MIMO Symbol Decoder Hybrid Linear Equalization and Maximum Likelihood Decoding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Multi-antenna systems face a complexity-performance tradeoff in decoding spatially multiplexed streams, particularly with higher modulation levels and a greater number of spatial streams, where increased decoding complexity is required for better performance.
Innovation Solution
Combining linear equalization decoding and maximum likelihood decoding (MLD) to achieve improved decoding performance with lower computational complexity, by grouping symbols for either linear equalization or MLD based on channel orthogonality or signal-to-interference and noise ratios (SINRs).
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If maximum likelihood decoding (MLD) is used to achieve high decoding performance, then decoding accuracy is improved, but computational complexity increases significantly
Solution Approach 1:
The patent segments the decoding process into two distinct stages: first applying linear equalization to pre-process and simplify the received signal, then applying maximum likelihood decoding only to the reduced-complexity intermediate results. This segmentation allows the system to benefit from both low-complexity linear equalization and high-accuracy MLD without bearing the full computational burden of standalone MLD.
Solution Approach 2:
Linear equalization serves as an intermediary step between signal reception and final MLD processing. This intermediate processing stage transforms the original complex decoding problem into a simplified version that requires fewer computational operations while maintaining the accuracy benefits of MLD for the final detection stage.
2Device complexity
If linear equalization decoding is used to reduce computational complexity, then decoding complexity is reduced, but decoding performance deteriorates
Solution Approach 1:
The patent merges two previously separate decoding approaches (linear equalization and MLD) into a unified hybrid decoder that combines their strengths. The linear equalization component handles the bulk of computational work with low complexity, while the MLD component provides accuracy enhancement, achieving a synergistic effect where the sum is greater than individual methods alone.
Solution Approach 2:
Different decoding strategies are applied to different parts of the decoding process based on their respective strengths: linear equalization is applied where computational efficiency is paramount, while MLD is applied where decoding accuracy is most critical. This localized application of different methods optimizes overall system performance.
3Productivity
If the number of spatial streams is increased to enhance data throughput, then system capacity is improved, but decoding complexity increases
Solution Approach 1:
By segmenting the decoding process into linear equalization and MLD stages, the patent enables the system to handle increased numbers of spatial streams more efficiently. The linear equalization stage scales better with increased stream count, preventing exponential complexity growth that would otherwise occur with pure MLD approaches.
Data Source
AI summary
Embodiments of MIMO symbol decoders and methods for decoding spatially multiplexed symbols using combined linear equalization decoding and maximum likelihood decoding are generally described herein. In some embodiments, one group of symbols (y) received through two or more spatial channels is decoded using a linear equalization decoding technique and another group of symbols is decoded using a maximum likelihood decoding (MLD) technique to generate a soft-bit output. The symbols may be grouped for either linear equalization decoding or MLD based on channel orthogonality or signal-to-interference and noise ratios (SINRs).