An oversampled subband adaptive filter without cross adaptive filters

Harteneck, M.; Pàez-Borrallo, J.M.; Stewart, Robert W.

doi:10.1016/s0165-1684(97)00179-5

Cited by 19 publications

(5 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Differences arise for the decimation of complex or real valued frequency bands. The decimation of real valued bandpass signals is generally complicated, and real valued signals have to be either modulated into the baseband prior to decimation by, for example, single sideband modulation (SSB, [1,19]), or their bandwidth and decimation ratio has to be chosen in accordance with the sampling theorem, leading to non-uniform filter banks [13,4]. In contrast, the decimation of complex valued bandpass signals with any integer factor N K is straightforward.…”

Section: Adaptive Filtering In Subbands 21 Structural Approaches To Safmentioning

confidence: 99%

A subband adaptive equalization structure

Weiss¹,

Garcia-Alis²,

Stewart³

1999

IEE Colloquium Novel DSP Algorithms and Architectures for Radio Systems

View full text Add to dashboard Cite

The potential presence of fractional delays, non-minimum phase parts, and a colouring of the channel output can require adaptive equalizers to adapt very long impulse responses. Besides resulting in a large computational complexity, this will in general cause slow convergence for LMS-type adaptive algorithms. In this paper, we address the equalization problem by a subband approach to reduce computational complexity and to improve convergence speed. We discuss, why amongst other possibilities of subband processing the oversampled approach is particularly appealing to significantly reduce computational complexity and improve convergence speed. Simulation results for a typical communication channels are presented and highlight the benefit of our method

show abstract

Section: Adaptive Filtering In Subbands 21 Structural Approaches To Safmentioning

confidence: 99%

A subband adaptive equalization structure

Weiss¹,

Garcia-Alis²,

Stewart³

1999

IEE Colloquium Novel DSP Algorithms and Architectures for Radio Systems

View full text Add to dashboard Cite

show abstract

“…As a result, a number of works using real-valued subband signals have been proposed. These include the use of non-uniform filter banks [6][7][8][9] to improve the subsampling rate of the filter bank. One major drawback of this scheme would be the need to handle different subsampling rates and hence the increase in difficulty of implementation.…”

Section: Introductionmentioning

confidence: 99%

Subband adaptive filtering with real-valued subband signals for acoustic echo cancellation

Chin

Farhang‐Boroujeny

2001

IEE Proc., Vis. Image Process.

View full text Add to dashboard Cite

Adaptive filtering in subbands have originally been proposed to overcome limitations of classic adaptive algorithms in some areas. In general, subband adaptive filters offer computational savings, as well as faster convergence over the classical Least Mean Squared (LMS) algorithm. However, improvements to current subband adaptive filters could be further enhanced by a more elegant choice of their design/structure. Classical subband adaptive filters employ DFT-based analysis and synthesis filter banks which results in subband signals that are complexvalued. In this paper, we modify the structure of subband adaptive filters by using single sideband (SSB) modulated analysis and synthesis filter banks, which result in subband signals that are real-valued. This simplifies realisation of subband adaptive filters.

show abstract

“…Furthermore, exploiting the relationship with cosine-modulated filter banks generates a subclass of the (G)DFT filter banks, but not the whole class [9,10]. As one might expect, direct formulation of the design of a GDFT prototype (over the whole available class) also leads to non-convex optimization problems [16,18,19,22]. (Some simple, but ad-hoc, prototype design methods have also been proposed [15,20].)…”

Section: Gdft Filter Banksmentioning

confidence: 99%

“…In particular, the performance of systems based on the class of oversampled generalized Discrete Fourier Transform (GDFT) filter banks is quite encouraging [16][17][18][19][20][21][22]. These filter banks are better able to suppress aliasing in the subbands than (uniform) oversampled cosinemodulated filter banks [18,19,21], and can be efficiently implemented using the GDFT [1]. In this paper we provide a flexible, efficient design technique for the prototype filter of an oversampled near perfect reconstruction (NPR) GDFT filter bank.…”

Section: Introductionmentioning

confidence: 99%

“…The design criteria are explicit bounds (derived herein) on the aliased components in the subbands (and the output), the imaged subband error components in the output, and the distortion induced by the filter bank. These bounds rigorously amalgamate several intuitively developed design criteria in the current literature [15,[18][19][20][21][22], and subsume the criteria derived in [23] and [16,. Our design criteria generate familiar constraints on the prototype filter: the aliasing criteria result in bounds on the stop-band energy and the maximum stop-band level, the imaging criterion results in an additional bound on the transition-band energy, and the distortion criterion results in a measure of the distance between the prototype filter and a 'self-orthogonal' filter.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Efficient design of oversampled NPR GDFT filter banks

Wilbur

Davidson

Reilly

2003 IEEE International Conference on Acoustics, Speech, and Signal Processing, 2003. Proceedings. (ICASSP '03).

View full text Add to dashboard Cite

In this paper, we present a flexible, efficient design technique for the prototype filter of an oversampled near perfect reconstruction (NPR) generalized Discrete Fourier Transform (GDFT) filter bank. Such filter banks have several desirable properties for subband processing systems which are sensitive to aliasing, such as subband adaptive filters. Our design criteria for the prototype filter are explicit bounds (derived herein) on the aliased components in the subbands and the output, the distortion induced by the filter bank, and the imaged subband errors in the output. It is shown that the design of an optimal prototype filter can be transformed into a convex optimization problem which can be efficiently solved. Our design technique provides an efficient and effective tool for exploring many of the inherent trade-offs in the design of the prototype filter, including the trade-off between aliasing in the subbands and the distortion induced by the filter bank. In our examples we calculate several of these trade-offs and demonstrate that our method can generate filters with significantly better performance than filters obtained using current design methods. * A condensed version of this report has been accepted, subject to minor revisions, for publication in the IEEE Transactions on Signal Processing

show abstract

An oversampled subband adaptive filter without cross adaptive filters

Cited by 19 publications

References 9 publications

A subband adaptive equalization structure

A subband adaptive equalization structure

Subband adaptive filtering with real-valued subband signals for acoustic echo cancellation

Efficient design of oversampled NPR GDFT filter banks

Contact Info

Product

Resources

About