Exploiting Opportunities in Pseudorandomness (NSF CCF-1763299)

Forthcoming
Murtagh, Jack, Omer Reingold, Aaron Sidford, and Salil Vadhan. “Deterministic approximation of random walks in small space.” Theory of Computing Special Issue on RANDOM '19 (Forthcoming). Publisher's VersionAbstract
Version History: v1, 15 Mar. 2019: https://arxiv.org/abs/1903.06361v1
v2 in ArXiv, 25 Nov. 2019: https://arxiv.org/abs/1903.06361v2
 
Prior Published Version (APPROX-RANDOM 2019), 20 Sep 2019: 
In Approximation, Randomization, and Combinatorial Optimization. Algorithms and Techniques (APPROX/RANDOM 2019), Dimitris Achlioptas and László A. Végh (Eds.). Vol. 145. Cambridge, Massachusetts (MIT): Leibniz International Proceedings in Informatics (LIPIcs), 2019.
 

We give a deterministic, nearly logarithmic-space algorithm that given an undirected graph \(G\), a positive integer \(r\), and a set \(S\) of vertices, approximates the conductance of \(S\) in the \(r\)-step random walk on \(G\) to within a factor of \(1+ϵ\), where \(ϵ > 0\) is an arbitrarily small constant. More generally, our algorithm computes an \(ϵ\)-spectral approximation to the normalized Laplacian of the \(r\)-step walk. Our algorithm combines the derandomized square graph operation (Rozenman and Vadhan, 2005), which we recently used for solving Laplacian systems in nearly logarithmic space (Murtagh, Reingold, Sidford, and Vadhan, 2017), with ideas from (Cheng, Cheng, Liu, Peng, and Teng, 2015), which gave an algorithm that is time-efficient (while ours is space-efficient) and randomized (while ours is deterministic) for the case of even \(r\) (while ours works for all \(r\)). Along the way, we provide some new results that generalize technical machinery and yield improvements over previous work. First, we obtain a nearly linear-time randomized algorithm for computing a spectral approximation to the normalized Laplacian for odd \(r\). Second, we define and analyze a generalization of the derandomized square for irregular graphs and for sparsifying the product of two distinct graphs. As part of this generalization, we also give a strongly explicit construction of expander graphs of every size.

TOC 2020.pdf APPROX-RANDOM 2019 ArXiv2019.pdf ArXiv 2019 v2.pdf
2021
Doron, Dean, Raghu Meka, Omer Reingold, Avishay Tal, and Salil Vadhan. “Pseudorandom generators for read-once monotone branching programs.” Electronic Colloquium on Computational Complexity (ECCC) 2021, no. 18 (2021). Publisher's VersionAbstract

Version History: Originally published as "Monotone branching programs: pseudorandomness and circuit complexity". 

Motivated by the derandomization of space-bounded computation, there has been a long line of work on constructing pseudorandom generators (PRGs) against various forms of read-once branching programs (ROBPs), with a goal of improving the \(O(\log^2n)\) seed length of Nisan’s classic construction to the optimal \(O(\log n)\).

In this work, we construct an explicit PRG with seed length \(\tilde{O}(\log n)\) for constant-width ROBPs that are monotone, meaning that the states at each time step can be ordered so that edges with the same labels never cross each other. Equivalently, for each fixed input, the transition functions are a monotone function of the state. This result is complementary to a line of work that gave PRGs with seed length \(O(\log n)\) for (ordered) permutation ROBPs of constant width, since the monotonicity constraint can be seen as the “opposite” of the permutation constraint.

Our PRG also works for monotone ROBPs that can read the input bits in any order, which are strictly more powerful than read-once \(\mathsf{AC^0}\). Our PRG achieves better parameters (in terms of the dependence on the depth of the circuit) than the best previous pseudorandom generator for read-once \(\mathsf{AC^0}\), due to Doron, Hatami, and Hoza.

Our pseudorandom generator construction follows Ajtai and Wigderson’s approach of iterated pseudorandom restrictions. We give a randomness-efficient width-reduction process which proves that the branching program simplifies to an \(O(\log n)\)-junta after only \(O(\log \log n)\) independent applications of the Forbes-Kelley pseudorandom restrictions.

ECCC 2021.pdf ECCC 2021 rev1.pdf
Pyne, Edward, and Salil Vadhan. “Pseudodistributions that beat all pseudorandom generators.” Electronic Colloquium on Computational Complexity (ECCC) 2021, no. 19 (2021). Publisher's VersionAbstract
A recent paper of Braverman, Cohen, and Garg (STOC 2018) introduced the concept of a pseudorandom pseudodistribution generator (PRPG), which amounts to a pseudorandom generator (PRG) whose outputs are accompanied with real coefficients that scale the acceptance probabilities of any potential distinguisher. They gave an explicit construction of PRPGs for ordered branching programs whose seed length has a better dependence on the error parameter  than the classic PRG construction of Nisan (STOC 1990 and Combinatorica 1992).

In this work, we give an explicit construction of PRPGs that achieve parameters that are impossible to achieve by a PRG. In particular, we construct a PRPG for ordered permutation branching programs of unbounded width with a single accept state that has seed length \(\tilde{O}(\log^{3/2}n)\)  for error parameter \( \epsilon = 1/ \mathrm{poly}(n)\), where \(n\) is the input length. In contrast, recent work of Hoza et al. (ITCS 2021) shows that any PRG for this model requires seed length \( \Omega(\log^2n)\) to achieve error \( \epsilon = 1/ \mathrm{poly}(n)\).

As a corollary, we obtain explicit PRPGs with seed length \(\tilde{O}(\log^{3/2}n)\)  and error \( \epsilon = 1/ \mathrm{poly}(n)\) for ordered permutation branching programs of width \(w = \mathrm{poly}(n) \)with an arbitrary number of accept states. Previously, seed length \(o(\log^2n)\) was only known when both the width and the reciprocal of the error are subpolynomial, i.e. \(w= n^{o(1)} \) and \(\epsilon = 1/n^{o(1)}\)(Braverman, Rao, Raz, Yehudayoff, FOCS 2010 and SICOMP 2014).

The starting point for our results are the recent space-efficient algorithms for estimating random-walk probabilities in directed graphs by Ahmadenijad, Kelner, Murtagh, Peebles, Sidford, and Vadhan (FOCS 2020), which are based on spectral graph theory and space-efficient Laplacian solvers. We interpret these algorithms as giving PRPGs with large seed length, which we then derandomize to obtain our results. We also note that this approach gives a simpler proof of the original result of Braverman, Cohen, and Garg, as independently discovered by Cohen, Doron, Renard, Sberlo, and Ta-Shma (personal communication, January 2021).
ECCC 2021.pdf
Hoza, William M., Edward Pyne, and Salil Vadhan. “Pseudorandom generators for unbounded-width permutation branching programs.” 12th Innovations in Theoretical Computer Science (ITCS '21) . Leibniz International Proceedings in Informatics (LIPIcs), 2021. Publisher's VersionAbstract

Version History: 

Preliminary version posted on ECCC TR20-138 (PDF version attached as ECCC 2020).

Talks: The ITCS talk for this paper, presented by Edward Pyne, is currently available on YouTube; click the embedded link to view. 

We prove that the Impagliazzo-Nisan-Wigderson [Impagliazzo et al., 1994] pseudorandom generator (PRG) fools ordered (read-once) permutation branching programs of unbounded width with a seed length of \(\tilde{O} (\log d + \log n ⋅ \log(1/\epsilon))\), assuming the program has only one accepting vertex in the final layer. Here, \(n\) is the length of the program, \(d\) is the degree (equivalently, the alphabet size), and \(\epsilon\) is the error of the PRG. In contrast, we show that a randomly chosen generator requires seed length \(\Omega (n \log d)\) to fool such unbounded-width programs. Thus, this is an unusual case where an explicit construction is "better than random."

Except when the program’s width \(w\) is very small, this is an improvement over prior work. For example, when \(w = \mathrm{poly} (n)\) and \(d = 2\), the best prior PRG for permutation branching programs was simply Nisan’s PRG [Nisan, 1992], which fools general ordered branching programs with seed length \(O (\log (wn/\epsilon) \log n)\). We prove a seed length lower bound of \(\tilde{\Omega} (\log d + \log n ⋅ \log(1/\epsilon)) \)for fooling these unbounded-width programs, showing that our seed length is near-optimal. In fact, when\( \epsilon ≤ 1/\log n\), our seed length is within a constant factor of optimal. Our analysis of the INW generator uses the connection between the PRG and the derandomized square of Rozenman and Vadhan [Rozenman and Vadhan, 2005] and the recent analysis of the latter in terms of unit-circle approximation by Ahmadinejad et al. [Ahmadinejad et al., 2020].

ECCC 2020.pdf ITCS 2021.pdf
2020
Doron, Dean, Jack Murtagh, Salil Vadhan, and David Zuckerman. “Spectral sparsification via bounded-independence sampling.” In 47th International Colloquium on Automata, Languages, and Programming (ICALP 2020), 168:39:1-39:21. Leibniz International Proceedings in Informatics (LIPIcs), Schloss-Dagstuhl-Leibniz-Zentrum für Informatik, 2020. Publisher's VersionAbstract
Version History:

We give a deterministic, nearly logarithmic-space algorithm for mild spectral sparsification of undirected graphs. Given a weighted, undirected graph \(G\) on \(n\) vertices described by a binary string of length \(N\), an integer \(k \leq \log n \) and an error parameter \(\varepsilon > 0\), our algorithm runs in space \(\tilde{O}(k \log(N ^. w_{max}/w_{min}))\) where \(w_{max}\) and \(w_{min}\) are the maximum and minimum edge weights in \(G\), and produces a weighted graph \(H\) with \(\tilde{O}(n^{1+2/k} / \varepsilon^2)\)expected edges that spectrally approximates \(G\), in the sense of Spielmen and Teng [ST04], up to an error of \(\varepsilon\).

Our algorithm is based on a new bounded-independence analysis of Spielman and Srivastava's effective resistance based edge sampling algorithm [SS08] and uses results from recent work on space-bounded Laplacian solvers [MRSV17]. In particular, we demonstrate an inherent tradeoff (via upper and lower bounds) between the amount of (bounded) independence used in the edge sampling algorithm, denoted by \(k\) above, and the resulting sparsity that can be achieved.

ECCC 2020.pdf ICALP 2020.pdf
Ahmadinejad, AmirMahdi, Jonathan Kelner, Jack Murtagh, John Peebles, Aaron Sidford, and Salil Vadhan. “High-precision estimation of random walks in small space.” 61st Annual IEEE Symposium on the Foundations of Computer Science (FOCS 2020). IEEE, 2020. Publisher's VersionAbstract
Version History: 
arXiv version (2019): http://arxiv.org/abs/1912.04524
 
Talks: View a talk on this paper presented by by John Peebles at FOCS 2020.
 
In this paper, we provide a deterministic \(\tilde{O}(\log N)\)-space algorithm for estimating the random walk probabilities on Eulerian directed graphs (and thus also undirected graphs) to within inverse polynomial additive error \((ϵ = 1/\mathrm{poly}(N)) \) where \(N\) is the length of the input. Previously, this problem was known to be solvable by a randomized algorithm using space \(O (\log N)\) (Aleliunas et al., FOCS '79) and by a deterministic algorithm using space \(O (\log^{3/2} N)\) (Saks and Zhou, FOCS '95 and JCSS '99), both of which held for arbitrary directed graphs but had not been improved even for undirected graphs. We also give improvements on the space complexity of both of these previous algorithms for non-Eulerian directed graphs when the error is negligible \((ϵ=1/N^{ω(1)})\), generalizing what Hoza and Zuckerman (FOCS '18) recently showed for the special case of distinguishing whether a random walk probability is 0 or greater than ϵ.

We achieve these results by giving new reductions between powering Eulerian random-walk matrices and inverting Eulerian Laplacian matrices, providing a new notion of spectral approximation for Eulerian graphs that is preserved under powering, and giving the first deterministic \(\tilde{O}(\log N)\)-space algorithm for inverting Eulerian Laplacian matrices. The latter algorithm builds on the work of Murtagh et al. (FOCS '17) that gave a deterministic \(\tilde{O}(\log N)\)-space algorithm for inverting undirected Laplacian matrices, and the work of Cohen et al. (FOCS '19) that gave a randomized \(\tilde{O} (N)\)-time algorithm for inverting Eulerian Laplacian matrices. A running theme throughout these contributions is an analysis of "cycle-lifted graphs," where we take a graph and "lift" it to a new graph whose adjacency matrix is the tensor product of the original adjacency matrix and a directed cycle (or variants of one).
ARXIV 2019.pdf FOCS 2020.pdf
Haitner, Iftach, Thomas Holenstein, Omer Reingold, Salil Vadhan, and Hoeteck Wee. “Inaccessible entropy II: IE functions and universal one-way hashing.” Theory of Computing 16, no. 8 (2020): 1-55. Publisher's VersionAbstract

Version History: published earlier in Henri Gilbert, ed., Advances in Cryptology—EUROCRYPT ‘10, Lecture Notes on Computer Science, as "Universal one-way hash functions via inaccessible entropy": 

https://link.springer.com/chapter/10.1007/978-3-642-13190-5_31

 

This paper revisits the construction of Universal One-Way Hash Functions (UOWHFs) from any one-way function due to Rompel (STOC 1990). We give a simpler construction of UOWHFs, which also obtains better efficiency and security. The construction exploits a strong connection to the recently introduced notion of inaccessible entropy (Haitner et al. STOC 2009). With this perspective, we observe that a small tweak of any one-way function \(f\) is already a weak form of a UOWHF: Consider \(F(x', i)\) that outputs the \(i\)-bit long prefix of \(f(x)\). If \(F\) were a UOWHF then given a random \(x\) and \(i\) it would be hard to come up with \(x' \neq x\) such that \(F(x, i) = F(x', i)\). While this may not be the case, we show (rather easily) that it is hard to sample \(x'\) with almost full entropy among all the possible such values of \(x'\). The rest of our construction simply amplifies and exploits this basic property.

With this and other recent works, we have that the constructions of three fundamental cryptographic primitives (Pseudorandom Generators, Statistically Hiding Commitments and UOWHFs) out of one-way functions are to a large extent unified. In particular, all three constructions rely on and manipulate computational notions of entropy in similar ways. Pseudorandom Generators rely on the well-established notion of pseudoentropy, whereas Statistically Hiding Commitments and UOWHFs rely on the newer notion of inaccessible entropy.

EUROCRYPT2010.pdf ToC 2020.pdf
2019
Agrawal, Rohit, Yi-Hsiu Chen, Thibaut Horel, and Salil Vadhan. “Unifying computational entropies via Kullback-Leibler divergence.” In Advances in Cryptology: CRYPTO 2019, A. Boldyreva and D. Micciancio, (Eds), 11693:831-858. Springer Verlag, Lecture Notes in Computer Science, 2019. Publisher's VersionAbstract
Version History: 
arXiv, first posted Feb 2019, most recently updated Aug 2019: https://arxiv.org/abs/1902.11202
 
We introduce hardness in relative entropy, a new notion of hardness for search problems which on the one hand is satisfied by all one-way functions and on the other hand implies both next-block pseudoentropy and inaccessible entropy, two forms of computational entropy used in recent constructions of pseudorandom generators and statistically hiding commitment schemes, respectively. Thus, hardness in relative entropy unifies the latter two notions of computational entropy and sheds light on the apparent “duality” between them. Additionally, it yields a more modular and illuminating proof that one-way functions imply next-block inaccessible entropy, similar in structure to the proof that one-way functions imply next-block pseudoentropy (Vadhan and Zheng, STOC ‘12).
CRYPTO 2019.pdf ArXiv2019.pdf