MLSE Trellis Initialization Using LUT State Metrics
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional trellis-based Maximal Likelihood Sequence Estimation (MLSE) decoders experience significant startup time due to the need to explore the trellis using run-up (RU) symbols, which increases decoding latency.
Innovation Solution
The proposed solution involves initializing a trellis of an MLSE engine using predetermined state information, which is stored in a Look-Up-Table (LUT) and based on simulations accounting for channel characteristics and DFE tap configurations. This partial initialization is followed by further initialization using a reduced set of RU symbols.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional trellis-based MLSE decoder uses run-up symbols to explore the trellis for initialization, then accurate decoding is achieved, but startup time and decoding latency increase significantly
Solution Approach 1:
The patent pre-computes and stores optimal initial state metrics for the trellis decoder in a lookup table before actual decoding operations. This preliminary action eliminates the need for runtime exploration using run-up symbols, directly reducing startup time while maintaining decoding accuracy through pre-optimized state initialization.
Solution Approach 2:
The patent creates a simplified copy of the trellis state information stored in a lookup table that contains pre-calculated initial state metrics. This copy allows the decoder to initialize quickly without performing the complete trellis exploration that would otherwise be required, thus reducing latency while preserving the essential state information needed for accurate decoding.
2Reliability
If traditional MLSE decoder performs complete trellis exploration using RU symbols, then all possible states are evaluated, but device complexity and processing overhead increase
Solution Approach 1:
The patent pre-evaluates all possible trellis states offline and stores the results in a lookup table. During actual operation, the decoder simply retrieves pre-computed initial state metrics rather than performing complete trellis exploration, thereby maintaining state evaluation completeness while dramatically reducing processing complexity and computational overhead.
Solution Approach 2:
The lookup table serves as a self-service mechanism that provides pre-computed state information without requiring the decoder to perform complex real-time calculations. The system initializes itself using stored information, eliminating the need for elaborate runtime processing while ensuring all states are properly evaluated.
3Loss of time
If fewer RU symbols are used for initialization, then startup time is reduced, but trellis initialization accuracy may be compromised
Solution Approach 1:
The patent performs the computationally intensive initialization work in advance by pre-computing optimal initial state metrics and storing them in a lookup table. This preliminary action allows the decoder to initialize quickly during operation without sacrificing accuracy, as the pre-computed values are derived from complete trellis analysis performed offline.
Solution Approach 2:
The patent creates a copy of the complete trellis state information in a compact lookup table format that can be quickly accessed. This copy contains all necessary initialization data in a condensed form, enabling fast initialization with full accuracy without requiring the original extensive run-up symbol sequences.
Data Source
AI summary
A method may include at least partially initializing a trellis of an SE engine at least partially based on predetermined state information about a communication channel associated with an incoming data stream; and processing, via the MLSE engine, the incoming data stream to further initialize the trellis and decode the incoming data stream.