ISIT 2019

230 papers
The 2019 IEEE International Symposium on Information Theory (ISIT) featured 244 papers covering a broad spectrum of information and coding theory. A primary focus was placed on secure computation through non-colluding server architectures and private information retrieval (PIR) for distributed storage and sublinear search. Research also emphasized non-convex signal recovery, specifically addressing global convergence landscapes, spectral initialization, and permutation recovery in unlabeled sensing. Additional trends included the application of information-theoretic principles to machine learning, such as federated learning with heterogeneous privacy constraints and MAP inference in deep networks. Coding for emerging channel models was also prominent, with work on duplication and substitution channels, as well as energy-efficient random access for fading channels. Theoretical refinements in estimation were explored through Bayesian Cramér-Rao bounds and the relationship between mutual information and moments. Prolific contributors to this year's proceedings included Vincent Y. F. Tan, Yury Polyanskiy, Salman Avestimehr, and Deniz Gündüz.

Top Research Directions

Research directions most represented by ISIT 2019 papers.

Non-Collusion Assumptions for Efficient Multi-Server Privacy 4
Straggler-Resilient Coded Computation 3
Single Letter Characterization of Information Theoretic Limits 2
Normalized Delivery Time in Fog Radio Access Networks 2
Beyond the High SNR Asymptotic Assumption 2
Information Theoretic Capacity Limits of Private Data Retrieval 2
Bandwidth-Efficient Repair of Concurrent Node Failures 2
Optimizing Repair Bandwidth for Distributed Erasure Codes 2
Quantifying Privacy Leakage with Information Theory 2
Mitigating the Straggler Bottleneck in Distributed Learning 2
The Empirical Validation Gap 2
Decoupling Statistical Dependencies via Poissonization 1
Information Contraction via Strong Data Processing Inequalities 1
Numerical Stability in Coded Distributed Computing 1
The Replica Method for High Dimensional Inference 1

Top Authors

AuthorPapers
Yury Polyanskiy9
Deniz Gunduz6
Vincent Y. F. Tan6
Christina Fragouli5
Salman Avestimehr5
Antonia Wachter-Zeh4
Farhad Shirani4
Flavio Calmon4
Jaekyun Moon4
Martina Cardone4
Ali Tajer3
Ayfer Ozgur3
Jonathan Scarlett3
Jy-yong Sohn3
Mehul Motani3
Nir Shlezinger3
Ofer Shayevitz3
Or Ordentlich3
Rajai Nasser3
Swanand Kadhe3

Impactful Papers

Papers from ISIT 2019 that shaped research clusters.

Foundational

Influential

Papers

Title
Synthesizing Differentially Private Datasets using Random Mixing
Systematic Matrix Multiplication Codes
Task-Based Quantization for Recovering Quadratic Functions Using Principal Inertia Components
The CEO Problem with rth Power of Difference Distortion
The Dirty Paper Wiretap Feedback Channel with or without Action on the State
The FM-linear Complexity of M-ary Sidel'nikov Sequences of Period p - 1 = f • Mλ
The Geometry of Community Detection via the MMSE Matrix
The Group Restricted Isometry Property for Subgaussian Block Diagonal Matrices
The Hat Guessing Number of Graphs
The Interactive Capacity of the Binary Symmetric Channel is at Least 1/40 the Shannon Capacity
The Interplay of Causality and Myopia in Adversarial Channel Models
The Optimal Power Control Policy for an Energy Harvesting System with Look-Ahead: Bernoulli Energy Arrivals
Throughput Scaling of Covert Communication over Wireless Adhoc Networks
Tightening Mutual Information Based Bounds on Generalization Error
Timely Coded Computing
Towards theoretically-founded learning-based denoising
Tradeoff Between Delay and High SNR Capacity in Quantized MIMO Systems
Tree Gradient Coding
Turing Computability of the Fourier Transform of Bandlimited Functions
Undetectable Radios: Covert Communication under Spectral Mask Constraints
Unique Information and Secret Key Decompositions
Universal D-Semifaithfull Coding for Countably Infinite Alphabets
Universally Decodable Matrices for Distributed Matrix-Vector Multiplication
Universal superposition codes: capacity regions of compound quantum broadcast channel with confidential messages
Updates with Multiple Service Classes
Using Erasure Feedback for Online Timely Updating with an Energy Harvesting Sensor
Variable-length compression and secrecy by design
Weakly-Private Information Retrieval
When a Heavy Tailed Service Minimizes Age of Information
Whittle Index Policy for Multichannel Scheduling in Queueing Systems

« prev Page 5/5