Efficient implementation of a least-squares based algorithm for noisy input-output systems

A new least-squares based method is established to perform unbiased parameter estimation of linear systems from noisy input and output measurements. The significance of the developed method lies in its improved computational efficiency since the underlying noisy system is now identified in a direct manner, with the augmented noisy system being introduced only as an auxiliary system but not being actually identified. Simulation results confirm that the presented efficient implementation scheme can retain the same estimation accuracy with reduced numerical costs.