Value of Information in Feedback Control: Quantification

Although transmission of a data packet containing sensory information in networked control system improves the quality regulation, it has indeed price from communication perspective. It is, therefore, rational that such be transmitted only if is valuable sense cost-benefit analysis. Yet, fact little known so far about this valuation and its connection with traditional event-triggered communication. In present article, we study intrinsic property systems by formulating rate-regulation trade-off between rate regulation cost an event trigger controller as two distributed decision makers, show conceivable quantifiable grounded on trade-off. particular, characterize equilibrium trade-off, quantify value ${\rm{VoI}}_k$ there variation so-called function respect to piece can communicated at each time notation="LaTeX">$k$ . We prove that, for multi-dimensional Gauss--Markov process, symmetric discrepancy state estimates controller, should nonnegative. Moreover, discuss computed arbitrary accuracy, approximated closed-form quadratic performance guarantee.