The aim of this paper is to provide two distributed versions of the Boneh-Goh-Nissim Cryptosystem (BGNC). We give a proof of the semantic security for the first one. This guaranties that our algorithm is semantically secure in the contest of active non-adaptive adversaries. Furthermore, we prove that the second version of our distributed scheme is computationally more efficient than the ElGamal destributed elliptic curve cryptosystem (EDECC) and secure under the Subgroup Decision problem (SDP) assumption.
We give three elliptic curves E over Q such that the Mordell–Weil group $E(Q\( (\sqrt{-3}) \)) has torsion subgroup isomorphic to Z/3Z × Z/3Z and rank at least 10. The largest rank listed for this torsion group in A.~Dujella's online table of records over quadratic fields, as consulted on 19~November 2026, is 9. The curves belong to the Hesse pencil \( x^3+y^3+z^3=3t\,xyz \). In this pencil the quadratic twist by -3 of a member is again a member, so the rank over Q\( (\sqrt{-3}) \) is the sum of the ranks over Q of two members of the pencil. One curve was found by enumerating rational points of small height in the projective plane and grouping them according to the member of the pencil on which they lie; the other two were found by a sieve in the parameter space that imposes split multiplicative reduction at many primes. The lower bound 10 is unconditional and rests on exact computations, which we give in full. Under the generalized Riemann hypothesis, a 2-descent with PARI/GP, completed for one of the six curves over Q involved by PARI's computation of the Cassels–Tate pairing on its 2-Selmer group, shows that the rank of each of the three curves over Q\( (\sqrt{-3}) \) is exactly 10; a descent by 3-isogeny, which does not need this hypothesis but which we have carried out only with our own implementation, gives the same upper bound.
The aim of this paper is to provide two distributed versions of the Boneh-Goh-Nissim Cryptosystem (BGNC). We give a proof of the semantic security for the first one. This guaranties that our algorithm is semantically secure in the contest of active non-adaptive adversaries. Furthermore, we prove that the second version of our distributed scheme is computationally more efficient than the ElGamal destributed elliptic curve cryptosystem (EDECC) and secure under the Subgroup Decision problem (SDP) assumption.