International Association
for Cryptologic Research

NTRU-like cryptosystems are among the most studied lattice-based post-quantum candidates. While most NTRU proposals have been introduced over a commutative ring of quotient polynomials, other rings can be used. Noncommutative algebra has been endorsed as a direction to build new variants of NTRU a long time ago. The first attempt to construct a noncommutative variant was due to Hoffstein and Silverman motivated by more resistance to lattice attack. The scheme has been built over the group ring of a dihedral group. However, their design differed from standard NTRU and soon was found vulnerable to algebraic attacks. In this work, we revive the group ring NTRU over the dihedral group as an instance of the GR-NTRU framework. Unlike many proposals of noncommutative variants in the literature, our work focuses on putting the scheme into practice. We clear all the aspects that make our scheme implementable by proposing an efficient inversion algorithm over the new setting of the noncommutative ring, describing the decryption failure model, and analyzing the lattice associated with our instantiation. Finally, we discuss the best-known attacks against our scheme and provide an implementation targeting 128-bit, 192-bit, and 256-bit levels of security as proof of its practicality.