Turbo Equalizer Segmentation for High-Throughput Optical Compensation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
High-speed optical fiber transmission systems face limitations in throughput due to nonlinear effects, polarization mode dispersion, and differential coding, which current Turbo equalization methods struggle to adapt to, especially at speeds greater than 100 Gbit/s.
Innovation Solution
The method involves dividing data blocks into overlapping segments for recursive processing and iterative decoding using an OP-BCJR unit and LDPC convolutional code decoding, reducing storage needs and improving throughput by processing data blocks that are 1/T of the code length, where T is the number of layers in the step-shaped check matrix.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a common sliding window BCJR with a serial structure is used in the Turbo equalizer, then the compensation accuracy for channel damage effects is improved, but the system throughput is limited due to the need to store the entire LDPC code word in the BCJR module
Solution Approach 1:
The patent divides the entire LDPC code word into multiple data segments that can be processed in parallel by multiple BCJR modules. Each BCJR module processes a specific segment independently, eliminating the need to store the entire code word in a single module. This segmentation approach maintains compensation accuracy while enabling parallel processing to improve system throughput.
Solution Approach 2:
The patent transitions from a serial processing structure to a parallel processing structure by introducing multiple BCJR modules operating simultaneously on different data segments. This dimensional change from single-threaded to multi-threaded processing allows the system to achieve higher throughput while maintaining the same compensation accuracy through parallel execution of the same algorithm.
2Reliability
If the BCJR module stores the entire LDPC code word to perform accurate recursive operations, then the compensation performance is improved, but the storage resources and device complexity increase significantly
Solution Approach 1:
The patent segments the large LDPC code word into smaller data segments, each of which can be stored and processed by individual BCJR modules. This reduces the storage burden on each module from the entire code word length to only a segment length, significantly decreasing total storage requirements while maintaining compensation performance through parallel processing of all segments.
3Reliability
If the Turbo equalizer uses a feedback structure with large-length LDPC code word and complex BCJR module, then the channel damage compensation is improved, but the system throughput is limited and cannot adapt to large-capacity high-speed transmission
Solution Approach 1:
The patent divides the processing task into multiple segments handled by parallel BCJR modules, each processing a portion of the LDPC code word simultaneously. This segmentation enables the system to maintain the feedback structure and complex BCJR processing needed for effective channel damage compensation while achieving high throughput through parallel execution.
Solution Approach 2:
The patent performs preliminary segmentation of the LDPC code word into multiple data segments before processing. This preliminary action allows the subsequent parallel processing to proceed efficiently, as each BCJR module receives pre-prepared segment data and can operate independently without waiting for other modules, thereby maximizing system throughput.
Data Source
AI summary
Embodiments of the present application relate to a method for implementing Turbo equalization compensation. The equalizer divides a first data block into n data segments, where D bits in two adjacent data segments in the n data segments overlap, performs recursive processing on each data segment in the n data segments, before the recursive processing, merges the n data segments to obtain a second data block; and performs iterative decoding on the second data block, to output a third data block, where data lengths of the first data block, the second data block, and the third data block are all 1/T of a code length of a LDPC convolutional code.