A Fast Algorithm for Finding the Roots of Polynomials over Finite Fields

Pingping Li, Leilei Yu, Linqi Song, Hanxu Hou*

*Corresponding author for this work

Research output: Chapters, Conference Papers, Creative and Literary WorksRGC 32 - Refereed conference paper (with host publication)peer-review

Abstract

Finding all roots of polynomials over finite fields is a crucial problem in algebraic coding techniques, such as decoding BCH codes and Reed-Solomon codes. The current commonly used methods are Chien search or fast Fourier transform (FFT), which evaluates all elements in a finite field and results in field multiplications whose number increases linearly with the field size. In this paper, we propose a new approach that enables the multiplicative complexity to grow sub-linearly with the field size. To our knowledge, this is the currently lowest multiplicative complexity for this problem. Additionally, the proposed approach maintains the additive complexity similar to that of FFT-based method. © 2023 IEEE.
Original languageEnglish
Title of host publication2023 IEEE the 15th International Conference on Wireless Communications and Signal Processing (WCSP 2023)
PublisherIEEE
Pages299-304
ISBN (Electronic)9798350324662, 9798350324655
ISBN (Print)9798350324679
DOIs
Publication statusPublished - Nov 2023
Event15th International Conference on Wireless Communications and Signal Processing (WCSP 2023) - Huabei Hotel, Hangzhou, China
Duration: 2 Nov 20234 Nov 2023
http://www.ic-wcsp.org/2023/show.asp?id=21

Publication series

NameIEEE International Conference on Wireless Communications and Signal Processing, WCSP

Conference

Conference15th International Conference on Wireless Communications and Signal Processing (WCSP 2023)
PlaceChina
CityHangzhou
Period2/11/234/11/23
Internet address

Research Keywords

  • error-correction codes
  • error-locator polynomial
  • fast Fourier transforms (FFT)
  • Polynomial evaluation

Fingerprint

Dive into the research topics of 'A Fast Algorithm for Finding the Roots of Polynomials over Finite Fields'. Together they form a unique fingerprint.

Cite this