Developing a novel DNA-based steganography algorithm using random table generation with segmentation

Genome steganography has emerged as a promising field for transmitting large amounts of data over an untrusted channel in the last two decades. DNA has many advantages over other multimedia cover mediums. Substitution is the most common approach to developing a DNA-based steganography algorithm. Components of the secret message are converted to DNA letters, which replace nucleotides in the cover sequence. The conversion is conducted through predetermined tables like binary coding rules and lookup/dictionary tables. These tables are static, limited to specific alphanumeric characters, and may compromise the hidden message if discovered by intruders. Most previously proposed algorithms adopt a simple sequential and ordered hiding pattern. This leads to poor utilization of the available hiding spots and creates a region of interest for attackers to apply steganalysis. In this paper, a novel DNA-based algorithm is proposed. Using two DNA sequences, primary and secondary, is the basis for this algorithm. Both sequences are segmented, where the primary sequence segments are for hiding the data, and the secondary sequence segments are for conveying the required information to find and extract the hidden data. Three enhanced tables are proposed: a modified ASCII table, a 4-bit binary coding rule table, and a two-tier lookup encoding table to convert the message to DNA form. The segmentation ensures that the hiding spots are randomly scattered across the primary cover DNA sequence, addressing the region of interest issue. The data is hidden in the cover using the least significant base substitutions. The result is an all-rounded algorithm that fulfills the desired performance measurements such as zero payloads, blindness, preserving functionality, high hiding capacity, low modification rate, and low cracking probability.