Hybrid Window Decoding for Joint Source Channel Anytime Coding System

Area of Science:

• Information Theory
• Coding Theory
• Digital Communications

Background:

• Joint Source Channel Anytime Coding (JSCAC) combines source and channel coding for improved performance.
• Existing JSCAC systems often use Joint Expanding Window Decoding (JEWD), which offers high error correction but also high complexity.
• There is a need for decoding algorithms that balance performance and complexity in JSCAC.

Purpose of the Study:

• To propose a novel Joint Hybrid Window Decoding (JHWD) algorithm for JSCAC systems.
• To reduce the decoding complexity of JSCAC while preserving error correction capabilities.
• To provide a low-complexity implementation option for JSCAC and related coding structures.

Main Methods:

• Development of a Joint Hybrid Window Decoding (JHWD) algorithm.
• Implementation of adaptive local expanding window decoding guided by syndrome-check-based detection.
• Utilizing a hybrid window structure to manage window expansion and sliding.
• Proposing an improved density evolution algorithm for asymptotic performance analysis.

Main Results:

• The JHWD algorithm effectively reduces decoding complexity associated with window expansion.
• Error propagation caused by window sliding is mitigated by the hybrid window approach.
• The proposed JHWD scheme demonstrates comparable error correction performance to state-of-the-art methods.
• Asymptotic performance analysis confirms the effectiveness of the JHWD algorithm.

Conclusions:

• The JHWD algorithm presents a viable low-complexity solution for JSCAC systems.
• This decoding scheme is suitable for various Joint Source Channel Coding (JSCC) systems employing spatially coupled codes.
• JHWD offers a practical trade-off between decoding complexity and error correction performance.