Improving Performance of Affine Projection–Like Algorithm Using Two Combination Methods

Yinxia Dong, Zongsheng Zheng, Xudong He, Baoquan Wang, Fengkun Liu, Hui Li and Linlan Wang

Journal of Applied Mathematics, 2025, vol. 2025, 1-8

Abstract: The affine projection–like (APL) algorithm has garnered significant attention due to its low steady-state mean-square deviation and simplicity. However, its performance is limited by a fixed step-size, which forces a trade-off between fast convergence and low steady-state error. To overcome this limitation, this paper introduces two novel variants of the APL algorithm: the convex combination affine projection–like (CC-APL) and the combined step-size affine projection–like (CSS-APL) algorithms. The CC-APL algorithm leverages a convex combination of adaptive filters with different step-sizes, allowing for dynamic adjustment between convergence speed and accuracy. The CSS-APL algorithm optimizes performance by integrating multiple step-sizes directly into the adaptation process. Both algorithms are designed to enhance the balance between convergence rate and steady-state mean-square deviation, addressing a key drawback of the traditional APL. Simulation results demonstrate that the proposed algorithms significantly improve performance compared to the conventional APL algorithm, especially in scenarios requiring both fast adaptation and low mean-square deviation.

Date: 2025
References: Add references at CitEc
Citations:

Downloads: (external link)
http://downloads.hindawi.com/journals/jam/2025/1787885.pdf (application/pdf)
http://downloads.hindawi.com/journals/jam/2025/1787885.xml (application/xml)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:hin:jnljam:1787885

DOI: 10.1155/jama/1787885

Access Statistics for this article

More articles in Journal of Applied Mathematics from Hindawi
Bibliographic data for series maintained by Mohamed Abdelhakeem ().