A polyphase IIR adaptive filter: error surface analysis and application

Burt, P.M.S.; Gerken, M.

doi:10.1109/icassp.1997.599508

Cited by 10 publications

(13 citation statements)

References 3 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Steiglitz-McBride method, SHARF (Simple Hyperstable Adaptive Recursive Filter) or the PLR (Pseudo Linear Regression) is prone to very slow convergence under adverse conditions [10].…”

Section: Any Constant Gain Iir Formulation Such As the Rpem (Recursivmentioning

confidence: 99%

“…In spite of its good general properties, the LMSOE algorithm may be inecient in the presence of underdamped or clustered poles [10,11,19], particularly when the parameters are approaching the global minimum (see Appendix C).…”

Section: Normalized Iir Lms (N-lmsoe)mentioning

confidence: 99%

“…Besides, experience and reports from several works indicate that the magnitude of the instant gradient vectors can change appreciably over the error surfaces (e.g., [10,26]). Under these circumstances, the normalization of the regressor φ i mitigates the inuence of the input signal amplitude on the gradient noise and simplify the choice of µ.…”

Section: Normalized Iir Lms (N-lmsoe)mentioning

confidence: 99%

“…In fact, IIR AFs tend to require fewer coecients than comparable FIR AFs, particularly when the underlying system they are trying to model has sharp resonances [8,10].…”

mentioning

confidence: 99%

“…Speed of convergence remains an open problem though [1,4,11,19,24], with IIR AFs sometimes exceeding the typical convergence times of FIR AFs by orders of magnitude [10]. Unlike the stability issues, this can not be solved by merely changing the lter structure, although a given IIR realization could show improved performance against another under specic conditions [13] 4 .…”

mentioning

confidence: 99%

See 4 more Smart Citations

Hybrid convex combinations for IIR system identification.

Ferro¹

View full text Add to dashboard Cite

The low complexity of IIR adaptive lters (AFs) is specially appealing to realtime applications but some drawbacks have been preventing their widespread use so far. For gradient based IIR AFs, adverse operational conditions cause convergence problems in system identication scenarios: underdamped and clustered poles, undermodelling or non-white input signals lead to error surfaces where the adaptation nearly stops on large plateaus or get stuck at sub-optimal local minima that can not be identied as such a priori. Furthermore, the non-stationarity in the input regressor brought by the lter recursivity and the approximations made by the update rules of the stochastic gradient algorithms constrain the learning step size to small values, causing slow convergence. In this work, we propose IIR performance enhancement strategies based on hybrid combinations of AFs that achieve higher convergence rates than ordinary IIR AFs while keeping the stability.

show abstract

“…Steiglitz-McBride method, SHARF (Simple Hyperstable Adaptive Recursive Filter) or the PLR (Pseudo Linear Regression) is prone to very slow convergence under adverse conditions [10].…”

Section: Any Constant Gain Iir Formulation Such As the Rpem (Recursivmentioning

confidence: 99%

Section: Normalized Iir Lms (N-lmsoe)mentioning

confidence: 99%

Section: Normalized Iir Lms (N-lmsoe)mentioning

confidence: 99%

“…In fact, IIR AFs tend to require fewer coecients than comparable FIR AFs, particularly when the underlying system they are trying to model has sharp resonances [8,10].…”

mentioning

confidence: 99%

mentioning

confidence: 99%

See 3 more Smart Citations

Hybrid convex combinations for IIR system identification.

Ferro¹

View full text Add to dashboard Cite

show abstract

On the uniqueness of stationary points in L2 approximation with a rational polyphase function

Burt¹,

Gerken²

2002

Signal Processing

View full text Add to dashboard Cite

FIR–IIR adaptive filters hybrid combination

Ferro¹,

Chamon

Lopes

2014

Electron. lett.

View full text Add to dashboard Cite

To enhance the performance in IIR system identification scenarios, a hybrid combination of FIR and IIR adaptive filters (AFs) via a supervisor that senses which one is performing best is proposed. The FIR-LMS AF is short, providing fast and robust convergence, whereas the IIR-LMS AF is slow but accurate. The stagnation effect caused by the different convergence rates is tackled through cyclic weight transfers FIR → IIR, which also ensure good tracking properties. A design technique for the transfers cycle length is proposed, providing good convergence while keeping computational cost low.

show abstract

A polyphase IIR adaptive filter: error surface analysis and application

Cited by 10 publications

References 3 publications

Hybrid convex combinations for IIR system identification.

Hybrid convex combinations for IIR system identification.

On the uniqueness of stationary points in L2 approximation with a rational polyphase function

FIR–IIR adaptive filters hybrid combination

Contact Info

Product

Resources

About