Key research themes
1. How can harmonic and waveform analysis improve the understanding and efficiency of power amplifier operation across power back-off regions?
This research theme focuses on extending classical harmonic and waveform analysis of power amplifiers (PAs) beyond the maximum output power condition, specifically to analyze behavior under power back-off. Considering that modern telecommunication signals exhibit high peak-to-average power ratios (PAPRs), analyzing PA behavior over varying power back-off levels becomes essential for improving efficiency and linearity. Theoretical analyses employ Fourier series expansions and simplified device models assuming constant transconductance and currents with clipping effects to generalize across bias classes (A, B, AB, C, etc.). The insights enable waveform engineering, harmonic shaping, and design optimization to minimize distortion and enhance PA performance.
2. What methods and design strategies optimize the efficiency and output power of broadband power amplifiers across wide frequency bands for advanced communication systems?
This theme examines the methodologies and circuit designs enabling broadband power amplifiers to deliver high output power and efficiency over wide frequency ranges, particularly targeting sub-6 GHz up to millimeter-wave bands for 5G, radar, and satellite communication. Techniques include leveraging device knee-profile modeling for power and efficiency contour prediction, advanced matching network synthesis such as multisection transformers, and integrated power combining using distributed networks. Emphasis is placed on utilizing computational tools (load-pull, source-pull) combined with empirical adjustments to optimize the complex trade-offs between gain, linearity, and efficiency while sustaining broadband operation.
3. How can advanced behavioral and piecewise modeling techniques improve the accuracy and efficiency of digital predistortion and linearization of RF power amplifiers with nonlinear amplitude-dependent characteristics?
This research area investigates improved behavioral models for RF power amplifiers with strong amplitude-dependent nonlinear distortion, including those employing advanced architectures like Doherty and load-modulated balanced PAs. Piecewise modeling approaches, including decomposed vector rotation (DVR) and mixture of experts (ME) models, are analyzed for their ability to capture complex nonlinearities and memory effects while providing smooth transitions across subregions of input amplitude. Optimization of model segmentation thresholds and parameters is crucial for computational efficiency and accuracy. The resulting models enable more effective digital predistortion to extend linearity, reduce adjacent channel interference, and maintain efficiency.