High-performance and low-complexity iterative large-number logic decoding method suitable for multi-system LDPC code

Abstract

The invention discloses a high-performance and low-complexity iterative large-number logic decoding method suitable for a multi-system LDPC code. The method mainly comprises three innovative points: (1) two improved methods for reducing implementation complexity are provided, and the operation of searching two minimum values is simplified into the operation of searching one minimum value when a priori reliability measurement value is calculated, so that the required comparison operation is reduced by half, and the performance loss is hardly caused; secondly, when the total reliability metric value is calculated, a large amount of integer multiplication is converted into a very small number of divisions, so that the bit width of the reliability metric value is reduced while multiplication operation is greatly reduced; (2) an unsaturated amplitude limiting method is provided, and the competitiveness of other domain elements is improved, so that the decoding performance is greatly improved, and the bit width of a reliability metric value is further reduced; and (3) aiming at the problem of huge storage consumption of multi-system LDPC decoding, a partially fixed storage method is provided, so that the corresponding storage consumption is greatly reduced. The invention provides an effective decoding method, the implementation complexity is reduced while the decoding performance isimproved, and the decoding method has certain innovativeness.

Description

technical field [0001] The invention relates to a decoding method in the technical field of communication coding, in particular to a high-performance, low-complexity iterative large-number logic decoding method suitable for multi-ary LDPC codes. Background technique [0002] In communication systems, there are often various noises in data from transmission to reception. Therefore, in the process of data transmission, channel coding needs to be introduced. [0003] Non-binary LDPC (NB-LDPC) codes have better decoding performance than binary LDPC codes in medium and short code lengths and high-order modulations, but high decoding complexity has always been the bottleneck of the code's application. In order to reduce its decoding complexity, many optimization algorithms based on Belief Propagation (BP) method have been proposed. Among them, the most commonly used ones are Extended Minimum Sum Method (EMSA), Minimum Maximum Method (MMA) and so on. But they are still relativel...

AI Technical Summary

Problems solved by technology

But relatively speaking, there will be some loss in performance. At the same time, compared with binary LDPC codes, the decoding complexity is still relatively large, so there is still a lot of research space for this method to improve performance and reduce implementation complexity.

Images