| Autor: |
N. Ambika, S. Radha, K. Muthumeenakshi |
| Rok vydání: |
2021 |
| Předmět: |
|
| Zdroj: |
Advances in Automation, Signal Processing, Instrumentation, and Control ISBN: 9789811582202 |
| Popis: |
The exponential advancement in the wireless networks increases the demand for radio spectrum. Sometimes, it leads to under-utilization of the spectrum, if the spectrum is not shared appropriately among the users. Cognitive radio is the technology that is competent of addressing the insufficiency of available radio spectrum through dynamic spectrum access. Primary users are the users of licensed spectrum, who may not use their authorized spectrum portions all the time. The spectrum holes are created when they do not use their authorized spectrum portions. The spectrum whole or white space is a band that is allocated to a PU, which is not being used all the time. They are usually identified by sensing the spectrum with respect to time and space. Machine learning and deep learning techniques are used to identify and classify the occupancy of a licensed spectrum. In this proposed work, a deep convolutional neural network model is used to classify the occupancy state of PU. The spectrogram dataset available from DySPAN is used, and overall classification accuracy of 98.99% is achieved. The performance curve shows the efficiency of the trained model. |
| Databáze: |
OpenAIRE |
| Externí odkaz: |
|
