Modulation classification is an important part of applications such as operator regulation, communications anti-jamming, user identification and cognitive radio. You can use deep learning algorithms to classify channel impaired signals. This example shows how to train a deep learning algorithm with synthetically generated channel-impaired waveforms for modulation classification. The example also uses two ADALM-PLUTO radios to test the trained network with real world signals.
Presenter: ETHEM MUTLU SÖZER is a principal software engineer at MathWorks Inc. in Natick, MA. He specializes in software development for signal processing and communications toolboxes to support SDR. Previously he was a research engineer at Massachusetts Institute of Technology, where he developed underwater acoustic communication hardware and software platforms. He has bachelor’s and master’s degrees from Middle East Technical University, and a Ph.D. from Northeastern University.