In this brief, an efficient approach using extreme learning machine (ELM) is first proposed for the behavioral modeling of radio frequency power amplifiers (RF PAs). As a single-hidden layer feedforward neural network algorithm, ELM offers significant speed advantages over conventional neural network learning algorithms. Compared to the existing behavioral modeling based on ANN, the proposed method also requires minimal human intervention. A Class-E PA is taken as an example for comparing ELM against traditional neural network learning algorithm. The modeling results of ELM for AM/AM and IMD3 agree well with the simulation results, and the speed advantage of the proposed method has also been confirmed.

Additional Metadata
Keywords Behavioral modeling, Computer-aided design, Extreme learning machine, Nonlinearity, Radio frequency power amplifiers
Persistent URL dx.doi.org/10.1109/MWSYM.2017.8058626
Conference 2017 IEEE MTT-S International Microwave Symposium, IMS 2017
Citation
Zhang, C.-Y. (Cheng-Yu), Zhu, Y.-Y. (Yuan-Yuan), Cheng, Q.-F. (Qian-Fu), Fu, H.-P. (Hai-Peng), Ma, J.-G. (Jian-Guo), & Zhang, Q.J. (2017). Extreme learning machine for the behavioral modeling of RF power amplifiers. In IEEE MTT-S International Microwave Symposium Digest (pp. 558–561). doi:10.1109/MWSYM.2017.8058626