Multi-Codeword MIMO Signal Detection Using Lattice Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In multi-codeword MIMO systems, the existing signal detection methods, such as ML, ZF, and MMSE, face challenges with high computational complexity and significant errors due to poor channel quality, especially with large numbers of antennas and ill-conditioned channel matrices.
Innovation Solution
A signal detecting method and device that converts complex number signals to real number signals, performs channel estimation, corrects the channel matrix using a transformation matrix, generates an equivalent detecting signal, and quantizes it to reduce errors, involving techniques like lattice reduction and LLL algorithms for matrix transformation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If ML detecting method is used, then detection precision is improved, but device complexity increases
Solution Approach 1:
The patent segments the complex ML detection process into two stages: first performing ZF or MMSE detection to obtain initial signal estimates, then performing lattice reduction only on the residual error. This segmentation allows the system to achieve ML-level detection precision while avoiding the exponential computational complexity of full ML detection by applying the computationally intensive lattice reduction operation only to reduced-dimensional residual signals rather than the complete signal vector.
Solution Approach 2:
The patent introduces an intermediary detection step using ZF or MMSE methods that produces preliminary signal estimates. These intermediate results serve as a bridge between the received signal and the final ML-detection-equivalent result, allowing the system to achieve high precision through a multi-stage process that is computationally more manageable than direct ML detection.
2Reliability
If MMSE detecting method is used, then reliability is improved, but manufacturing precision deteriorates
Solution Approach 1:
The patent applies lattice reduction continuously and iteratively to the residual error signals after ZF/MMSE detection. This continuous refinement process repeatedly reduces the lattice basis and updates the signal estimates, progressively improving detection accuracy beyond the initial MMSE results while maintaining the reliability benefits of noise-aware detection.
Solution Approach 2:
The patent implements a feedback mechanism where the residual error from MMSE detection is fed back into the lattice reduction process, which in turn generates corrected signal estimates that are fed back to refine the detection further. This feedback loop allows the system to iteratively improve detection accuracy while maintaining the reliability advantages of MMSE's noise consideration.
3Measurement precision
If lattice reduction is performed without quantization, then detection precision is improved, but loss of information increases
Solution Approach 1:
The patent changes the parameter representation of the detecting signal by performing quantization on the lattice reduction results. This parameter transformation converts continuous high-precision lattice coordinates into discrete quantized values that can be efficiently transmitted and processed, reducing information loss during transmission while preserving the essential detection precision gains from lattice reduction through careful quantization design.
Data Source
Figure 1~2
Figure 3
Figure 4~5
AI summary
A signal detecting method and device for a multi-codeword multi-input multi-output (MIMO) system are provided, which are used in the field of communication and solve the problem that it is complex to implement a signal detecting method in the multi-codeword MIMO system and the error between the signal obtained through detection and the signal actually transmitted from the transmitting end is large. The method includes: converting a received complex number signal into a real number signal; performing channel estimation to obtain a channel matrix; correcting the channel matrix to obtain an equivalent corrected matrix; generating an equivalent detecting signal according to the equivalent corrected matrix and the real number signal obtained through conversion; quantizing the equivalent detecting signal to obtain an equivalent quantized detecting signal; and generating a signal transmitted from a transmitting end according to the equivalent quantized detecting signal. The method and device can be applied in a receiving device of a multi-codeword MIMO system.