Poly-Stranded FEC Layout for Parallel Optical Error Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional forward error correction techniques in optical communication systems, such as product codes and braided codes, have inferior error rates and are not amenable to efficient parallel implementation, which is crucial for high-speed communication systems.
Innovation Solution
The development of poly-stranded error correcting codes, which organize data bits into a specific arrangement allowing for parallel encoding and decoding, using multiple strands of blocks with different error correcting codes applied to each strand to generate parity bits, facilitating high-speed and reliable data transmission.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional forward error correction techniques (product codes, braided codes) are used, then implementation is simpler, but error rates are inferior and parallel processing efficiency is reduced
Solution Approach 1:
The patent segments the data bits into multiple strands, where each strand is independently encoded using a specific error correcting code. This segmentation allows for parallel processing of multiple strands simultaneously, improving both error correction capability and processing efficiency. The data is divided into first strands and second strands, with each strand containing multiple blocks that are encoded independently.
Solution Approach 2:
The patent introduces a multi-dimensional structure by organizing data into strands with blocks arranged in specific patterns (e.g., diagonal arrangements). This dimensional organization enables more efficient parallel encoding and decoding operations while maintaining strong error correction properties. The multi-strand structure adds a new dimension to the traditional single-code approach.
2Productivity
If high-speed transmission is implemented, then data rate increases, but parallel processing capability is required which conventional codes lack
Solution Approach 1:
By dividing the data into multiple independent strands that can be processed in parallel, the system achieves high-speed transmission. Each strand can be encoded and decoded independently using the same or different error correcting codes, enabling efficient parallel processing architecture that meets high-speed requirements.
Solution Approach 2:
The patent employs multiple error correcting codes (e.g., BCH codes, Reed-Solomon codes, LDPC codes) that can be universally applied to different strands. This multi-functionality allows the system to handle various data rates and error conditions using a unified multi-strand framework, simplifying the implementation of parallel processing.
3Reliability
If redundancy is added to reduce transmission errors, then error correction capability improves, but transmission efficiency decreases
Solution Approach 1:
The patent applies error correcting codes selectively to different strands rather than uniformly to all data. This partial action approach allows the system to achieve the required error correction capability for each strand while optimizing the overall redundancy level. Not all strands require the same degree of protection, enabling efficient resource allocation.
Solution Approach 2:
The system can change the error correcting code parameters (e.g., code rate, block length) for different strands based on their specific requirements. This allows optimization of the balance between error correction capability and transmission efficiency for each strand, rather than using a fixed redundancy level for all data.
Data Source
AI summary
Techniques for performing forward error correction of data to be transmitted over an optical communications channel. The techniques include: receiving data bits; organizing the data bits into an arrangement having a plurality of blocks organized into rows and columns and into a plurality of strands including a first strand of blocks that includes a back portion comprising a first row of the plurality of blocks, and a front portion comprising blocks from at least two different columns in at least two different rows other than the first row of blocks; and encoding at least some of the data bits in the arrangement using a first error correcting code at least in part by generating first parity bits by applying the first error correcting code to first data bits in the front portion of the first strands and second data bits in the back portion of the first strand.