Accuracy Enhancement of Automatic Modulation Recognition Using Deep Learning Paradigm
Salah Ayad Jassim, Ibrahim Khider
Abstract
Communication systems consist of advanced system performance that enables signal reception at the destination. In such a system, the difficult problem is to restore the original transmission. In this study, Automatic Modulation Recognition (AMR) is used to improve the precision of modulation recognition. This method is essential for advanced communication systems that require a minimal delay, such as Realtime broadcasting. Using paradigms of deep learning, the modulation approach of a received signal is identified. Feedforward neural network with Particle Swarm Optimization (PSO) integration is proposed for this purpose. The suggested model exhibits optimal recognition accuracy of 97.3 percent, as reported.
Topics & Concepts
Computer scienceModulation (music)Feed forwardParticle swarm optimizationTransmission (telecommunications)Artificial intelligenceArtificial neural networkSIGNAL (programming language)Communications systemDeep learningSpeech recognitionPattern recognition (psychology)Machine learningTelecommunicationsControl engineeringEngineeringPhilosophyAestheticsProgramming languageWireless Signal Modulation Classification