DOE PAGES title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: Securely Aggregated Coded Matrix Inversion

Journal Article · · IEEE Journal on Selected Areas in Information Theory

Coded computing is a method for mitigating straggling workers in a centralized computing network, by using erasure-coding techniques. Federated learning is a decentralized model for training data distributed across client devices. In this work we propose approximating the inverse of an aggregated data matrix, where the data is generated by clients; similar to the federated learning paradigm, while also being resilient to stragglers. To do so, we propose a coded computing method based on gradient coding. We modify this method so that the coordinator does not access the local data at any point; while the clients access the aggregated matrix in order to complete their tasks. Here, the network we consider is not centrally administrated, and the communications which take place are secure against potential eavesdroppers.

Research Organization:
Georgia Institute of Technology, Atlanta, GA (United States)
Sponsoring Organization:
USDOE
Grant/Contract Number:
NA0003921
OSTI ID:
2283180
Journal Information:
IEEE Journal on Selected Areas in Information Theory, Journal Name: IEEE Journal on Selected Areas in Information Theory Vol. 4; ISSN 2641-8770
Publisher:
IEEECopyright Statement
Country of Publication:
United States
Language:
English

References (36)

Convex Optimization book January 2004
Sequential Lagrange multiplier condition forϵ-optimal solution in convex programming journal October 2008
Lagrange Coded Computing with Sparsity Constraints conference September 2019
Secure Linear MDS Coded Matrix Inversion conference September 2022
Approximate Gaussian Elimination for Laplacians - Fast, Sparse, and Simple conference October 2016
Coded Federated Learning conference December 2019
Balanced Sparsest generator matrices for MDS codes conference July 2013
Balanced Reed-Solomon codes conference July 2016
Encoded distributed optimization conference June 2017
Improving Distributed Gradient Descent Using Reed-Solomon Codes conference June 2018
Coded Federated Computing in Wireless Networks with Straggling Devices and Imperfect CSI conference July 2019
Product Lagrange Coded Computing conference June 2020
Numerically Stable Binary Gradient Coding conference June 2020
Orthonormal Sketches for Secure Coded Regression conference June 2022
Generalized Lagrange Coded Computing: A Flexible Computation-Communication Tradeoff conference June 2022
Computational Code-Based Privacy in Coded Federated Learning conference June 2022
Balanced Reed-Solomon codes for all parameters conference September 2016
Coded Computing for Low-Latency Federated Learning Over Wireless Edge Networks journal January 2021
Analog Lagrange Coded Computing journal March 2021
Optimal Communication-Computation Trade-Off in Heterogeneous Gradient Coding journal September 2021
List-Decodable Coded Computing: Breaking the Adversarial Toleration Barrier journal September 2021
Successive Approximation Coding for Distributed Matrix Multiplication journal June 2022
Maximum-Likelihood Decoding of Reed–Solomon Codes is NP-Hard journal July 2005
Explicit Codes Achieving List Decoding Capacity: Error-Correction With Optimal Redundancy journal January 2008
Speeding Up Distributed Machine Learning Using Codes journal March 2018
On the Optimal Recovery Threshold of Coded Matrix Multiplication journal January 2020
Straggler Mitigation in Distributed Matrix Multiplication: Fundamental Limits and Optimal Coding journal March 2020
Gradient Coding From Cyclic MDS Codes and Expander Graphs journal December 2020
Numerically Stable Coded Matrix Computations via Circulant and Rotation Matrix Embeddings journal April 2022
Accuracy and Stability of Numerical Algorithms book January 2002
Numerical Linear Algebra book January 1997
How to share a secret journal November 1979
Coded Computing: Mitigating Fundamental Bottlenecks in Large-Scale Distributed Computing and Machine Learning journal January 2020
Convex Optimization: Algorithms and Complexity journal January 2015
Solution of Vandermonde Systems of Equations journal October 1970
A Simple Proof of the Gale-Ryser Theorem journal April 1996