Multi-Level Modulation Soft Information via Nearest-Neighbor Mapping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Calculating soft information in multi-level modulation systems is complex due to the need for high computational resources and power consumption, especially when determining the likelihood of bit values in noisy signals with many possible symbol mappings.
Innovation Solution
A low complexity technique is employed by using a second symbol mapping to find nearest neighbors in the signal constellation set, which are then translated back into the original symbol mapping to estimate digital information, reducing computational complexity and power consumption.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If full-precision log-likelihood ratio computation is performed by comparing received signal to each signal point in the multi-level constellation set, then measurement precision is improved, but device complexity increases
Solution Approach 1:
The patent applies partial action by computing log-likelihood ratios only for a limited set of candidate symbols rather than all possible symbols. The detector identifies a reduced subset of symbols that are most likely to represent the transmitted signal based on the received signal characteristics, and performs precise LLR computation only for these candidates. This partial computation approach maintains adequate measurement precision while significantly reducing device complexity compared to exhaustive computation.
2Productivity
If soft information is computed frequently to determine digital information in multi-level systems, then productivity is improved, but use of energy increases
Solution Approach 1:
The patent reduces energy consumption by performing partial computations instead of exhaustive symbol comparisons. The detector uses a simplified approach to identify candidate symbols and computes soft information only for these limited candidates, enabling frequent computations at lower power levels. This maintains productivity by achieving sufficient detection speed while dramatically reducing the energy cost per computation.
Solution Approach 2:
The detection process is segmented into multiple stages: initial candidate symbol identification based on simplified criteria, followed by precise soft information computation only for these candidates. This segmentation allows the system to perform detections frequently by breaking the complex computation into lighter, more energy-efficient steps that can be executed rapidly.
3Productivity
If the number of signal levels is increased to increase system capacity, then productivity is improved, but device complexity increases
Solution Approach 1:
The patent enables high-capacity multi-level modulation systems to operate with reduced complexity by computing soft information for only a subset of candidate symbols rather than all possible symbols. As the number of signal levels increases, the exhaustive computation of LLRs for all symbols becomes prohibitively complex, but this patent identifies and computes only for the most likely candidates, maintaining high system capacity while controlling device complexity.
Data Source
AI summary
Apparatus and methods are provided for calculating soft information in a multi-level modulation scheme using one or more nearest neighbors. The nearest neighbors correspond to signal points in a signal constellation set nearest to the value of a received signal. The nearest neighbors of a received signal can be found by using a second symbol-to-signal point mapping for the signal constellation set that is different from the mapping actually used by the signal modulator. The second symbol mapping can be used to simplify the discover of nearest neighbors. Once the nearest neighbors are found in the second symbol mapping, the nearest symbols can be translated back into the actual symbol mapping using, for example, table lookup. The nearest neighbors in the actual symbol mapping can then be used to compute soft information in the form of, for example, log-likelihood ratios (LLRs).