On parameter identifiability of linear multivariable systems with communication access constraints

The problem of on-line parameter identification was discussed for linear multivariable discrete time stochastic systems with communication access constraints. Based on the concept of parameter estimability defined with mutual information, a condition for identifiability was proved under the assumption that the time-varying parameter can be modeled as a Gauss-Markov process, and the sensors access status is described by binary-value function. Analytical analysis and simulation results show that, there is a proper communication strategy which preserving the identifiability of the system under access constraints.