Variable-length constrained sequence coding

点击翻译按钮获取中文摘要

This research focuses on the continued development and application of variable-length constrained sequence codes in digital transmission and storage systems.******Encoding and decoding of data is an indispensable part of all digital communication systems, including both transmission and storage systems. These coding techniques fall into the general classes of:******Source codes: used to represent the source information in as compact a form as possible;******Constrained sequence (CS) codes: used to generate encoded symbol streams that satisfy the constraints of the channel and attempt to ensure proper operation of practical demodulation circuitry, and******Error control codes: used to detect and correct symbols that have been incorrectly detected by the demodulator.******When compressing data with a source code, variable-length (VL) coding techniques are almost always used owing to their ability to efficiently represent the data with as few encoded symbols as possible. In contrast, practitioners of both constrained sequence codes and error control codes continue to focus on the development and implementation of fixed-rate codes.******However, many new communication systems are inherently based on the use of VL sequences which enables the use of simple and efficient VL CS codes. Packet-switched transmission networks are prime candidates for VL CS codes, as are storage systems that use VL compression techniques.******My research will focus on (1) the continued development of and (2) the application of variable-length constrained sequence codes for digital transmission and storage systems. The starting point of this work will be the state-independent VL CS coding technique recently developed in my group whose implementation is simpler, and which offers higher efficiency, than any other class of CS codes published to date.******Extensions of a fundamental nature to this exciting work will include:******(1a) extending the current state-independent coding techniques to support state-dependent encoding******(1b) developing simplified search technique for optimal codes******(1c) determining appropriate techniques to mitigate error propagation during decoding******(1d) optimizing word mappings to limit error propagation during VL CS decoding******(1e) integrating VL CS codes with error control codes, and******(1f) extending the development of binary VL CS codes to systems with multidimensional signaling alphabets.******We will also apply these principles to optimize the design of VL CS codes for the important applications of:******(2a) recording codes for flash memory, and******(2b) spectral shaping codes to avoid narrowband interference in single-carrier digital video transmission systems.******This research will therefore continue the development of a new approach to construct efficient constrained sequence codes and will demonstrate their design and application in new and evolving digital communication systems.**