When high PAPR reduction meets CNN: A PRD framework
One of the most important factors limiting the performance of OFDM (Orthogonal Frequency Division Multiplexing) system is high PAPR (Peak to Average Power Ratio). Great efforts have been made in suppressing PAPR, but their implementation often requires pre-processing all input signals, leading to ex...
Guardado en:
| Autores principales: | , , , |
|---|---|
| Formato: | article |
| Lenguaje: | EN |
| Publicado: |
AIMS Press
2021
|
| Materias: | |
| Acceso en línea: | https://doaj.org/article/8ce80e5396bf407db3f4c27e3e0adf14 |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| Sumario: | One of the most important factors limiting the performance of OFDM (Orthogonal Frequency Division Multiplexing) system is high PAPR (Peak to Average Power Ratio). Great efforts have been made in suppressing PAPR, but their implementation often requires pre-processing all input signals, leading to excessive calculation overhead. When the transmission speed is high, much more time will be taken to process the input signal with the traditional methods, which will reduce the performance of the system. In this background, this paper firstly presents an algorithm, called PRD, to identify the high PAPR sequence without IFFT (Inverse Fast Fourier Transform) operations, in which a CNN (Convolutional Neural Network) for identifying PAPR sequences is trained first before applying further PAPR reduction schemes. Experimental results show that the proposed algorithm can identify the high PAPR sequences with 92.3% accuracy and reduce PAPR with extremely low calculations. |
|---|