Quasi-MLD MIMO Signal Detection via Disjoint Subgroup Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing wireless communication systems face challenges in achieving high spectral efficiency and low complexity in Multiple Input Multiple Output (MIMO) transmissions, particularly for large MIMO systems like 4x4 and 8x8 MIMO, due to the high computational complexity of Maximum Likelihood Detection (MLD) and sub-optimal performance of Linear Minimal Mean Square Error (LMMSE) receivers, especially under varying channel conditions and layer numbers.
Innovation Solution
A method and user equipment (UE) that split received MIMO signals into disjoint subgroups, allowing for simultaneous detection using MLD, with an intelligent splitting scheme based on achievable information data rates and interference cancellation, reducing complexity and improving performance through Log-Likelihood Ratio combination and QR-decomposition.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If Maximum Likelihood Detection (MLD) is used for large MIMO systems (4x4, 8x8), then detection performance is improved, but computational complexity becomes prohibitively high
Solution Approach 1:
The patent divides the MIMO signal into G disjoint subgroups, where each subgroup contains a subset of MIMO layers. This segmentation allows the detection process to be divided into multiple smaller detection tasks rather than one large complex task. Each subgroup is detected independently using MLD, reducing the computational burden while maintaining overall detection performance through the combination of subgroup results.
2Device complexity
If near-MLD algorithms (SOSD, QRD-M, K-Best) are used to reduce complexity, then computational complexity is reduced, but detection performance deteriorates
Solution Approach 1:
By segmenting the MIMO signal into G subgroups, the patent enables the use of MLD on smaller subsystems rather than the full system. This segmentation approach allows near-MLD algorithms to operate on reduced complexity subproblems while collectively achieving performance close to full MLD, thus balancing complexity reduction with performance maintenance.
Solution Approach 2:
The patent applies partial MLD by detecting only a subset of MIMO layers at a time through the G subgroups. Instead of performing complete MLD on all L layers simultaneously, the system performs MLD on partial subgroups and combines the results, achieving sufficient performance without the full computational burden of complete MLD.
3Productivity
If the number of MIMO layers is increased to achieve high spectral efficiency, then spectral efficiency is improved, but detection complexity increases exponentially
Solution Approach 1:
The patent segments the MIMO layers into G disjoint subgroups, where each subgroup contains a manageable number of layers. This segmentation transforms the exponential complexity growth into a more manageable structure, as the detection complexity becomes proportional to the number of subgroups times the complexity of detecting a single subgroup, rather than the exponential complexity of detecting all layers simultaneously.
Solution Approach 2:
The patent introduces flexible parameters including the number of subgroups G and the size of each subgroup, which can be dynamically adjusted based on the number of MIMO layers L and channel conditions. This dynamic adaptability allows the system to maintain optimal performance while controlling complexity as the number of layers increases to achieve higher spectral efficiency.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A UE and a method for a UE are presented. The UE comprises a receiver unit and a processing circuit, and is configured for receiving wireless signals. The processing circuit is arranged for performing pre-detection of the received signals providing an initial estimation of transmitted signals. The processing circuit is also arranged for splitting the transmitted signal into disjoint subgroups, each one covering a subgroup of all layers used for the transmitted signal such that the subgroups together cover all the layers. The processing circuit is also arranged for interference cancellation performed on the subgroups of transmitted signals based on the initial estimation of the transmitted signals. The processing circuit is also arranged for detection of the subgroups of transmitted signals by utilisation of an MLD algorithm, wherein the subgroup of layers within each one of the subgroups of transmitted signals are detected simultaneously.