GFP-T Superblock Error Correction With Two-Page Buffering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for processing GFP-T superblocks in high-speed transport networks, such as 10 GbE, 40 GbE, and 100 GbE, are inadequate in reducing latency and error correction efficiency due to reliance on multiple syndrome tables and buffer stages, which increase wait times and are not sufficient for contemporary transmission rates.
Innovation Solution
A two-page buffer system is employed where the payload data of a GFP-T superblock is buffered in one page and the flag byte in a separate buffer, with a single CRC operation performed, and the results checked against a single syndrome table for error correction, allowing continuous data flow and minimizing latency by eliminating the need for multiple syndrome tables.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If multiple syndrome tables and buffer stages are used for error correction in GFP-T superblocks, then error correction capability is improved, but latency increases and processing efficiency deteriorates
Solution Approach 1:
The patent segments the error correction process by separating the CRC calculation from the syndrome table lookup. The 64-byte payload is divided into 8-byte segments for parallel processing, with each segment having its own syndrome table entry. This allows the system to maintain comprehensive error correction capability while reducing the critical path latency by enabling concurrent operations.
Solution Approach 2:
The patent pre-calculates and stores syndrome table values in a lookup table before data arrives. By preparing the syndrome tables in advance and organizing them for rapid access, the system eliminates the need for complex real-time calculations during data processing, thereby reducing latency while maintaining robust error correction.
2Measurement precision
If multiple syndrome tables and buffer stages are used for error correction, then error detection accuracy is improved, but processing speed deteriorates
Solution Approach 1:
The patent replaces complex sequential mechanical processing with a parallel lookup-based system. Instead of sequentially checking multiple syndrome tables with multiple buffer stages, the system uses a single syndrome table lookup for each 8-byte segment with parallel CRC calculation, achieving the same error detection accuracy at higher speeds.
Solution Approach 2:
The patent applies partial action by processing data in 8-byte segments rather than requiring complete 64-byte block processing before error correction. This allows early detection and correction of errors in each segment independently, maintaining high detection accuracy while enabling continuous data flow and improving overall processing throughput.
3Adaptability or versatility
If frame-by-frame mapping is used in GFP-F, then protocol compatibility is improved, but latency increases for high-volume data transport
Solution Approach 1:
The patent segments the GFP-T frame into multiple 8-byte segments that can be processed independently and in parallel. This segmentation allows the system to maintain the structured approach needed for protocol compatibility while enabling concurrent processing of multiple segments, thereby reducing the wait time associated with sequential frame-by-frame processing.
Solution Approach 2:
The patent enables continuous data flow by allowing CRC calculation and syndrome table lookup to occur in parallel and continuously as data arrives, rather than waiting for complete frames. This continuous processing maintains protocol compatibility through proper CRC verification while eliminating idle wait times between frame processing cycles.
Data Source
AI summary
The present invention discloses a method and apparatus for processing and error correction of a GFP-T superblock, where the 64 bytes of payload data of a first superblock are buffered in the first page of a two-page buffer. The flag byte is buffered in a separate buffer, and a CRC operation is performed in a separate logic element. The result of the CRC operation is checked against a single syndrome table which may indicate single- or multi-bit errors. As the payload data of the first superblock is processed and read out of the first page of the two-page buffer, the payload data of a second superblock is written into the second page of the two-page buffer to be processed and corrected.