Lattice Enumeration-Aided MIMO Detection Reducing Complexity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing MIMO detection algorithms face complexity issues that increase exponentially with the number of transmit antennas, leading to impractical solutions with significant performance sacrifices.
Innovation Solution
The implementation of a lattice enumeration-aided detector (LEAD) that approximates a hyperellipsoid detection search space using eigenvectors and eigenvalues of the effective channel, allowing for improved detection in a regular alphabet independent of the channel, and employing QR decomposition and successive interference cancellation to simplify the detection process.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If the number of transmit and receive antennas is increased to increase data throughput and diversity, then system capacity increases linearly and fading probability decreases exponentially, but detection complexity increases significantly
Solution Approach 1:
The patent segments the detection process into two phases: first identifying a reduced subset of candidate symbols from the complete signal constellation, then performing final detection only on this reduced set. This segmentation reduces the exponential search space into manageable segments, lowering detector complexity while maintaining performance in MIMO systems with multiple antennas.
2Measurement precision
If the optimal brute force detector is used to achieve best performance, then detection accuracy is maximized, but complexity increases exponentially with the number of channel inputs
Solution Approach 1:
The patent applies partial action by performing exhaustive search only on a reduced subset of candidate symbols rather than the complete constellation. The detector identifies and evaluates only the most promising candidates based on initial metrics, performing partial exhaustive search that achieves near-optimal performance with significantly reduced computational complexity compared to complete brute force detection.
3Loss of information
If list-sphere detection is used to compute log-likelihood ratio information, then reliability information for each bit is provided, but processing resources are significantly consumed
Solution Approach 1:
The patent segments the LLR computation process by first identifying a reduced candidate subset and then computing reliability information only for these candidates. This segmentation allows the system to provide bit-level reliability information through LLR computation while consuming significantly fewer processing resources compared to computing LLRs for the complete signal constellation.
Data Source
AI summary
Embodiments provide systems and methods for improved multiple-input, multiple-output (MIMO) detection comprising generating at least one list of candidate vectors by employing lattice enumeration which approximates hyperellipsoid detection search space and calculating a reliability of the candidate vectors. At least one advantage to embodiments is that improved detection occurs because detection can be performed in a search space defined by the eigenvectors (which define the general shape of an ellipsoid/hyperellipsoid, depending upon number of dimensions) and eigenvalues (which provide the appropriate scaling in each direction of the eigenvectors) of the effective channel.