Soft-Aided Staircase Code Decoding for Miscorrection Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current hard-decision-based forward error correction (FEC) codes, such as staircase codes, face limitations in error-correcting capability and are prone to miscorrections due to bounded-distance decoding, which degrades performance in high data rate applications like optical transport networks.
Innovation Solution
A decoding algorithm that partially uses soft information from the channel by marking highly reliable and unreliable bits to improve miscorrection-detection and error-correcting capabilities, without significantly increasing complexity or latency, applicable to various hard-decision FEC codes like staircase and product codes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If bounded-distance decoding (BDD) is used for hard-decision FEC codes, then decoding complexity is reduced, but error-correcting capability is limited to t=[(d0−1)/2] and miscorrections occur
Solution Approach 1:
The patent applies preliminary action by performing miscorrection detection before final decoding decisions are made. The algorithm detects potential miscorrections in advance by checking syndrome conditions and bit reliability metrics, then corrects them before outputting the final decoded result, thereby preventing the limited error-correcting capability of standard BDD from becoming a bottleneck.
Solution Approach 2:
The patent implements feedback by using the syndrome information and bit reliability metrics from the initial BDD to guide subsequent correction steps. The decoded output is fed back through a verification process that checks for miscorrections using syndrome conditions, and unreliable bits are identified and re-decoded based on this feedback, effectively extending the error-correcting capability beyond the standard BDD limit.
2Reliability
If soft-decision FEC codes are used, then net coding gain (NCG) is improved, but power consumption and decoding latency increase
Solution Approach 1:
The patent applies partial action by selectively using soft information only where needed. Instead of implementing full soft-decision decoding, the algorithm uses reliability metrics from the hard decision process to identify and correct only the most unreliable bits and detect miscorrections, achieving significant NCG improvement while avoiding the full power consumption and latency overhead of complete soft-decision decoding.
3Reliability
If anchor-based decoding algorithm is used, then miscorrection prevention is improved, but decoding complexity increases due to tracking anchor codewords
Solution Approach 1:
The patent extracts the essential function of anchor-based decoding (miscorrection prevention) without implementing the full anchor tracking mechanism. The algorithm extracts and uses only the syndrome-based miscorrection detection capability, removing the complex anchor codeword tracking requirement while maintaining the core benefit of preventing miscorrections.
4Reliability
If modified iterative BDD with channel reliabilities is used, then decoding performance is improved, but additional memory and processing are required
Solution Approach 1:
The patent applies partial action by using channel reliability information selectively rather than storing and processing all soft information. The algorithm computes reliability metrics only for bits that need correction or verification, avoiding the large memory requirements of storing complete soft information while still achieving improved decoding performance through targeted reliability-based corrections.
Data Source
AI summary
A hard-decision (HD) forward error correcting (FEC) coded signal is decoded by a decoder to produce decoded bits using marked reliable bits of the HD-FEC coded signal and marked unreliable bits of the HD-FEC coded signal. The marked reliable and unreliable bits are computed by calculation and marking blocks based on an absolute value of log-likelihood ratios of the HD-FEC coded signal. The HD-FEC coded signal may be, for example, a staircase code coded signal or a product code coded signal.