithm to conduct thi
s synthesis is proposed, to achieve locally optimal performance for the clo
sed-loop system. (C) 2000 Elsevier Science Ltd. All rights reserved.