Stabilization of Systems Over Bit-Rate-Constrained Networked Control Architectures

Abstract

We consider the problem of control of open-loop unstable systems over a networked control architecture. We focus on the case where the system state is measured and where control commands are transmitted over a fixed bit-rate channel. We describe a novel nonlinear decoder model predictive controller which utilizes 1 bit per sample. The required bit rate can then be adjusted by varying the number of samples per unit time. We establish practical stability for the algorithm. We also compare the minimum fixed bit rate required by our algorithm with known results on the minimum bit rate necessary to achieve stability.