A novel system identification in networked environment

This paper is motivated by the practical modeling considerations of linear time-invariant systems working in a networked environment. Specifically, we limit the problem to off-line identification of open-loop stable continuous time systems in network. The identification problem is formulated under the configuration of event-driven actuators subject to random network induced delays and packet dropouts; the associated networked identification is based on the general non-uniformly sampled data. According to deal with colored process noise and unknown initial conditions, we proposed a modified version of the generalized linear least squares algorithm with a filter to estimate the parameters of the system model. A simulation example is presented to demonstrate the performance of the proposed algorithm.