MIMO Space-Time Decoding via Segmented Parallel Constellation Enumeration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The Selective Spanning with Fast Enumeration (SSFE) algorithm for space-time decoding in MIMO systems faces issues such as excessive power consumption, high processing delay, and error propagation due to cyclic search instructions and dynamic skip execution, which hinder real-time signal processing and resource utilization.
Innovation Solution
A new method, referred to as NSSFE, performs QR decomposition on the channel matrix, enumerates candidate constellation points using a fixed vector M, calculates Euclid distances, and decides space-time decoding based on these increments, avoiding the selection-decision process and ensuring all points are within the constellation, reducing calculation complexity and error propagation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the Selective Spanning with Fast Enumeration (SSFE) algorithm is used for space-time decoding, then detection performance is improved, but processing delay increases and power consumption becomes excessive
Solution Approach 1:
The patent segments the constellation points into multiple groups and processes them in parallel using multiple threads. Each thread handles a specific group of constellation points independently, eliminating the cyclic sequential processing of the SSFE algorithm. This segmentation enables simultaneous computation across different constellation point groups, significantly reducing processing delay while maintaining detection performance.
Solution Approach 2:
The patent performs preliminary organization of constellation points into structured groups before the main decoding process. By pre-arranging the candidate constellation points and their associated metrics in a format suitable for parallel processing, the algorithm eliminates the need for dynamic skip execution and cyclic search instructions during runtime, thereby reducing both processing delay and power consumption.
2Reliability
If the Selective Spanning with Fast Enumeration (SSFE) algorithm is used for space-time decoding, then detection performance is improved, but resource utilization rate decreases due to dynamic skip execution
Solution Approach 1:
The patent divides the decoding process into segmented parallel threads, each handling specific constellation point groups. This eliminates the dynamic skip execution characteristic of SSFE, where certain processing units remain idle while others work. All processing resources are continuously utilized across parallel threads, maximizing resource utilization rate while preserving the enhanced detection performance of SSFE.
Solution Approach 2:
The patent transforms the static sequential structure of SSFE into a dynamic parallel architecture. By using multiple threads that can be simultaneously executed, the system adapts to available computational resources and maintains high resource utilization. The parallel structure allows the algorithm to dynamically utilize all available processing units without the waste caused by dynamic skip instructions.
3Ease of manufacture
If traditional MIMO detection algorithms are used, then system implementation is simpler, but error propagation occurs reducing detection accuracy
Solution Approach 1:
The patent implements a parallel thread structure where each thread independently processes specific constellation point groups without relying on sequential decisions from previous steps. This eliminates the error propagation mechanism present in traditional algorithms where incorrect decisions in early stages affect subsequent detections. The independent parallel processing ensures that errors do not propagate across the detection process, improving detection accuracy while maintaining implementation simplicity.
4Productivity
If the number of antennas is increased to improve transmission rate, then system rate increases linearly, but algorithm complexity grows exponentially
Solution Approach 1:
The patent segments the large-scale MIMO detection problem into multiple smaller sub-problems handled by different parallel threads. Each thread processes a subset of constellation points for the given number of antennas, preventing the exponential complexity growth that would otherwise occur with increased antenna count. This segmentation allows the system to scale to higher transmission rates with more antennas while maintaining manageable per-thread complexity.
Solution Approach 2:
The patent introduces a parallel processing dimension by using multiple threads simultaneously executing detection algorithms. This transforms the single-thread exponential complexity problem into a multi-threaded problem where the total complexity is distributed across threads. The parallel dimension allows the system to handle increased antenna counts and transmission rates without experiencing exponential complexity growth in any single processing unit.
Data Source
Figure 1~2
Figure 3~4
Figure 5
AI summary
The embodiment of present invention discloses a processing method and apparatus for performing space-time decoding on MIMO signals. The method includes: step 1, performing QR decomposition on a channel matrix H with a matrix size of Nr×Nt in a MIMO system and performing an initialization operation, wherein Nt is the number of transmitting antennas and Nr is the number of receiving antennas; step 2, enumerating candidate constellation points of transmitting signals at each level of transmitting antennas according to an input parameter M = [M1, M2,···, MNt], a receiving signal y and a result of the QR decomposition of the channel matrix H, wherein M is the number of the numerated candidate constellation points at each level, y=Hs+n, s is a transmitting signal with a matrix size of Nt×1, n is white gaussian noise with a matrix size of Nr×1; step 3, calculating increments of a Euclid distance due to addition of selection of a candidate constellation point at present level according to the obtained candidate constellation points; and step 4, deciding and outputting space-time decoding according to the sum of calculated increments of Euclid distances of candidate constellation points at all levels.