Sliding-Window RLS Low-Cost Implementation of Proportionate Affine Projection Algorithms

Author(s)

Zakharov, Yuriy ^[1] ; Nascimento, Vitor H. ^[2]

Affiliation

^[1] Univ York, Dept Elect, York YO10 5DD, N Yorkshire - England

^[2] Univ Sao Paulo, Dept Elect Syst Engn, BR-05508970 Sao Paulo - Brazil

• Document type: Journal article
• Source: IEEE-ACM TRANSACTIONS ON AUDIO SPEECH AND LANGUAGE PROCESSING; v. 22, n. 12, p. 1815-1824, DEC 2014.
• Web of Science Citations: 5

Abstract

This paper addresses adaptive filtering for sparse identification. Proportionate affine projection algorithms (PAPAs) are known to be efficient techniques for this purpose. We show that the PAPA performance may improve with an increase in the projection order M (for example, such as M = 512), which, however, also results in an increased complexity; the complexity is in general O((MN)-N-2) or at least O(MN) operations per sample, where N is the filter length. We show that PAPAs are equivalent to specific sliding-window recursive least squares (SRLS) adaptive algorithms with time-varying and tap-varying diagonal loading (SRLS-VDLs). We then propose an approximation to the SRLS-VDLs based on dichotomous coordinate descent (DCD) iterations with a complexity of O(NuN) which does not depend on M; it depends on the number of DCD iterations N-u, which as we show can be significantly smaller than M, thus allowing a low-complexity implementation of PAPA adaptive filters. (AU)

• FAPESP's process: 11/06994-2 - Low-cost algorithms for acoustic signal processing
• Grantee: Vitor Heloiz Nascimento
• Support Opportunities: Regular Research Grants
• FAPESP's process: 12/50565-1 - Adaptive compressive sensing-aware techniques: desinf algorithms and applications
• Grantee: Vitor Heloiz Nascimento
• Support Opportunities: Regular Research Grants