Random Construction of Partial MDS Codes

Alessandro Neri,Anna-Lena Horlemann-Trautmann

DOI: https://doi.org/10.48550/arXiv.1801.05848

2018-01-18

Abstract:This work deals with partial MDS (PMDS) codes, a special class of locally repairable codes, used for distributed storage system. We first show that a known construction of these codes, using Gabidulin codes, can be extended to use any maximum rank distance code. Then we define a standard form for the generator matrices of PMDS codes and use this form to give an algebraic description of PMDS generator matrices. This implies that over a sufficiently large finite field a randomly chosen generator matrix in PMDS standard form generates a PMDS code with high probability. This also provides sufficient conditions on the field size for the existence of PMDS codes.

Information Theory

What problem does this paper attempt to address?

The problem that this paper attempts to solve is how to construct Partial Maximum Distance Separable (PMDS) codes in distributed storage systems to achieve efficient data recovery. Specifically, the paper focuses on how to construct PMDS codes by random methods and determine the sufficient conditions for the existence of these codes over a sufficiently large finite field.

### Background of the Paper

In a distributed storage system, a file \(x\in F_q^k\) is encoded and stored as a codeword \(c\in F_q^n\), distributed across multiple storage nodes. Each node is assumed to store one coordinate of the codeword. When some nodes fail, it is hoped that the lost information can be recovered with minimal effort. The locality of the code plays an important role in this context, which is represented by the number of nodes that need to be contacted in order to repair a failed node. PMDS codes are a special type of locally repairable codes that can achieve maximum recoverability in this setting, that is, any theoretically correctable erasure pattern can be corrected by such codes.

### Research Objectives

1. **Expand Known Constructions**: The paper first shows that the method of constructing PMDS codes using Gabidulin codes can be extended to using any Maximum Rank Distance (MRD) codes.
2. **Define Standard Form**: The paper defines the standard form of the PMDS code generator matrix and uses this form to give an algebraic description of the PMDS generator matrix.
3. **Random Construction**: The paper studies the probability that a generator matrix generates a PMDS code when randomly selecting the non - specified coordinate values of the generator matrix under the PMDS standard form. This shows that, over a sufficiently large finite field, a randomly selected generator matrix generates a PMDS code with high probability.
4. **Existence Conditions**: The paper derives a lower bound on the necessary field size for the existence of PMDS codes and improves this lower bound by a step - by - step construction method.

### Main Contributions

- **Extended Construction Method**: It is proved that PMDS codes can be constructed using any MRD code, not just Gabidulin codes.
- **Standard Form**: The standard form of the PMDS code generator matrix is defined and its algebraic description is given.
- **Probability Analysis**: The probability of generating a PMDS code when randomly selecting a generator matrix is derived, and a lower bound on the field size is given.
- **Step - by - Step Construction**: A step - by - step construction method is proposed to further improve the lower bound on the field size.

### Conclusion

Through random methods and step - by - step construction methods, the paper provides an effective construction scheme for PMDS codes in distributed storage systems and determines the sufficient conditions for the existence of these codes over a sufficiently large finite field. These results are of great significance for improving the reliability and efficiency of distributed storage systems.