Year: 2025
Type: dissertation
Abstract: DNA-based data storage systems have garnered significant attention due to their remarkable properties, including high information density, durability, and replicability. However, conventional DNA storage faces critical challenge which is the high synthesis costs. To address this issues, this thesis investigates the coding theory and algorithmic frameworks for composite DNA-based data storage, where information is represented by probability vectors rather than standard nucleotide sequences. By integrating contributions from three core research areas—error correction, decoding, and secret sharing—this work provides a comprehensive approach to enhancing the reliability, efficiency, and security of composite DNA-based data storage systems.First, this thesis introduces a coding scheme for limited-magnitude probability errors (LMPE) that arise during composite DNA synthesis and sequencing. We develop theoretical bounds and coding constructions for LMPE, focusing on minimizing redundancy and computational complexity. The proposed systematic codes demonstrate asymptotic optimality and practicalapplicability in composite DNA-based data storage systems. Second, we presents a novel ramp secret sharing scheme (ARSSS) designed for composite DNA-based data storage systems. Unlike traditional secret sharing approaches based on finite fields, this scheme operates directly on probability vectors. The proposed method reduces sequencing requirements, enhances data security through asymptotic information-theoretic.
Author: Zhang, Wenkai
Institution: University of California, Irvine
Language: English
Topic: DNA and Biological Computing
Subfield: Theoretical Computer Science
Field: Mathematics
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7046147409
