On The Decoding Error Weight of One or Two Deletion Channels

• Published in: arXiv.org
• Main Authors: Sabary, Omer, Bar-Lev, Daniella, Gershon, Yotam, Yucovich, Alexander, Yaakobi, Eitan
• Format: Paper
• Language: English
• Published: Ithaca Cornell University Library, arXiv.org 07.01.2022
• Subjects: Channels; Decoding; Deletion; Insertion; Maximum likelihood decoding; Storage systems; Symbols

This paper tackles two problems that are relevant to coding for insertions and deletions. These problems are motivated by several applications, among them is reconstructing strands in DNA-based storage systems. Under this paradigm, a word is transmitted over some fixed number of identical independent channels and the goal of the decoder is to output the transmitted word or some close approximation of it. The first part of this paper studies the deletion channel that deletes a symbol with some fixed probability \(p\), while focusing on two instances of this channel. Since operating the maximum likelihood (ML) decoder in this case is computationally unfeasible, we study a slightly degraded version of this decoder for two channels and its expected normalized distance. We identify the dominant error patterns and based on these observations, it is derived that the expected normalized distance of the degraded ML decoder is roughly \(\frac{3q-1}{q-1}p^2\), when the transmitted word is any \(q\)-ary sequence and \(p\) is the channel's deletion probability. We also study the cases when the transmitted word belongs to the Varshamov Tenengolts (VT) code or the shifted VT code. Additionally, the insertion channel is studied as well as the case of two insertion channels. These theoretical results are verified by corresponding simulations. The second part of the paper studies optimal decoding for a special case of the deletion channel, the \(k\)-deletion channel, which deletes exactly \(k\) symbols of the transmitted word uniformly at random. In this part, the goal is to understand how an optimal decoder operates in order to minimize the expected normalized distance. A full characterization of an efficient optimal decoder for this setup, referred to as the maximum likelihood* (ML*) decoder, is given for a channel that deletes one or two symbols.