Multilayer Perceptron Modulation Recognition via Data Fusion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing modulation recognition algorithms for MIMO systems face high complexity and low recognition precision, especially in large-scale systems with high-order modulation modes, due to exponential calculation growth and limited handling of additive noise.
Innovation Solution
A fast modulation recognition method using multimodally-distributed test data fusion with preprocessing, decision statistics data generation, and input feature creation for an MLP classifier, which reduces algorithm complexity and improves recognition precision through normalization, distribution tests, and data fusion.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If likelihood-based modulation recognition algorithms are used in large-scale MIMO systems, then recognition precision can be improved, but calculation complexity grows exponentially
Solution Approach 1:
The patent segments the calculation process by pre-calculating and storing likelihood values for different modulation modes and signal-to-noise ratios in lookup tables. During actual recognition, the algorithm only needs to perform table lookups and simple comparisons rather than complex exponential calculations, thus maintaining high precision while reducing real-time computational complexity
Solution Approach 2:
The patent performs preliminary calculations by pre-computing likelihood values for various modulation modes and SNR conditions before actual signal recognition. These pre-computed values are stored in lookup tables, allowing the system to quickly retrieve and compare pre-calculated results during runtime without performing expensive exponential calculations in real-time
2Measurement precision
If eigenvalue-based recognition mechanisms are used, then recognition precision can be improved, but anti-interference capability deteriorates due to inability to handle additive noise
Solution Approach 1:
The patent changes the approach from direct eigenvalue analysis to likelihood-based recognition that incorporates signal-to-noise ratio as a key parameter. By using likelihood functions that explicitly model noise characteristics and SNR conditions, the system can distinguish between signal variations caused by modulation and those caused by noise, thereby maintaining high precision while improving anti-interference capability
3Productivity
If high-order modulation modes are considered in large-scale MIMO systems, then communication capacity is improved, but calculation complexity grows exponentially
Solution Approach 1:
The patent segments the complex calculation by organizing likelihood values into structured lookup tables indexed by modulation mode and SNR level. This segmentation allows the system to handle high-order modulation modes efficiently by breaking down the recognition process into discrete table lookups and comparisons rather than continuous complex calculations
Solution Approach 2:
The patent performs preliminary computation of likelihood values for all possible high-order modulation modes and SNR conditions beforehand. These pre-computed results are stored in lookup tables, enabling the system to support high-order modulation modes that increase communication capacity while avoiding exponential calculation complexity during actual signal recognition through simple table queries
Data Source
AI summary
The present invention discloses a fast modulation recognition method for a multilayer perceptron (MLP) based on multimodally-distributed test data fusion. The method sequentially includes: preprocessing a received signal, obtaining a signal feature sequence, generating a matrix of decision statistics data o*hj−, generating an MLP an input feature by fusing the decision statistics data, recognizing a modulation mode by using the MLP, and matching an output with a corresponding classification label. The present invention has a low algorithm complexity as compared with a classical likelihood algorithm, and at the same time improves the recognition precision of a single distribution test algorithm.