Fig. 1

The proposed architecture of DNA data storage, stored data and results of scheme performance. (A) A long binary stream is segmented into source packets with the same length before performing packet level repeat accumulate encoding where parity packets are generated. All packets are attached with unique addresses to locate their orders and Cyclic Redundancy Check (CRC) to detect errors occurring interiorly. After that, the digital information is converted to DNA sequences according to the hybrid mapping strategy. The mapped DNA sequences are synthesized for storage. Then, DNA sequencing is performed to read the oligos. At the last stage, we recover the original data by performing de-mapping and decoding with the sequencing result. (B) Six files of total size of 379.1 KB are encoded. (C) By taking the comparison table in [13] as a reference, we compare existing oligo-based DNA data storage schemes where improvements are shown in several aspects of performance. (D) The proposed scheme achieves 1.67 bits/nt net information density, which is higher than the existing oligo-based DNA data storage schemes. (E) The net information density of the proposed coding scheme with different magnitudes of data size is computational analysed, in which the density slightly decreases with the exponential increase of the data size (blue triangle). The estimated density results are higher than the existing work [11, 13, 14] with the equivalent data scales (black circle)