Sphere Decoding Detection with Adaptive Radius
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing sphere decoding detection methods face high computational complexity, particularly in hardware implementation, and suffer from performance loss in signal detection, especially in low SNR regions and for low-order modulation.
Innovation Solution
A sphere decoding detection method that performs pre-processing using a semi-definite relaxation detector to obtain an initial signal estimation, adjusts the constellation space size based on SNR, and employs depth-first and sphere constraint rules to search for optimal signal points within a multidimensional sphere, updating the radius dynamically to reduce the number of nodes accessed in the tree search.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If standard sphere decoding detection method is used, then bit error performance approaches MLD, but computational complexity reaches exponent level making hardware implementation difficult
Solution Approach 1:
The patent implements dynamic adjustment of the initial radius parameter based on signal-to-noise ratio conditions. The radius is adapted dynamically: smaller radius in high SNR regions reduces search space and complexity, while larger radius in low SNR regions maintains detection performance. This dynamic parameter adjustment resolves the contradiction by making complexity adaptive to channel conditions rather than fixed at exponent level.
Solution Approach 2:
The patent changes key parameters of the sphere decoding algorithm: (1) initial radius parameter adjusted according to SNR, (2) modulation order adaptation to limit constellation space size, (3) depth-first search order optimization. These parameter changes transform the fixed high-complexity MLD-equivalent algorithm into an adaptive algorithm whose complexity can be controlled while maintaining acceptable performance across different SNR regions.
2Device complexity
If Fincke-Pohst SD algorithm is used with fixed initial radius, then search space is narrowed once, but selection of initial radius D is sensitive and causes performance loss
Solution Approach 1:
The patent makes the initial radius parameter dynamic rather than fixed. The radius is determined adaptively based on the current SNR condition: in high SNR regions, a smaller radius suffices to achieve good performance with reduced complexity; in low SNR regions, a larger radius is used to ensure adequate search space coverage. This dynamic adjustment eliminates the sensitivity problem of fixed radius selection.
Solution Approach 2:
The patent changes the initial radius parameter based on SNR conditions and modulation order. By adjusting this critical parameter, the algorithm achieves a balance between search space reduction (complexity control) and performance maintenance. The parameter change strategy transforms the rigid Fincke-Pohst approach into a flexible adaptive algorithm.
3Device complexity
If Schnorr-Euchner algorithm with depth-first search is used, then complexity is reduced, but maximum number of nodes M limitation causes signal performance loss
Solution Approach 1:
The patent implements dynamic adjustment of the effective search space through adaptive initial radius selection based on SNR. In high SNR conditions, the adaptive radius naturally limits the number of nodes to be searched, achieving complexity reduction without performance loss. In low SNR conditions, the larger adaptive radius allows sufficient node exploration to maintain performance. This dynamic approach replaces the fixed node limit M with a more flexible adaptive mechanism.
Solution Approach 2:
The patent changes the initial radius parameter adaptively based on SNR and modulation order, which indirectly controls the number of nodes accessed in the depth-first search. This parameter change strategy allows the algorithm to achieve complexity reduction in high SNR regions while maintaining adequate performance in low SNR regions, avoiding the performance loss caused by fixed node limits.
4Device complexity
If QR preprocessing based depth-first SD detection is used, then complexity is reduced in high SNR region, but it is not suitable for low SNR region detection
Solution Approach 1:
The patent enhances the universality of the sphere decoding algorithm by making it effective across both high SNR and low SNR regions. Through adaptive initial radius selection and modulation order adaptation, the algorithm achieves good performance in high SNR regions (with reduced complexity) while also maintaining adequate performance in low SNR regions. This multi-functional capability resolves the limitation of QR-preprocessing-based methods that work only in high SNR regions.
Solution Approach 2:
The patent changes the initial radius parameter adaptively based on SNR conditions, enabling the algorithm to function effectively in both high and low SNR regions. In high SNR regions, the smaller adaptive radius reduces complexity while maintaining performance. In low SNR regions, the larger adaptive radius ensures sufficient search space coverage. This parameter adaptation strategy extends the algorithm's versatility across different SNR regions.
Data Source
Figure 1
Figure 2
Figure 3~4
AI summary
Disclosed are a sphere decoding detection method and apparatus, including: preprocessing a received signal to obtain a signal approximate estimation value Xpre of the received signal, deducing an initial square radius D2 of sphere decoding detection according to Xpre, and determining the size l of a constellation space according to the current signal to noise ratio of the received signal; according to depth first and sphere constraint rules, searching for a search path depending on the size l of the constellation space and an initial square radius D2, and all nodes through which the search path passes falling in a sphere which takes the initial square radius as a radius; after a search path is searched out, and when the sum of local Euclidean distances of the searched-out search path is less than the current square radius, updating the square radius, and re-searching for a search path in a multi-dimensional sphere which uses the received signal as the center of the sphere and the updated square radius as the radius until a search path cannot be searched out, and determining a candidate signal point corresponding to the latest saved search path as the optimum signal estimation point.