MIMO ML Demodulation Using Bitwise Constellation Partitioning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current MIMO communication systems face challenges in efficiently decoding signals due to high complexity and processing requirements, especially when dealing with multiple antennas and varying channel conditions, which can lead to increased processing time and resource utilization.
Innovation Solution
The implementation of a multi-stream maximum-likelihood (ML) demodulation method that calculates likelihood values without explicitly calculating distance values, using techniques like Maximal Ratio Combining and symbol slicing to reduce computational complexity by grouping candidate values and determining global or local optimum values efficiently.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If conventional MIMO demodulation methods are used, then signal decoding can be performed, but computational complexity and processing time increase significantly
Solution Approach 1:
The patent applies segmentation by dividing the constellation into multiple subsets based on bit positions. Instead of evaluating all possible constellation points globally, the algorithm segments the search space into smaller subsets corresponding to different bit positions, allowing independent optimization for each bit and significantly reducing computational complexity.
Solution Approach 2:
The patent changes the parameter representation by using likelihood values directly without explicitly calculating distance values. By transforming the optimization criterion from distance-based to likelihood-based, the algorithm avoids computationally intensive distance calculations while maintaining decoding accuracy.
2Measurement precision
If maximum-likelihood demodulation is implemented, then decoding accuracy is improved, but processing time increases
Solution Approach 1:
The patent segments the demodulation process into bit-position-specific operations. By dividing the constellation evaluation into smaller subsets based on bit positions, the algorithm maintains maximum-likelihood accuracy while reducing the time required to evaluate each candidate point through localized optimization.
Solution Approach 2:
The patent performs preliminary organization of constellation points into subsets based on bit positions before the actual demodulation process. This preliminary action allows the algorithm to quickly access and evaluate only relevant candidate points for each bit, avoiding unnecessary computations and reducing overall processing time.
3Ease of operation
If explicit distance value calculation is performed, then demodulation can be completed, but computational burden increases
Solution Approach 1:
The patent fundamentally changes the parameter used for demodulation by working directly with likelihood values instead of calculating explicit distance values. This parameter transformation eliminates computationally burdensome distance calculations while preserving the ability to complete demodulation accurately through likelihood-based comparison.
Solution Approach 2:
The patent extracts and removes the computationally intensive distance calculation step from the demodulation process. By taking out the explicit distance value calculation and replacing it with direct likelihood value usage, the algorithm significantly reduces computational burden while maintaining demodulation functionality.
Data Source
AI summary
A method of decoding a signal includes receiving multiple data symbols corresponding to multiple transmitted data symbols. The method further includes selecting one candidate data symbol from multiple candidate data symbols corresponding to the multiple transmitted data symbols. The method further includes determining a global optimum candidate value for the selected one candidate data symbol. The method further includes grouping all possible values of the selected one candidate data symbol into two or more bit groups. The method further includes selecting a bit group from the two or more bit groups that does not include the global optimum value. The method further includes determining a local optimum candidate value from values in the selected bit group. The method further includes calculating a likelihood value for a bit in the selected one candidate data symbol based at least in part on the determined global and local optimum candidate values.