SM-MIMO Signal Detection Using QR Decomposition and Interference Cancellation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Signal detection schemes in SM-MIMO communication systems, such as ZF, MMSE, OSIC, sphere decoding, and QRD-M, suffer from performance deterioration and increased calculation complexity compared to the ML scheme, making it difficult to achieve optimal LLR values and channel decoding performance.
Innovation Solution
The proposed apparatus and method involve QR decomposition of the channel matrix to generate Q and R matrices, determining detection orders, and sequentially removing interference to detect candidate symbols and calculate LLR values, thereby reducing calculation complexity while maintaining performance similar to the ML scheme.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the ML scheme is used for signal detection, then detection performance is improved, but calculation complexity increases significantly
Solution Approach 1:
The patent segments the signal detection process into multiple stages: QR decomposition of the channel matrix, ordered symbol detection based on diagonal elements, and interference cancellation. This segmentation transforms the complex ML detection into manageable sequential steps, reducing overall calculation complexity while maintaining performance.
Solution Approach 2:
The patent performs preliminary QR decomposition of the channel matrix before actual symbol detection. This preliminary action transforms the detection problem into a simpler form where symbols can be detected sequentially with reduced complexity, while the QR decomposition results are reused throughout the detection process.
2Device complexity
If simplified schemes like ZF or MMSE are used, then calculation complexity is reduced, but detection performance deteriorates significantly
Solution Approach 1:
The patent implements interference cancellation based on previously detected symbols. Each detected symbol is used to cancel its interference contribution from subsequent symbols, creating a feedback mechanism that progressively improves detection accuracy. This feedback approach allows the system to achieve performance close to ML while maintaining lower complexity.
3Measurement precision
If sphere decoding is used to achieve ML performance, then detection performance is improved, but calculation quantity becomes very large for worst channel conditions
Solution Approach 1:
The patent dynamically determines the detection order based on the diagonal elements of the R matrix from QR decomposition. Symbols are detected in ascending order of their corresponding diagonal elements, which adaptively optimizes the detection sequence for current channel conditions. This dynamic approach avoids the exhaustive search required by sphere decoding while maintaining performance.
Data Source
AI summary
A Spatial Multiplexing Multiple Input Multiple Output (SM-MIMO) communication system, which generates Q and R matrices through QR decomposing a channel matrix in an array order corresponding to an order of transmission antennas of a signal transmission device, determines a detection order of transmission symbols included in a transmission symbol vector, QR decomposes a reception signal vector and generates the QR decomposed reception signal vector as a transformed reception signal vector, substitutes all of transmittable candidate symbols for a transmission symbol corresponding to a first detection order among the transmission symbols and sequentially removes interference in the transformed reception signal vector, detects candidate symbols of transmission symbols corresponding to a remaining detection order from the interference-removed transformed reception signal vector, combines the detected candidate symbols, detects candidate symbol vectors, and calculates Log-Likelihood Ratio values of bits included in the transmission symbols by using the detected vectors.