We propose a new general primitive called lossy trapdoor functions (lossy TDFs), and realize it under a variety of different number theoretic assumptions, including hardness of the decisional Diffie-Hellman (DDH) problem and the worst-case hardness of standard lattice problems. Using lossy TDFs, we develop a new approach for constructing many important cryptographic primitives, including standard trapdoor functions, CCA-secure cryptosystems, collisionresistant hash functions, and more. All of our constructions are simple, efficient, and black-box. Taken all together, these results resolve some long-standing open problems in cryptography. They give the first known (injective) trapdoor functions based on problems not directly related to integer factorization, and provide the first known CCA-secure cryptosystem based solely on worst-case lattice assumptions.

We show that finding small solutions to random modular linear equations is at least as hard as approximating several lattice problems in the worst case within a factor almost linear in the dimension of the lattice. The lattice problems we consider are the shortest vector problem, the shortest independent vectors problem, the covering radius problem, and the guaranteed distance decoding problem (a variant of the well known closest vector problem). The approximation factor we obtain is nlog O(1) n for all four problems. This greatly improves on all previous work on the subject starting from Ajtai’s seminal paper (STOC, 1996), up to the strongest previously known results by Micciancio (SIAM J. on Computing, 2004). Our results also bring us closer to the limit where the problems are no longer known to be in NP intersect coNP. Our main tools are Gaussian measures on lattices and the high-dimensional Fourier transform. We start by defining a new lattice parameter which determines the amount of Gaussian noise that one has to add to a lattice in order to get close to a uniform distribution. In addition to yielding quantitatively much stronger results, the use of this parameter allows us to simplify many of the complications in previous work. Our technical contributions are two-fold. First, we show tight connections between this new parameter and existing lattice parameters. One such important connection is between this parameter and the length of the shortest set of linearly independent vectors. Second, we prove that the distribution that one obtains after adding Gaussian noise to the lattice has the following interesting property: the distribution of the noise vector when conditioning on the final value behaves in many respects like the original Gaussian noise vector. In particular, its moments remain essentially unchanged. 1

We show how to construct a variety of “trapdoor ” cryptographic tools assuming the worstcase hardness of standard lattice problems (such as approximating the shortest nonzero vector to within small factors). The applications include trapdoor functions with preimage sampling, simple and efficient “hash-and-sign ” digital signature schemes, universally composable oblivious transfer, and identity-based encryption. A core technical component of our constructions is an efficient algorithm that, given a basis of an arbitrary lattice, samples lattice points from a Gaussian-like probability distribution whose standard deviation is essentially the length of the longest vector in the basis. In particular, the crucial security property is that the output distribution of the algorithm is oblivious to the particular geometry of the given basis. ∗ Supported by the Herbert Kunzel Stanford Graduate Fellowship. † This material is based upon work supported by the National Science Foundation under Grants CNS-0716786 and CNS-0749931. Any opinions, findings, and conclusions or recommedations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation. ‡ The majority of this work was performed while at SRI International. 1 1

We construct public-key cryptosystems that are secure assuming the worst-case hardness of approximating the length of a shortest nonzero vector in an n-dimensional lattice to within a small poly(n) factor. Prior cryptosystems with worst-case connections were based either on the shortest vector problem for a special class of lattices (Ajtai and Dwork, STOC 1997; Regev, J. ACM 2004), or on the conjectured hardness of lattice problems for quantum algorithms (Regev, STOC 2005). Our main technical innovation is a reduction from certain variants of the shortest vector problem to corresponding versions of the “learning with errors ” (LWE) problem; previously, only a quantum reduction of this kind was known. In addition, we construct new cryptosystems based on the search version of LWE, including a very natural chosen ciphertext-secure system that has a much simpler description and tighter underlying worst-case approximation factor than prior constructions. Keywords: Lattice-based cryptography, learning with errors, quantum computation

Abstract The generalized knapsack function is defined as fa(x) = Pi ai * xi, where a = (a1,..., am)consists of m elements from some ring R, and x = (x1,..., xm) consists of m coefficients froma specified subset S ` R. Micciancio (FOCS 2002) proposed a specific choice of the ring R andsubset S for which inverting this function (for random a, x) is at least as hard as solving certainworst-case problems on cyclic lattices. We show that for a different choice of S ae R, the generalized knapsack function is in factcollision-resistant, assuming it is infeasible to approximate the shortest vector in n-dimensionalcyclic lattices up to factors ~ O(n). For slightly larger factors, we even get collision-resistancefor any m> = 2. This yields very efficient collision-resistant hash functions having key size andtime complexity almost linear in the security parameter n. We also show that altering S isnecessary, in the sense that Micciancio's original function is not collision-resistant (nor even universal one-way).Our results exploit an intimate connection between the linear algebra of n-dimensional cycliclattices and the ring Z [ ff]/(ffn- 1), and crucially depend on the factorization of ffn- 1 intoirreducible cyclotomic polynomials. We also establish a new bound on the discrete Gaussian distribution over general lattices, employing techniques introduced by Micciancio and Regev(FOCS 2004) and also used by Micciancio in his study of compact knapsacks. 1 Introduction A function family {fa}a2A is said to be collision-resistant if given a uniformly chosen a 2 A, it is infeasible to find elements x1 6 = x2 so that fa(x1) = fa(x2). Collision-resistant hash functions are one of the most widely-employed cryptographic primitives. Their applications include integrity checking, user and message authentication, commitment protocols, and more. Many of the applications of collision-resistant hashing tend to invoke the hash function only a small number of times. Thus, the efficiency of the function has a direct effect on the efficiency of the application that uses it. This is in contrast to primitives such as one-way functions, which typically must be invoked many times in their applications (at least when used in a black-box way) [9].

In recent years, several papers have established limits on the computational difficulty of lattice problems, focusing primarily on the ℓ2 (Euclidean) norm. We demonstrate close analogues of these results in ℓp norms, for every 2 < p ≤ ∞. In particular, for lattices of dimension n: • Approximating the closest vector problem, the shortest vector problem, and other related problems to within O ( √ n) factors (or O ( √ n log n) factors, for p = ∞) is in coNP. • Approximating the closest vector and bounded distance decoding problems with preprocessing to within O ( √ n) factors can be accomplished in deterministic polynomial time. • Approximating several problems (such as the shortest independent vectors problem) to within Õ(n) factors in the worst case reduces to solving the average-case problems defined in prior works (Ajtai, STOC 1996; Micciancio and Regev, SIAM J. on Computing 2007; Regev, STOC 2005). Our results improve prior approximation factors for ℓp norms by up to √ n factors. Taken all together, they complement recent reductions from the ℓ2 norm to ℓp norms (Regev and Rosen, STOC 2006), and provide some evidence that lattice problems in ℓp norms (for p> 2) may not be substantially harder than they are in the ℓ2 norm. One of our main technical contributions is a very general analysis of Gaussian distributions over lattices, which may be of independent interest. Our proofs employ analytical techniques of Banaszczyk that, to our knowledge, have yet to be exploited in computer science. 1

Abstract We demonstrate an average-case problem which is as hard as finding fl(n)-approximateshortest vectors in certain n-dimensional lattices in the worst case, where fl(n) = O(plog n).The previously best known factor for any class of lattices was fl(n) = ~O(n).To obtain our results, we focus on families of lattices having special algebraic structure. Specifically, we consider lattices that correspond to ideals in the ring of integers of an algebraicnumber field. The worst-case assumption we rely on is that in some `p length, it is hard to findapproximate shortest vectors in these lattices, under an appropriate form of preprocessing of the number field. Our results build upon prior works by Micciancio (FOCS 2002), Peikert andRosen (TCC 2006), and Lyubashevsky and Micciancio (ICALP 2006). For the connection factors fl(n) we achieve, the corresponding decisional promise problemson ideal lattices are not known to be NP-hard; in fact, they are in P. However, the search approximation problems still appear to be very hard. Indeed, ideal lattices are well-studiedobjects in computational number theory, and the best known algorithms for them seem to perform no better than the best known algorithms for general lattices.To obtain the best possible connection factor, we instantiate our constructions with infinite families of number fields having constant root discriminant. Such families are known to existand are computable, though no efficient construction is yet known. Our work motivates the search for such constructions. Even constructions of number fields having root discriminant upto O(n2/3-ffl) would yield connection factors better than the current best of ~O(n).

In this chapter we describe some of the recent progress in lattice-based cryptography. Lattice-based cryptographic constructions hold a great promise for post-quantum cryptography, as they enjoy very strong security proofs based on worst-case hardness, relatively efficient implementations, as well as great simplicity. In addition, lattice-based cryptography is believed to be secure against quantum computers. Our focus here

We prove the equivalence, up to a small polynomial approximation factor p n / log n, of the lattice problems uSVP (unique Shortest Vector Problem), BDD (Bounded Distance Decoding) and GapSVP (the decision version of the Shortest Vector Problem). This resolves a long-standing open problem about the relationship between uSVP and the more standard GapSVP, as well the BDD problem commonly used in coding theory. The main cryptographic application of our work is the proof that the Ajtai-Dwork ([AD97]) and the Regev ([Reg04a]) cryptosystems, which were previously only known to be based on the hardness of uSVP, can be equivalently based on the hardness of worst-case GapSVP O(n 2.5) and GapSVP O(n 2), respectively. Also, in the case of uSVP and BDD, our connection is very tight, establishing the equivalence (within a small constant approximation factor) between the two most central problems used in lattice based public key cryptography and coding theory. 1

We introduce a computational problem of distinguishing between two specific quantum states as a new cryptographic problem to design a quantum cryptographic scheme that is “secure ” against any polynomial-time quantum adversary. Our problem QSCDff is to distinguish between two types of random coset states with a hidden permutation over the symmetric group of finite degree. This naturally generalizes the commonly-used distinction problem between two probability distributions in computational cryptography. As our major contribution, we show three cryptographic properties: (i) QSCDff has the trapdoor property; (ii) the average-case hardness of QSCDff coincides with its worst-case hardness; and (iii) QSCDff is computationally at least as hard in the worst case as the graph automorphism problem. These cryptographic properties enable us to construct a quantum public-key cryptosystem, which is likely to withstand any chosen plaintext attack of a polynomialtime quantum adversary. We further discuss a generalization of QSCDff, called QSCDcyc, and introduce a multi-bit encryption scheme relying on the cryptographic properties of QSCDcyc.