

Dergi
Glsvlsi 2026 Proceedings of the Great Lakes Symposium on VLSI 2026
Cilt
—
ISSN
—
Yayın Tarihi
22 Haziran 2026
Ascon is the National Institute of Standards and Technology (NIST) standard for lightweight cryptography, providing Authenticated Encryption with Associated Data (AEAD), hash, eXtendable-Output Function (XOF), and Customizable XOF (CXOF). Prior studies mainly target AEAD implementations of earlier Ascon versions with a little focus on hash and unified designs, and generally explore the area and latency tradeoff by varying permutation rounds per one clock cycle. In this work, we present a comprehensive design space exploration of the complete Ascon family of the NIST standard for the first time through two design architectures: (i) the conventional round-based variants that perform 1, 2, and 4 rounds per one clock cycle, i.e., v1r1c, v2r1c, and v4r1c, respectively, and (ii) the operation-based variants that execute one 64-bit and 32-bit operation per one clock cycle using a cycle-optimized schedule that completes one permutation round with a total of 28 and 56 cycles, i.e., v1op1c_64b_28 and v1op1c_32b_56, respectively. Our design space exploration includes the Ascon-AEAD128 encryption, decryption, unified encryption/decryption, unified Ascon-Hash256, Ascon-XOF128, Ascon-CXOF128, and a complete Ascon unified family. Experimental results show that the round-based variants achieve low latency and energy consumption, while the operation-based variants lead to designs with low area. Our implementations also have smaller hardware complexity than the state-of-the-art designs. For example, on the Ascon-AEAD128 designs performing encryption, our round-based v1r1c variant achieves an area reduction of when compared to a round-based state-of-the-art design, while our v1op1c_64b_28 variant reduces latency and energy consumption by and, respectively when compared to the state-of-the-art design using one 64-bit operation per one clock cycle, completing one permutation round in 59 clock cycles.