Realtime Software Defined Self-Interference Cancellation Based on Machine Learning for In-Band Full Duplex Wireless Communications
| Autor: | Junhong Xu, Shaoen Wu, Hanqing Guo, Shangyue Zhu |
|---|---|
| Rok vydání: | 2018 |
| Předmět: |
Scheme (programming language)
business.industry Computer science Testbed Software-defined radio Machine learning computer.software_genre Software Single antenna interference cancellation Limit (music) Key (cryptography) Wireless Artificial intelligence business computer computer.programming_language |
| Zdroj: | ICNC |
| Popis: | The key enabling technology to in-band full duplex wireless is the self-interference cancellation. This paper proposes a realtime software defined digital self-interference cancellation based on machine learning. The contributions of this work include: (1) a digital self-cancellation scheme that approaches to the limit of the hardware, and (2) an open software defined cancellation framework that supports the realtime cancellation using machine learning algorithms with synchronized iterative training and prediction. This solution is implemented and evaluated on a software defined radio testbed. It results in performance of 50 dB in digital cancellation, which is close to the hardware limit and significantly outperformed available literature digital cancellation solutions. It also demonstrates stable performance across spectral band. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|