Stochastic Recursive Algorithm with Modified SPR Condition

A new adaptive stochastic algorithm is introduced which guarantees the convergence when the passivity condition fails without a priori information about the unknown model. The proposed algorithm consists of three stages. In the first stage an autoregressive model is fitted to estimate the parameters of the autoregressive polynomial. In the second stage, a white noise dither signal is used to estimate the autoregressive part of the autoregressive moving average model. In the third stage, an estimate of the actual noise is generated to obtain an improved autoregressive moving average model. The simulation results given shows that the proposed algorithm compares favorably with the algorithm introduced by Mayne and Clark and also Landau.