MIMO Signal Detection Using Substitute Vectors and Thresholds
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In multiple input multiple output (MIMO) systems using spatial multiplexing, existing signal detection methods such as maximum likelihood (ML) bit metric detection require high complexity, while reduced complexity methods like zero forcing (ZF) and minimum mean square estimator (MMSE) degrade performance, and non-linear methods like ordered successive interference cancellation (OSIC) also fall short compared to ML.
Innovation Solution
A method involving the detection of substitute vectors, calculation of metrics based on square roots of Euclidean distances, and determination of threshold values to calculate soft values for improved transmission symbol detection with reduced complexity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If maximum likelihood (ML) bit metric detection is used for optimal transmission signal detection, then detection performance is improved, but hardware complexity is exponentially increased with respect to the size of constellation and the number of transmitting antennas
Solution Approach 1:
The patent segments the detection process into two stages: first performing linear detection (ZF or MMSE) to obtain initial signal estimates, then applying non-linear processing (OSIC or ML detection) only to refine the results. This segmentation allows the system to achieve near-ML performance while avoiding the exponential complexity of pure ML detection across all antennas and constellation sizes.
Solution Approach 2:
The patent applies preliminary linear detection (ZF or MMSE) before performing the final ML detection. By pre-processing the received signal to obtain initial signal estimates and cancel interference in advance, the system reduces the search space for ML detection, thereby lowering hardware complexity while maintaining detection performance.
2Device complexity
If zero forcing (ZF) or minimum mean square estimator (MMSE) methods are used to reduce complexity, then hardware complexity is reduced, but detection performance is degraded compared to ML bit metric detection
Solution Approach 1:
The patent merges linear detection methods (ZF or MMSE) with non-linear detection methods (OSIC or ML detection) into a hybrid detection scheme. The linear detection provides initial estimates and interference cancellation, while the non-linear detection refines the results, combining the advantages of both approaches to achieve low complexity with high performance.
Solution Approach 2:
The patent uses linear detection (ZF or MMSE) as a preliminary step before final detection. This preliminary action provides initial signal estimates that guide the subsequent non-linear detection process, allowing the system to achieve near-ML performance with significantly reduced complexity compared to pure ML detection.
3Ease of operation
If ordered successive interference cancellation (OSIC) is used to reduce complexity and improve implementation ease, then ease of operation is improved, but detection performance is degraded compared to ML bit metric detection
Solution Approach 1:
The patent applies OSIC as a preliminary interference cancellation step before performing final ML detection. By successively canceling interference from detected signals and updating the received signal, the system creates better conditions for the final detection stage, improving performance while maintaining implementation ease.
Solution Approach 2:
The patent segments the detection process into multiple stages: OSIC-based interference cancellation followed by final ML detection. This segmentation allows the system to benefit from the implementation simplicity of OSIC while achieving near-ML performance through the final detection stage.
Data Source
AI summary
The present invention relates to a method of calculating a soft value and a method of detecting a transmission signal. The present invention estimates a channel on the basis of a received signal and rearranges a plurality of data streams. Further, a plurality of substitute vectors are selected from the rearranged data streams and a metric corresponding to each of the substitute vector is calculated. Further, a threshold value is calculated from a metric calculated for each of the substitute vectors and a soft value of each bit of a transmission signal is calculated from the metric and threshold value corresponding to each of the substitute vectors.