CAT and DOG: Improved Codes for Private Distributed Matrix Multiplication

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Hofmeister, Christoph, Bitar, Rawad, Wachter-Zeh, Antonia
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909711857614848
author Hofmeister, Christoph
Bitar, Rawad
Wachter-Zeh, Antonia
author_facet Hofmeister, Christoph
Bitar, Rawad
Wachter-Zeh, Antonia
contents We present novel constructions of polynomial codes for private distributed matrix multiplication (PDMM/SDMM) using outer product partitioning (OPP). We extend the degree table framework from the literature to cyclic-addition degree tables (CATs). By using roots of unity as evaluation points, we enable modulo-addition in the table. Based on CATs, we present an explicit construction, called CATx, that requires fewer workers than existing schemes in the low-privacy regime. Additionally, we present new families of schemes based on conventional degree tables, called GASPrs and DOGrs, that outperform the state-of-the-art for a wide range of parameters.
format Preprint
id arxiv_https___arxiv_org_abs_2501_12371
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle CAT and DOG: Improved Codes for Private Distributed Matrix Multiplication
Hofmeister, Christoph
Bitar, Rawad
Wachter-Zeh, Antonia
Information Theory
We present novel constructions of polynomial codes for private distributed matrix multiplication (PDMM/SDMM) using outer product partitioning (OPP). We extend the degree table framework from the literature to cyclic-addition degree tables (CATs). By using roots of unity as evaluation points, we enable modulo-addition in the table. Based on CATs, we present an explicit construction, called CATx, that requires fewer workers than existing schemes in the low-privacy regime. Additionally, we present new families of schemes based on conventional degree tables, called GASPrs and DOGrs, that outperform the state-of-the-art for a wide range of parameters.
title CAT and DOG: Improved Codes for Private Distributed Matrix Multiplication
topic Information Theory
url https://arxiv.org/abs/2501.12371