A Novel Soft-In Soft-Out Decoding Algorithm for VT Codes on Multiple Received DNA Strands

In recent years, DNA-based data storage has received extensive attention as a promising technology due to its high density, long-term durability, and low power consumption. One of the most challenging issues in its field is coping with insertion, deletion, and substitution (IDS) errors introduced during DNA synthesis and sequencing. Many traditional codes have been introduced as error correction methods, among which VT codes have attracted widespread interest since it is asymptotically optimal. In this work, we provide a novel soft-input soft-output (SISO) decoding algorithm for VT codes and generalize it to decode multiple received sequences. Monte Carlo simulations show that the bit error rate of the SISO decoder is one order of magnitude lower as compared with that of the conventional hard-decision decoder. Additionally, the generalized decoding algorithm over multiple received sequences achieves significant performance gains compared to a single-sequence transmission case. We further provide two reduced-complexity strategies for our SISO decoding algorithm that greatly reduce the decoding complexity by truncating traces of small probability.