Adaptive AR model identification based on the faest filters

A new adaptive approach for simultaneously selecting the order and identifying the parameters of an AutoRegressive model (AR) is presented. The proposed algorithm is based on the reformulation of the problem in the standard state space form and the subsequent implementation of a bank of fast a posteriori error sequential technique (FAEST) filters, each fitting a different order model. The problem is reduced then to selecting the true model, using the Multi-Model Partitioning (MMP) theory. Simulations illustrate that the proposed method is selecting the correct model order and identifies the model parameters, even in the case that the true model order does not belong to the bank of FAEST filters. The use of FAEST filters, reduce the computational effort, especially in the case of large order AR models. Finally, the algorithm is parallel by nature and thus suitable for VLSI implementation.