Data Rate Limitations for Stabilization of Uncertain Systems over Lossy Channels
It provides fundamental limits for networked control of uncertain systems, extending prior results to handle parametric uncertainties.
The paper derives necessary and sufficient conditions for mean square stabilization of uncertain discrete-time linear systems over lossy channels, expressed in terms of data rate, packet loss probability, uncertainty bounds, and unstable eigenvalues. The condition is exact for scalar plants.
This paper considers data rate limitations for mean square stabilization of uncertain discrete-time linear systems via finite data rate and lossy channels. For a plant having parametric uncertainties, a necessary condition and a sufficient condition are derived, represented by the data rate, the packet loss probability, uncertainty bounds on plant parameters, and the unstable eigenvalues of the plant. The results extend those existing in the area of networked control, and in particular, the condition is exact for the scalar plant case.