Automatic Modulation Classification: A Deep Architecture Survey

Automatic modulation classification (AMC), which aims to blindly identify the type of an incoming signal at receiver in wireless communication systems, is a fundamental processing technique physical layer improve spectrum utilization efficiency. Motivated by deep learning (DL) high-impact success many informatics domains, including radio for communications, numerous recent AMC methods exploiting networks have been proposed overcome existing drawbacks traditional approaches. DL capable underlying characteristics signals effectively pattern recognition, turn improves performance under presence channel impairments. In this work, we first provide concepts various architectures, such as neural networks, recurrent long short-term memory, and convolutional necessary background. We then convey comprehensive study where technical analysis deliberated perspective state-of-the-art architectures. Remarkably, several sophisticated structures advanced designs are investigated different data types sequential signals, images, constellation images deal with Finally, discuss some primary research challenges potential future directions area classification.