Speaker: Mohammad Mahmoody, Assistant Professor, University of Virginia

Speaker: Avishay Tal

Affiliation: IAS

Title:Time-Space Hardness of Learning Sparse Parities

Abstract:

How can one learn a parity function, i.e., a function of the form $f(x) = a_1 x_1 + a_2 x_2 + … + a_n x_n (mod 2)$ where a_1, …, a_n are in {0,1}, from random labeled examples? One approach is to gather O(n) random labeled examples and perform Gaussian-elimination. This requires a memory of size O(n^2) and poly(n) time. Another approach is to go over all possible 2^n parity functions and to verify them by checking O(n) random examples per each possibility. This requires a memory of size O(n), but O(2^n * n) time. In a recent work, Raz [FOCS, 2016] showed that if an algorithm has memory of size much smaller than n^2, then it has to spend exponential time in order to learn a parity function. In other words, fast learning requires a good memory. In this work, we show that even if the parity function is known to be extremely sparse, where only log(n) of the a_i’s are nonzero, then the learning task is still time-space hard. That is, we show that any algorithm with linear size memory and polynomial time fails to learn log(n)-sparse parities. Consequently, the classical tasks of learning linear-size DNF formulae, linear-size decision trees, and logarithmic-size juntas are all time-space hard. Based on joint work with Gillat Kol and Ran Raz.

SPEAKER: Yevgeniy Dodis, New York University

TITLE: Fixing Cracks in the Concrete: Random Oracles with Auxiliary Input, Revisited

ABSTRACT: We revisit security proofs for various cryptographic primitives in the random oracle model with auxiliary input (ROM-AI): a (computationally unbounded) attacker A can compute arbitrary S bits of leakage z=z(O) about the random oracle O before attacking the system, and then use additional T oracle queries to O during the attack. This model was explicitly studied by Unruh in 2007 (CRYPTO 2007), but dates back to the seminal paper of Hellman in 1980 about time-space tradeoffs for inverting random functions, and has natural applications in settings where traditional random oracle proofs are not useful: (a) security against non-uniform attackers; (b) space-time tradeoffs; (c) security against preprocessing; (d) resilience to backdoors in hash functions. We obtain a number of new results about ROM-AI, but our main message is that ROM-AI is the “new cool kid in town”: it nicely connects theory and practice, has a lot of exciting open questions, leads to beautiful math, short definitions, elegant proofs, surprising algorithms, and is still in its infancy. In short, you should work on it! Joint Work with Siyao Guo and Jonathan Katz.

Speaker: Ori Rottenstriech, Princeton

Title: Novel Approaches to Challenges in Emerging Network Paradigms

Abstract:

SDN (Software defined networking) and NFV (Network Function Virtualization) are two emerging network paradigms that enable simplification, flexibility and cost-reduction in network management. We believe that the new paradigms will lead to many interesting research questions. We study how to rely on them for dealing with two common network challenges.

We consider switches that imply network policies in SDN through rule matching tables of limited size. We study the applicability of rule caching and lossy compression to create packet classifiers requiring much less memory than the theoretical size limits of semantically-equivalent representations. We would like to find limited-size classifiers that can correctly classify a high portion of the traffic. We address different goals with unique settings and explain how to deal with the traffic that cannot be classified correctly.

Network functions such as load balancing and deep packet inspection are often implemented in dedicated hardware called middleboxes. Those can suffer from temporary unavailability due to misconfiguration or software and hardware malfunction. We suggest to rely on virtualization for planning and deploying backup schemes for network functions. The schemes guarantee high levels of survivability with significant reduction in resource consumption. We discuss different goals that network designers should take into account. We describe a graph theoretical model for finding properties of efficient solutions and developing algorithms that can build them.

Bio: Ori Rottenstriech is a postdoctoral research associate at the Department of Computer Science, Princeton University. He received his Ph.D. from the Electrical Engineering department of the Technion. His research interests include the intersection of computer networking and algorithms.

Speaker: Sepehr Assadi, UPenn

Title:

Abstract:

Speaker: Dana Dachman Soled, UMD

Speaker: Mohammad Hajiesmaili
Affiliation: Johns Hopkins University

Title: Online storage management in electricity market

Abstract:

With unprecedented benefits in terms of efficiency, economy, reliability, and environmental awareness, in the recent years, there has been a rapid proliferation of renewable energy sources such as solar and wind in electric power systems. Despite these benefits, the inherent uncertainty in renewables places severe challenges on the management of the entire energy systems, including electricity market. Leveraging energy storage systems is a promising approach to mitigate the uncertainty of renewables, by charging and discharging during the mismatched periods. Energy storage systems, however, offers a new design space for additional optimization. That is, a storage system can capture energy during periods when the market prices are low and surrender stored energy when energy prices are high.
In this talk, we consider different scenarios of storage management in both supply and demand sides of the electricity market. The uncertainties in both renewable output and electricity market price, emphasizes the need for online solution design. The underlying theoretical problems could be described as extensions of conversion problems in financial markets, i.e., the search for best prices to buy and/or sell assets. The difference with the conversion problems, is that in addition to the uncertainty in the price, our problems suffer from another uncertainty originated from renewable output. We follow online algorithm design and use competitive ratio as the performance measure of our algorithms. We present our recent results in designing competitive online algorithms that achieve constant competitive ratios. In addition, we briefly talk about the case of utilizing aggregate potentials distributed small-scale storage systems, such as EVs or residential storages, to participate in electricity market through an aggregator. This setting is more challenging than the previous one, since the distributed sources also arrive in online manner with heterogeneous profiles.

Overall, we believe that changing the landscape of electric power system from a centralized predictable system to a distributed uncertain system opens a new research direction for leveraging online framework designs in this relatively under-explored area.

Speaker: Kuan Cheng
Affiliation: Johns Hopkins University

Title: Near-Optimal Secret Sharing and Error Correcting Codes in $\AC^0$

Abstract:

We study the question of minimizing the computational complexity of (robust) secret sharing schemes and error correcting codes. In standard instances of these objects, both encoding and decoding involve linear algebra, and thus cannot be implemented in the class $\AC^0$. The feasibility of non-trivial secret sharing schemes in $\AC^0$ was recently shown by Bogdanov et al.\ (Crypto 2016) and that of (locally) decoding errors in $\AC^0$ by Goldwasser et al.\ (STOC 2007).

In this paper, we show that by allowing some slight relaxation such as a small error probability, we can construct much better secret sharing schemes and error correcting codes in the class $\AC^0$. In some cases, our parameters are close to optimal and would be impossible to achieve without the relaxation. Our results significantly improve previous constructions in various parameters.

Our constructions combine several ingredients in pseudorandomness and combinatorics in an innovative way. Specifically, we develop a general technique to simultaneously amplify security threshold and reduce alphabet size, using a two-level concatenation of protocols together with a random permutation. We demonstrate the broader usefulness of this technique by applying it in the context of a variant of secure broadcast.

Based on a joint work with Yuval Ishai and Xin Li.

Speaker: Ilan Komargodski
Affiliation: Cornell Tech

Title: White-Box vs. Black-Box Complexity of Search Problems: Ramsey and Graph Property Testing

Abstract: Ramsey theory assures us that in any graph there is a clique or independent set of a certain size, roughly logarithmic in the graph size. But how difficult is it to find the clique or independent set? This problem is in TFNP, the class of search problems with guaranteed solutions.  If the graph is given explicitly, then it is possible to do so while examining a linear number of edges. If the graph is given by a black-box, where to figure out whether a certain edge exists the box should be queried, then a large number of queries must be issued.

1) What if one is given a program or circuit (“white-box”) for computing the existence of an edge. Does the search problem remain hard?
2) Can we generically translate all TFNP black-box hardness into white-box hardness?
3) Does the problem remain hard if the black-box instance is small?

We will answer all of these questions and discuss related questions in the setting of property testing.

Joint work with Moni Naor and Eylon Yogev.

Speaker: Ran Ben Basat
Affiliation: Technion

Title: Classic Network Measurement meets Virtual Switching

Abstract: In modern cloud infrastructures, each physical server often runs multiple virtual machines and employs a software Virtual Switch (VS) to handle their traffic. In addition to switching, the VS performs network measurements, such as identifying the most frequent flows, which are essential for networking applications such as load balancing and intrusion detection.

Unlike traditional streaming algorithms, which minimize the space requirements, the bottleneck in virtual switching measurement is the CPU utilization. In this talk, I will present new hardware-oriented algorithms and acceleration methods that optimize the update time for software, at the cost of a slight memory overhead.

Bio: Ran is a Ph.D. candidate at the Technion, Israel. He does research in streaming algorithms for networking applications, focusing on efficient processing and query speeds.

Speaker: Samson Zhou
Affiliation: Purdue University

Title: Pattern Matching over Noisy Data Streams

Abstract: The identification of low-complexity structure in strings is a fundamental building block for many algorithms in computational biology or natural language processing. The general paradigm in these algorithms is to find either highly repetitive structure, in the form of periodicity or palindromes in a pre-processing stage, to filter out locations where a certain pattern cannot occur, thus improving efficiency.

Unfortunately, we must expect massive data to contain a number of small imperfections, such as through human error or mutations. This motivates the need to study structure in models of sublinear space, resilient to sources of noise. In this talk, we introduce several types of structure and noise, and discuss efficient algorithms to identify these structures over data streams.

As a warm-up, we provide an algorithm for identifying a longest common aligned substring of two inputs, resilient up to d errors of insertions, substitutions, or deletions. We then present a streaming algorithm for identifying the longest palindrome, resilient up to a threshold of d substitution errors. Finally, we discuss the problem of finding all periods of a string including up to d persistent changes or erasures. For each of these scenarios, we also provide complementary lower bounds.

Joint work with Funda Ergun, Elena Grigorescu, and Erfan Sadeqi Azer.

BIO:
Samson is a PhD candidate in the Department of Computer Science at Purdue University, under the supervision of Greg Frederickson and Elena Grigorescu. He received his undergraduate education at MIT, where he obtained a Bachelor’s in math and computer science, as well as a Master’s in computer science. He is a member of the Theory Group at Purdue, and his current research interests are sublinear and approximation algorithms, with an emphasis on streaming algorithms.

Speaker: Venkata Gandikota
Affiliation: Johns Hopkins University

Title: NP-Hardness of Reed-Solomon Decoding and the Prouhet-Tarry-Escott Problem

Abstract: Establishing the complexity of Bounded Distance Decoding for Reed-Solomon codes is a fundamental open problem in coding theory, explicitly asked by Guruswami and Vardy (IEEE Trans. Inf. Theory, 2005). The problem is motivated by the large current gap between the regime when it is NP-hard, and the regime when it is efficiently solvable (i.e., the Johnson radius).

We show the first NP-hardness results for asymptotically smaller decoding radii than the maximum likelihood decoding radius of Guruswami and Vardy. Specifically, for Reed-Solomon codes of length N and dimension K = O(N), we show that it is NP-hard to decode more than N-K-O(log N / log log N) errors.

These results follow from the NP-hardness of a generalization of the classical Subset Sum problem to higher moments, called Moments Subset Sum, which has been a known open problem, and which may be of independent interest. We further reveal a strong connection with the well-studied Prouhet-Tarry-Escott problem in Number Theory, which turns out to capture a main barrier in extending our techniques. We believe the Prouhet-Tarry-Escott problem deserves further study in the theoretical computer science community.

This is a joint work with Badih Ghazi (MIT) and Elena Grigorescu (Purdue).

Speaker: Samson Zhou
Affiliation: Purdue University

Title: Password Hashing and Graph Pebbling

Abstract: Although the passwords of users are no longer being stored, we show an offline attacker is compelled to crack all stolen passwords under current security recommendations. Memory hard functions have been proposed as moderately expensive cryptographic tools for password hashing. The cryptanalysis of these functions has focused on the cumulative memory complexity and the energy complexity of the function. The first metric measures how much memory users must commit to evaluating a function, while the second metric measures how much energy users must commit. We prove these evaluations reduce to pebbling games on graphs but show that a tool for exact cryptanalysis of functions is unlikely to exist. Nevertheless, we provide asymptotic upper and lower bounds on several families of functions, including Argon2i, the winner of the password hashing competition that is currently being considered for standardization by the Cryptography Form Research Group of the Internet Research Task Force.

Joint work with Jeremiah Blocki, Ben Harsha, Ling Ren

BIO:
Samson is a PhD candidate in the Department of Computer Science at Purdue University, under the supervision of Greg Frederickson and Elena Grigorescu. He received his undergraduate education at MIT, where he obtained a Bachelor’s in math and computer science, as well as a Master’s in computer science. He is a member of the Theory Group at Purdue and a winner of the Sigma Xi Research Awards Competition for graduate students in engineering. His current research interests are sublinear and approximation algorithms, with an emphasis on streaming algorithms.

Speaker: Sai Lakshmi Bhavana Obbattu
Affiliation: IISC Bangalore, India

Title: Privacy Amplification from Non-malleable Codes

The goal of a Privacy Amplification (PA) protocol is to allow two parties, who start out sharing a non-uniform secret ‘w’, to agree on a uniform secret ‘k’, in the presence of a computationally unbounded man-in-the-middle adversary. An interactive PA protocol is rated based on three parameters: 1) Number of rounds, 2) Entropy loss (entropy of w – |k|), and 3) Min-entropy requirement for w, while the asymptotically optimal parameters are 2, O(s) and O(s+log n) respectively (where s is the security parameter and n =|w|). There have been two popular approaches to solve this problem: one using use bit authentication protocols and the other using non-malleable extractors, but none of the prior protocols using these approaches had all asymptotically optimal parameters.

We give an alternate approach to solve the problem using Non-malleable Codes (NMCs). This approach results in a 8-round protocol with min-entropy requirement O(s+log n) and an entropy loss of O(s log s). Augmented NMCs with better parameters would result in optimal entropy loss of O(s). Our result is one of the first information theoretic applications of NMCs. In this talk, I will introduce NMCs and show connection of NMCs to PA.

In a concurrent and independent work, Xin Li gives a protocol with asymptotically optimal parameters based on non-malleable extractors. Because all known approaches have large hidden constants, exploring alternatives is necessary if we hope to get practical concrete parameters

The talk is based on:

Eshan Chattopadhyay, Bhavana Kanukurthi, Sai Lakshmi Bhavana Obbattu and Sruthi Sekar. Privacy Amplification from Non-malleable Codes. (eprint.iacr.org/2018/293)

Bhavana Kanukurthi, Sai Lakshmi Bhavana Obbattu and Sruthi Sekar. Non-malleable Randomness Encoders and their Applications (Eurocrypt 2018)

Bio: Sai Lakshmi Bhavana Obbattu is a doctoral student at Indian Institute of Science(IISc), Bangalore, advised by Dr. Bhavana Kanukurthi. Her publication venues include the Theory of Cryptography Conference (TCC) and Eurocrypt. Her TCC publication on Four-state Non-malleable Codes was invited to the Journal of Cryptology. She received her Integrated Dual Degree (B.Tech and M.Tech) from IIT(BHU), Varanasi. Her research interests include Non-malleable codes, Privacy Amplification and Applied Multi-party computation.

Speaker: Akash Kumar
Affiliation: Purdue University
Location: Malone 338 (note change of location)

Title:
Finding Minors in Sublinear time in Bounded degree graphs with (almost) optimal one-sided query complexity.

Abstract:
Let G be an undirected, bounded degree graph with n vertices. Fix a finite graph H, and suppose one must remove \varepsilon n edges from G to make it H-minor free (for some small constant \varepsilon > 0). We give an n^{1/2+o(1)}-time randomized procedure that, with high probability, finds an H-minor in such a graph. For an example application, suppose one must remove \varepsilon n edges from a bounded degree graph G to make it planar. This result implies an algorithm, with the same running time, that produces a K_{3,3} or K_5 minor in G. No sublinear time bound was known for this problem, prior to this result.

By the graph minor theorem, we get an analogous result for any minor-closed property. Up to n^{o(1)} factors, this resolves a conjecture of Benjamini-Schramm-Shapira (STOC 2008) on the existence of one-sided property testers for minor-closed properties. Furthermore, our algorithm is nearly optimal, by an \Omega(\sqrt{n}) lower bound of Czumaj et al (RSA 2014).

Prior to this work, the only graphs H for which non-trivial property testers were known for H-minor freeness are the following: H being a forest or a cycle (Czumaj et al, RSA 2014), K_{2,k}, (k\times 2)-grid, and the k-circus (Fichtenberger et al, Arxiv 2017).

(Joint work with C. Seshadhri and Andrew Stolman).

Welcome and Introductions

Speaker: Zhengzhong Jin
Affiliation: JHU

Title: Deterministic Document Exchange Protocols, and Almost Optimal Binary Codes for Edit Errors

Abstract:
We study two basic problems regarding edit error, i.e. document exchange and error correcting codes for edit errors (insdel codes). For message length n and edit error upper bound k, it is known that in both problems the optimal sketch size or the optimal number of redundant bits is Θ(k log n/k). However, known constructions are far from achieving these bounds.

We significantly improve previous results on both problems. For document exchange, we give an efficient deterministic protocol with sketch size O(k log^2 n/k). This significantly improves the previous best known deterministic protocol, which has sketch size O(k^2+k log^2 n). For binary insdel codes, we obtain the following results:

1. An explicit binary insdel code which encodes an n-bit message x against k errors with redundancy O(k log^2 n/k). In particular this implies an explicit family of binary insdel codes that can correct ε fraction of insertions and deletions with rate 1−O(ε log^2(1/ε))=1−\tilde {O}(ε).

2. An explicit binary insdel code which encodes an n-bit message x against k errors with redundancy O(k log n). This is the first explicit construction of binary insdel codes that has optimal redundancy for a wide range of error parameters k, and this brings our understanding of binary insdel codes much closer to that of standard binary error correcting codes.

In obtaining our results we introduce the notion of ε-self matching hash functions and ε-synchronization hash functions. We believe our techniques can have further applications in the literature.

Speaker: Marius Zimand
Affiliation: Towson University

Title: An operational characterization of mutual information in algorithmic information theory

Abstract: An operational interpretation of the concept of mutual information in the framework of Kolmogorov complexity has been elusive till now. We show that the mutual information of any pair of strings x and y is equal, up to logarithmic precision, to the length of the longest shared secret key that two parties, one having x and the complexity profile of the pair and the other one having y and the complexity profile of the pair, can establish via a probabilistic protocol with interaction on a public channel. We establish the communication complexity of secret key agreement protocols that produce a secret key of maximal length, for protocols with public randomness. We show that if the communication complexity drops below the established threshold then only very short secret keys can be obtained.

This is joint work with Andrei Romashchenko.

Speaker: Yasamin Nazari
Affiliation: JHU

Title: Distributed Distance-Bounded Network Design Through Distributed Convex Programming

Abstract:
Solving linear programs is often a challenging task in distributed settings. While there are good algorithms for solving packing and covering linear programs in a distributed manner, this is essentially the only class of linear programs for which such an algorithm is known. In this work we provide a distributed algorithm for solving a different class of convex programs which we call “distance-bounded network design convex programs”. These can be thought of as relaxations of network design problems in which the connectivity requirement includes a distance constraint (most notably, graph spanners). Our algorithm runs in O((D/ϵ)logn) rounds in the LOCAL model and finds a (1+ϵ)-approximation to the optimal LP solution for any 0<ϵ≤1, where D is the largest distance constraint. While solving linear programs in a distributed setting is interesting in its own right, this class of convex programs is particularly important because solving them is often a crucial step when designing approximation algorithms. Hence we almost immediately obtain new and improved distributed approximation algorithms for a variety of network design problems, including Basic 3- and 4-Spanner, Directed k-Spanner, Lowest Degree k-Spanner, and Shallow-Light Steiner Network Design with a spanning demand graph. Our algorithms do not require any "heavy" computation and essentially match the best-known centralized approximation algorithms, while previous approaches which do not use heavy computation give approximations which are worse than the best-known centralized bounds.