Narayanan, Aravind Tharayil and Minati, Ludovico and Hagihara, Aran and Kobayashi, Jun and Shimura, Toshihiro and Kawano, Yoichi and Chakraborty, Parthojit and Bartels, Jim and Tokgöz, Korkut Kaan and Dosho, Shiro and Suzuki, Toshihide and Ito, Hiroyuki (2024) A neural network-based DPD coefficient determination for PA linearization in 5G and beyond-5G mmWave systems. IEICE Electronics Express, 21 (10). ISSN 1349-2543
Full text not available from this repository. (Request a copy)
Official URL: https://dx.doi.org/10.1587/elex.21.20240186
Abstract
This work presents a neural networkDPDformmWave RF-PAs. Differently from existing neural network-based DPDs, the neural network in the proposed DPD does not reside in the forward data path. Instead, it estimates the polynomial coefficients from the complex Fourier amplitudes of harmonics during a calibration sweep. It can compensate for PA nonlinearity under various operating conditions with lower hardware complexity compared to conventional DPDs. The proposed design is validated on a 28 GHzCMOSphased-array transceiver. In 256-QAM 5G-OFDMA-mode, the proposed neural network DPD achieved an improvement in EVM from -28:7 dB to -32:0 dB, while maintaining an ACLR of -33:4 dBc.
| Item Type: | Article |
|---|---|
| Additional Information: | Document type: Letter |
| Uncontrolled Keywords: | 5G; beamforming; DPD; NN; transceiver |
| Divisions: | Faculty of Engineering and Natural Sciences |
| Depositing User: | Korkut Kaan Tokgöz |
| Date Deposited: | 05 Jul 2024 15:26 |
| Last Modified: | 27 Sep 2024 10:11 |
| URI: | https://research.sabanciuniv.edu/id/eprint/49528 |
Actions (login required)
- View Item