efthimia145
commented
3 years ago Modulation classification methods:
- Advanced image processing with Machine Learning Algorithms -> fail in noisy constellation patterns
- Fuzzy c-means algorithm -> statistical features -> Bayes classifier (Mobasseri)
- CNN for intelligent constellation diagram analysis: 2 conv layers connected with pooling layers / 6-modulation-format (Wang)
- CNN applied to a 3-channel color image conversion of the constellation diagram. (Peng)
- CNN applied to a combination of a regular constellation image and a contrast enhanced grid constellation image. - mostly for improvement in the accuracy of high order modulations(Huang)
- CNN applied to spectrogram images after a time-frequency analysis. CNN -> 4 conv neural networks
- Automatic modulation classification
- (more soon..)
We need to perform literature review for modulation classification methods. We have to collect and categorize the existing methods.