Skip to Content
MathematicsMathematics
  • Article
  • Open Access

6 November 2026

20 Pages

Minimal-Polynomial-Based Lightweight Cryptography for IoT and Wireless Hardware: Security Analysis and FPGA Implementation

and
1
Biomedical Device Technology, Vocational School of Health Services, Ankara Medipol University, Ankara 06050, Türkiye
2
Department of Electrical and Electronics Engineering, Ankara Medipol University, Ankara 06050, Türkiye
*
Author to whom correspondence should be addressed.
Mathematics2026, 14(21), 4082;https://doi.org/10.3390/math14214082 
(registering DOI)

Abstract

This paper proposes a lightweight polynomial-based encryption architecture for resource-constrained IoT and wireless communication hardware. Instead of substitution boxes, the method uses key-dependent polynomial multiplication with minimal polynomials over finite fields. Plaintext rows are XORed with secret keys and a nonce-dependent pseudorandom mask and multiplied by a minimal polynomial; multiplication by a complementary factor recovers the message exactly. A security analysis shows that, because the transformation is linear over GF(2), reuse of the same key and polynomial allows for known- and chosen-plaintext key recovery; per-message nonce masking is therefore required. Software experiments (four message lengths, 100 trials) verified lossless decryption. The encryption time of the key-assisted method was comparable to that of SPECK64/128 at the same implementation level, whereas AES-128 was faster for long messages, and the avalanche ratio of the proposed method (0.070) was well below the value of about 0.5 obtained by the reference ciphers. On a Xilinx Artix-7 FPGA, the proposed core used 1211 LUTs and 205 registers without block RAM or DSPs and reached 154.1 MHz, compared with 2305 LUTs and 2733 registers for an iterative AES-128 core. The method is a compact, S-box-free polynomial processing layer rather than a stand-alone replacement for standard ciphers.

Article Metrics

Citations

Article Access Statistics

Multiple requests from the same IP address are counted as one view.