On Lattices, Learning with Errors, Random Linear Codes, and Cryptography

摘要

Our main result is a reduction from worst-case lattice problems such as GapSVP and SIVP to a certain learning problem. This learning problem is a natural extension of the "learning from parity with error" problem to higher moduli. It can also be viewed as the problem of decoding from a random linear code. This, we believe, gives a strong indication that these problems are hard. Our reduction, however, is quantum. Hence, an efficient solution to the learning problem implies a quantum algorithm for GapSVP and SIVP. A main open question is whether this reduction can be made classical (i.e., nonquantum).rnWe also present a (classical) public-key cryptosystem whose security is based on the hardness of the learning problem. By the main result, its security is also based on the worst-case quantum hardness of GapSVP and SIVP. The new cryptosystem is much more efficient than previous lattice-based cryptosystems: the pu b lic key is of size. O(n~2) and encrypting a message increases its size by a factor of O(n) (in previous cryptosystems these values are O(n~4) and O(n~2), respectively). In fact, under the assumption that all parties share a random bit string of length O(n~2), the size of the public key can be reduced to O(n).