Optical Receiver Soft-Decision Circuit With Reduced Bit Width
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional soft-decision-data generating circuits for optical communication systems face challenges in reducing circuit size and processing load due to increased bit width, and struggle to generate soft-decision data with reduced bit width from digital signals, especially when noise distribution is asymmetrical.
Innovation Solution
An optical receiving apparatus that includes an A/D converting circuit, a received-signal demodulating circuit, and a soft-decision-data generating circuit which generates n-bit soft-decision data using the most significant bit as hard-decision data and reliability information from comparisons of bits on the MSB and LSB sides, allowing for reduced bit width and optimized error correction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If soft-decision data bit width is increased to improve error correction capability, then error correction performance is improved, but circuit size and processing load increase
Solution Approach 1:
The patent extracts only the necessary components from the full m-bit received signal to generate n-bit soft-decision data (where n < m). By selecting specific bits and generating reliability information only where needed, the circuit size is reduced while maintaining error correction capability. This is achieved by taking out the essential information elements rather than processing the entire m-bit signal.
Solution Approach 2:
The patent segments the m-bit received signal into multiple parts: MSB-side bits for hard decision, LSB-side bits for reliability information, and generates soft-decision data through segmented processing. This segmentation allows different parts of the signal to be processed differently, reducing overall circuit complexity while preserving error correction performance.
2Measurement precision
If multiple soft-decision threshold comparisons are implemented to generate accurate soft-decision data, then soft-decision accuracy is improved, but processing load and circuit complexity increase
Solution Approach 1:
Instead of performing comparisons with all 2^n-1 soft-decision thresholds, the patent extracts reliability information from only k bits (where k < m) of the received signal. This extraction approach maintains soft-decision accuracy by capturing essential reliability information while dramatically reducing the number of comparison operations required.
Solution Approach 2:
The patent performs preliminary hard decision on the MSB-side bits before generating soft-decision data. This preliminary action establishes a baseline decision result that reduces the complexity of subsequent soft-decision processing, as the hard decision outcome can be used to simplify the comparison operations needed for soft-decision accuracy.
3Device complexity
If bit width reduction is applied to decrease circuit size, then circuit size is reduced, but soft-decision data generation becomes more difficult
Solution Approach 1:
The patent divides the m-bit received signal into MSB-side bits and LSB-side bits, with the LSB-side bits further segmented into k bits for reliability information and (m-k) bits discarded. This segmentation creates a systematic method for bit width reduction that is easier to implement than arbitrary reduction, as it follows a clear structural division of the signal.
Solution Approach 2:
The patent changes the parameter of soft-decision data bit width from the conventional m bits to n bits (where n < m), specifically setting n based on the number of reliability information bits k. This parameter change is achieved through a standardized process of selecting k bits and generating reliability information, making the reduction systematic rather than arbitrary.
Data Source
AI summary
An optical receiving apparatus includes: an A/D converting circuit; a received-signal demodulating circuit that demodulates a received digital signal from the A/D converting circuit into an m-bit received signal; a soft-decision-data generating circuit that generates n-bit (n≦m) soft-decision data based on the m-bit received signal; and an error correcting circuit that performs error correction based on the n-bit soft-decision data and outputs an error-corrected received signal. The soft-decision-data generating circuit generates soft-decision data of n bits (n=p+1) that corresponds to a determination result according to 2n−1 soft-decision thresholds, by using an MSB of the m-bit received signal as hard-decision data, and by using, as reliability information, a result of comparison between a plurality of bits (k bits, where k≦m) on an MSB side of the m-bit received signal and a fixed threshold, or p bits (p≦m−k) selected from (m−k) bits on an LSB side of the m-bit received signal.