MIMO Detection Reducing Complexity via Channel Matrix Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current MIMO detection methods face high complexity, particularly in higher-order scenarios, making them difficult to implement in practice while lacking a balance between performance and simplicity.
Innovation Solution
The method involves dividing the channel matrix into groups, solving independent equations, and combining optimal solutions to detect the transmit signal vector, reducing complexity by at least one order of magnitude without compromising performance. This is achieved by removing column vectors, performing singular value decomposition, calculating Euclidean distances, and selecting appropriate transmit signal vectors to form a second transmit signal vector set for detection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If maximum likelihood detection (MLD) algorithm is used for MIMO detection, then detection performance is improved, but detection complexity increases significantly
Solution Approach 1:
The channel matrix H is divided into multiple sub-channels by removing groups of column vectors, transforming the original large-scale MIMO detection problem into multiple smaller independent detection problems. This segmentation reduces the search space from Q^MT to (Q^K)^(MT/N), significantly lowering computational complexity while maintaining detection performance through subsequent combination of results.
2Productivity
If the quantity of transmit antennas MT increases to achieve higher spectrum efficiency, then system capacity is improved, but detection complexity increases exponentially
Solution Approach 1:
By removing N column vectors corresponding to N transmit antennas to form sub-channels, the method enables scalable MIMO detection. When MT increases, the system simply divides into more sub-channels, keeping each sub-channel's detection complexity manageable while linearly increasing overall system capacity.
Data Source
AI summary
Embodiments of the present patent application relate to a MIMO detection method and apparatus for a transmit signal. The method includes removing one group of column vectors from an original channel matrix to generate a corresponding first channel matrix and multiplying a conjugate transpose matrix of a first matrix by the first channel matrix to obtain an equivalent channel matrix, and multiplying the conjugate transpose matrix of the first matrix by an original received signal vector to obtain an equivalent received signal vector. The method also includes calculating a Euclidean distance between the equivalent received signal vector and a first vector to obtain a Euclidean distance set. The method also includes obtaining a second transmit signal vector set by using the Euclidean distance set, and detecting a transmit signal vector according to the original received signal vector, the original channel matrix, the second transmit signal vector and the second transmit signal vector set.