Improved NP-inapproximability for 2-variable linear equations

Johan Håstad, Sangxia Huang, Rajsekar Manokaran, Ryan O'Donnell and John Wright

An instance of the 2-Lin(2) problem is a system of equations of the form x_i+x_j = b (mod 2). Given such a system in which it's possible to satisfy all but an Ɛ fraction of the equations, we show it is NP-hard to satisfy all but a C*Ɛ fraction of the equations, for any C<11/8=1.375 (and any 0<Ɛ<=1/8). The previous best result, standing for over 15 years, had 5/4 in place of 11/8. Our result provides the best known NP-hardness even for the Unique-Games problem, and it also holds for the special case of Max-Cut. The precise factor 11/8 is unlikely to be best possible; we also give a conjecture concerning analysis of Boolean functions which, if true, would yield a larger hardness factor of 3/2.

Our proof is by a modified gadget reduction from a pairwise-independent predicate. We also show an inherent limitation to this type of gadget reduction. In particular, any such reduction can never establish a hardness factor C greater than 2.54. Previously, no such limitation on gadget reductions was known.