Quasi-Reduced State Equalizer for Low-Latency ISI Mitigation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
High symbol rates in next-generation communications networks face significant challenges due to inter-symbol interference (ISI), which introduces errors in signal decoding, particularly in wireless and optical channels, and existing techniques for reducing ISI are complex and latency-prone, especially at data rates exceeding 100 Gbps.
Innovation Solution
The implementation of a low-complexity quasi-reduced state soft-output equalizer that computes accumulated path metrics for candidate states of leading symbols and selectively propagates these metrics over a trellis to candidate states of trailing symbols, reducing processing load and latency while generating high-quality soft-outputs for bit-level decoding without requiring backward recursion.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional MAP equalization is used to reduce ISI, then decoding accuracy is improved, but processing latency increases
Solution Approach 1:
The patent segments the trellis structure into distinct stages (first trellis stage for leading symbol, second trellis stage for trailing symbol) and processes them in a pipelined manner. This allows independent computation of path metrics for each stage, enabling parallel processing that reduces latency while maintaining decoding accuracy through systematic metric computation and propagation.
Solution Approach 2:
The patent computes path metrics for the leading symbol's candidate states before processing the trailing symbol. By performing preliminary metric computation and propagation in advance through the trellis structure, the system prepares data for subsequent decoding stages, reducing overall processing latency while preserving decoding accuracy.
2Productivity
If reduced-state trellis equalization is applied to lower complexity, then processing speed is improved, but soft-output quality deteriorates
Solution Approach 1:
The patent applies different processing qualities to different parts of the trellis structure. The leading symbol's candidate states undergo full metric computation with high precision, while the trailing symbol utilizes propagated metrics efficiently. This local differentiation maintains soft-output quality where critical while optimizing processing speed in less critical paths.
Solution Approach 2:
The patent changes the parameter of state representation by computing and propagating accumulated path metrics (APMs) through the trellis stages. By transforming the representation from raw symbol data to accumulated metrics, the system maintains information quality while enabling more efficient processing in subsequent decoding stages.
3Measurement precision
If full trellis computation is performed for all symbols, then decoding accuracy is maintained, but device complexity increases
Solution Approach 1:
The patent extracts and propagates only the essential accumulated path metrics from the leading symbol's candidate states to the trailing symbol's candidate states, rather than performing complete recomputation. This extraction of critical information reduces computational complexity while maintaining decoding accuracy through selective metric propagation.
Solution Approach 2:
The patent performs preliminary computation of accumulated path metrics for leading symbol candidate states before the trailing symbol processing. By preparing these metrics in advance and propagating them through the trellis, the system avoids redundant computations during trailing symbol processing, reducing overall device complexity while maintaining accuracy.
Data Source
Figure 1
Figure 2
Figure 3~6
AI summary
Quasi-reduced state trellis equalization techniques achieve low-latency inter-symbol interference (ISI) equalization by selecting a subset of accumulated path metrics (APMs) for a leading symbol to propagate over a trellis to candidate states of a trailing symbol. This simplifies the computation of APMs for candidate states of the trailing symbol. Thereafter, APMs for candidate states of the trailing symbol are computed based on the subset of APMs for the leading symbol that were propagated over the trellis. Propagating fewer than all APMs for the leading symbol to the trailing symbol reduces the complexity of APM computation at the trailing symbol.