Learning with errors (LWE) is a problem in machine learning. A generalization of the parity learning problem, it has recently been used to create public-key cryptosystems based on worst-case hardness of some lattice problems. The problem was introduced by Oded Regev in 2005.
Given access to samples where and , with assurance that, for some fixed linear function with high probability and deviates from it according to some known noise model, the algorithm must be able to recreate or some close approximation of it.
Denote by the additive group on reals modulo one. Denote by the distribution on obtained by choosing a vector uniformly at random, choosing according to a probability distribution on and outputting for some fixed vector where the division is done in the field of reals, and the addition in .
The learning with errors problem is to find , given access to polynomially many samples of choice from .
For every , denote by denote the one-dimensional Gaussian with density function where , and let be the distribution on obtained by considering modulo one. The version of LWE considered in most of the results would be
The LWE problem described above is the search version of the problem. In the decision version, the goal is to distinguish between noisy inner products and uniformly random samples from (practically, some discretized version of it). Regev showed that the decision and search versions are equivalent when is a prime bounded by some polynomial in .
Solving decision assuming search
Intuitively, it is easy to see that if we have a procedure for the search problem, the decision version can be solved easily: just feed the input samples for the decision version to the procedure for the search problem, and check if the returned by the search procedure can generate the input pairs modulo some noise.
Solving search assuming decision
For the other direction, suppose that we have a procedure for the decision problem, the search version can be solved as follows: Recover the one co-ordinate at a time. Suppose we are guessing the first co-ordinate, and we are trying to test if for a fixed . Choose a number uniformly at random. Denote the given samples by . Feed the transformed pairs to the decision problem. It is easy to see that if the guess was correct, the transformation takes the distribution to itself, and otherwise takes it to the uniform distribution. Since we have a procedure for the decision version which distinguishes between these two types of distributions, and errs with very small probability, we can test if the guess equals the first co-ordinate. Since is a prime bounded by some polynomial in , can only take polynomially many values, and each co-ordinate can be efficiently guessed with high probability.
For a n-dimensional lattice , let smoothing parameter denote the smallest such that where is the dual of and is extended to sets by summing over function values at each element in the set. Let denote the discrete Gaussian distribution on of width for a lattice and real . The probability of each is proportional to .
The discrete Gaussian sampling problem(DGS) is defined as follows: An instance of is given by an -dimensional lattice and a number . The goal is to output a sample from . Regev shows that there is a reduction from to for any function .
Regev then shows that there exists an efficient quantum algorithm for given access to an oracle for for integer and such that . This implies the hardness for . Although the proof of this assertion works for any , for creating a cryptosystem, the has to be polynomial in .
Peikert proves that there is a probabilistic polynomial time reduction from the problem in the worst case to solving using samples for parameters , , and .
Use in Cryptography
The LWE problem serves as a versatile problem used in construction of several cryptosystems. In 2005, Regev showed that the decision version of LWE is hard assuming quantum hardness of the lattice problems (for as above) and with t=Õ(n/). In 2009, Peikert proved a similar result assuming only the classical hardness of the related problem . The disadvantage of Peikert's result is that it bases itself on a non-standard version of an easier (when compared to SIVP) problem GapSVP.
Regev proposed a public-key cryptosystem based on the hardness of the LWE problem. The cryptosystem as well as the proof of security and correctness are completely classical. The system is characterized by and a probability distribution on . The setting of the parameters used in proofs of correctness and security is
- , a prime number between and .
- for an arbitrary constant
The cryptosystem is then defined by:
- Private Key: Private key is an chosen uniformly at random.
- Public Key: Choose vectors uniformly and independently. Choose error offsets independently according to . The public key consists of
- Encryption: The encryption of a bit is done by choosing a random subset of and then defining as
- Decryption: The decryption of is if is closer to than to , and otherwise.
The proof of correctness follows from choice of parameters and some probability analysis. The proof of security is by reduction to the decision version of LWE: an algorithm for distinguishing between encryptions (with above parameters) of and can be used to distinguish between and the uniform distribution over
Template:Expand section Peikert proposed a system that is secure even against any chosen-ciphertext attack.
- Oded Regev, “On lattices, learning with errors, random linear codes, and cryptography,” in Proceedings of the thirty-seventh annual ACM symposium on Theory of computing (Baltimore, MD, USA: ACM, 2005), 84-93, http://portal.acm.org/citation.cfm?id=1060590.1060603.
- Chris Peikert, “Public-key cryptosystems from the worst-case shortest vector problem: extended abstract,” in Proceedings of the 41st annual ACM symposium on Theory of computing (Bethesda, MD, USA: ACM, 2009), 333-342, http://portal.acm.org/citation.cfm?id=1536414.1536461.
- Chris Peikert and Brent Waters, “Lossy trapdoor functions and their applications,” in Proceedings of the 40th annual ACM symposium on Theory of computing (Victoria, British Columbia, Canada: ACM, 2008), 187-196, http://portal.acm.org/citation.cfm?id=1374406.
- Craig Gentry, Chris Peikert, and Vinod Vaikuntanathan, “Trapdoors for hard lattices and new cryptographic constructions,” in Proceedings of the 40th annual ACM symposium on Theory of computing (Victoria, British Columbia, Canada: ACM, 2008), 197-206, http://portal.acm.org/citation.cfm?id=1374407.