English

Multilinear Algebra for Minimum Storage Regenerating Codes

Information Theory 2022-01-07 v1 Commutative Algebra math.IT

Abstract

An (n,k,d,α)(n, k, d, \alpha)-MSR (minimum storage regeneration) code is a set of nn nodes used to store a file. For a file of total size kαk\alpha, each node stores α\alpha symbols, any kk nodes recover the file, and any dd nodes can repair any other node via each sending out α/(dk+1)\alpha/(d-k+1) symbols. In this work, we explore various ways to re-express the infamous product-matrix construction using skew-symmetric matrices, polynomials, symmetric algebras, and exterior algebras. We then introduce a multilinear algebra foundation to produce (n,k,(k1)tt1,(k1t1))\bigl(n, k, \frac{(k-1)t}{t-1}, \binom{k-1}{t-1}\bigr)-MSR codes for general t2t\geq2. At the t=2t=2 end, they include the product-matrix construction as a special case. At the t=kt=k end, we recover determinant codes of mode m=km=k; further restriction to n=k+1n=k+1 makes it identical to the layered code at the MSR point. Our codes' sub-packetization level---α\alpha---is independent of nn and small. It is less than L2.8(dk+1)L^{2.8(d-k+1)}, where LL is Alrabiah--Guruswami's lower bound on α\alpha. Furthermore, it is less than other MSR codes' α\alpha for a subset of practical parameters. We offer hints on how our code repairs multiple failures at once.

Keywords

Cite

@article{arxiv.2006.16998,
  title  = {Multilinear Algebra for Minimum Storage Regenerating Codes},
  author = {Iwan Duursma and Hsin-Po Wang},
  journal= {arXiv preprint arXiv:2006.16998},
  year   = {2022}
}

Comments

36 pages, 9 figures, 3 tables

R2 v1 2026-06-23T16:44:46.391Z