摘要

We prove that for an arbitrarily small constant epsilon > 0, the preprocessing versions of the closest vector problem and the nearest codeword problem are hard to approximate within a factor 2(log1-epsilon) (n), under the assumption that NP not subset of SIZE(2log(O(1/epsilon)) (n)).

  • 出版日期2014
  • 单位Microsoft