Failure of a parallel adaptive identifier with adaptive error filtering

Johnson, C.R.; Taylor, Trevor

doi:10.1109/tac.1980.1102538

Cited by 21 publications

(3 citation statements)

References 4 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In fact, it was shown in Ref. (15) that the minimum phase condition of D(q −1 , n) must also be satisfied in order to guarantee that e OE (n) converges to zero in the mean sense. This additional condition implies that a continuous minimum phase monitoring should be performed in the polynomial D(q −1 , n) to assure global convergence of the SHARF algorithm.…”

Section: Rementioning

confidence: 99%

New Approach for IIR Adaptive Lattice Filter Structure Using Simultaneous Perturbation Algorithm

2010

View full text Add to dashboard Cite

Adaptive infinite impulse response (IIR), or recursive, filters are less attractive mainly because of the stability and the difficulties associated with their adaptive algorithms. Therefore, in this paper the adaptive IIR lattice filters are studied in order to devise algorithms that preserve the stability of the corresponding direct-form schemes. We analyze the local properties of stationary points, a transformation achieving this goal is suggested, which gives algorithms that can be efficiently implemented. Application to the Steiglitz-McBride (SM) and Simple Hyperstable Adaptive Recursive Filter (SHARF) algorithms is presented. Also a modified version of Simultaneous Perturbation Stochastic Approximation (SPSA) is presented in order to get the coefficients in a lattice form more efficiently and with a lower computational cost and complexity. The results are compared with previous lattice versions of these algorithms. These previous lattice versions may fail to preserve the stability of stationary points.

show abstract

Section: Rementioning

confidence: 99%

New Approach for IIR Adaptive Lattice Filter Structure Using Simultaneous Perturbation Algorithm

2010

View full text Add to dashboard Cite

show abstract

“…This is achieved via an augmented parameterization. Johnson and Taylor subsequently pointed out [3] that, despite the asymptotic convergence of the a posterori adaptation error to zero, nothing could be said about the convergence of the a priori model error. This claim was supported by a suitably generated simulation example [3].…”

Section: Introduc~onmentioning

confidence: 99%

“…Johnson and Taylor subsequently pointed out [3] that, despite the asymptotic convergence of the a posterori adaptation error to zero, nothing could be said about the convergence of the a priori model error. This claim was supported by a suitably generated simulation example [3]. A partial answer to this problem was given in [4] where it was suggested that the adjustable compensator should be kept in a stability region by some form of projection procedure.…”

Section: Introduc~onmentioning

confidence: 99%