Multidimensional Slicing for Low-Complexity Sequence Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional reduced complexity sequence estimation algorithms, such as the M-algorithm, face significant complexity and performance degradation, especially at low Signal-to-Noise Ratios (SNR), due to high complexity and Correct Path Loss (CPL) events, which are exacerbated by large symbol constellations and channel effects.
Innovation Solution
The method employs multidimensional slicing to generate branch vector hypotheses, reducing complexity by using a partial lattice that incorporates tap coefficients with higher magnitudes, allowing for more reliable branch metric calculations and reducing CPL events, thereby improving performance and reducing computational complexity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional M-algorithm is used for sequence estimation, then near ML performance can be achieved, but computational complexity becomes prohibitively high for large constellation sizes
Solution Approach 1:
The patent segments the full constellation set into multiple subsets or slices. Instead of evaluating all M constellation points for each survivor path, the algorithm divides the constellation into manageable slices and only evaluates relevant subsets, thereby reducing the computational complexity from O(M) to O(M/K) where K is the number of slices, while maintaining near-ML performance through selective evaluation of promising candidates
Solution Approach 2:
The patent applies local quality by focusing computational resources on specific regions or slices of the constellation that are most likely to contain the correct path. Rather than uniformly evaluating all constellation points, the algorithm identifies and concentrates search efforts in local regions with higher probability, optimizing the trade-off between performance and complexity
2Reliability
If the number of survivors is increased to maintain near ML performance at low SNR, then reliability improves, but complexity increases significantly
Solution Approach 1:
The patent applies partial action by evaluating only a subset of constellation points rather than all M points for each survivor. By using slicing to identify and evaluate only the most promising candidates in specific regions, the algorithm achieves near-ML performance with fewer evaluations, avoiding the need to increase the number of survivors and their associated complexity
3Measurement precision
If exhaustive path search is performed for all survivor candidates, then measurement precision of path metric improves, but processing time increases
Solution Approach 1:
The patent applies preliminary action by pre-dividing the constellation into slices and pre-identifying which slices are relevant for each survivor path before performing metric calculations. This preliminary organization allows the algorithm to skip unnecessary evaluations and directly compute metrics for only the most promising candidates, reducing processing time while maintaining path metric accuracy
Data Source
AI summary
A receiver is configured to receive a sample of an inter-symbol correlated (ISC) signal, the sample corresponding to a time instant when phase and/or amplitude of the ISC signal is a result of correlation among a plurality of symbols of a transmitted symbol sequence. The receiver may linearize the sample of the ISC signal. The receiver may calculate a residual signal value based on the linearized sample of the ISC signal. The receiver may generate an estimate of one or more of said plurality of symbols based on the residual signal value. The linearization may comprise applying an estimate of an inverse of a non-linear model. The non-linear model may be a model of nonlinearity experienced by the ISC signal in a transmitter from which the ISC signal originated, in a channel through which the ISC signal passed en route to the receiver, and/or in a front-end of the receiver.