Practical aspects of system stabilization with minimum data rates

Distributed control systems implement feedback links over computer networks to transmit quantized data with finite precision. Uncertainties introduced by quantization errors can degrade the designed closed-loop performance or even destabilize the closed-loop system. The theoretical lower bounds on minimum stabilizing data rates over the feedback links were recently found for large classes of systems. In this paper, we propose several practical implementations using floating point quantization and state feedback to achieve these theoretical lower bounds. In particular, we analyze two system realizations, viz. for digital systems with plants implemented by microcontrollers where the designer has access to internal data registers and those originating from analog counterparts where no such information is available. Simulation results show that these realizations can actually achieve and in particular cases even beat the theoretical minimum. They are of significant value to the system designers.