Scholarlink

<p>In non-cooperative communication systems, recognition of the modulation scheme at the receiver end is an important functionality in wireless communication systems. The research problem of automatic modulation classification is, therefore, gaining increasing significance in the areas of both civil and defense applications. In this work, we present a deep learning enabled automatic modulation classifier for a class of modulation schemes. Our proposal is based on a Convolutional-Long Short-Term Memory Deep Neural Network model architecture particularly focusing on low signal-to-noise ratio communication links. Radio ML 2016.10b dataset generated by GNU radio and recognized as the benchmark dataset is used for training and testing of the proposed modulation classifier. Through extensive experimental results, it is shown that the proposed model architecture offers significantly improved classification accuracy up to 90.85% at 0 dB signal-to-noise ratio of the communication link. Our model offers a competitive solution to automatic modulation classification problem in complex radio environment.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Deep learning-based automatic modulation classifier

  • 期刊论文

  • 发表日期：

    2025年6月16日

查看全文
  • Brahmjit Singh,
  • Vishakha Bajaj,
  • Poonam Jindal,
  • Chandra Prakash

摘要

In non-cooperative communication systems, recognition of the modulation scheme at the receiver end is an important functionality in wireless communication systems. The research problem of automatic modulation classification is, therefore, gaining increasing significance in the areas of both civil and defense applications. In this work, we present a deep learning enabled automatic modulation classifier for a class of modulation schemes. Our proposal is based on a Convolutional-Long Short-Term Memory Deep Neural Network model architecture particularly focusing on low signal-to-noise ratio communication links. Radio ML 2016.10b dataset generated by GNU radio and recognized as the benchmark dataset is used for training and testing of the proposed modulation classifier. Through extensive experimental results, it is shown that the proposed model architecture offers significantly improved classification accuracy up to 90.85% at 0 dB signal-to-noise ratio of the communication link. Our model offers a competitive solution to automatic modulation classification problem in complex radio environment.