CAREER: Mixed-Domain Pre-Distortion of RF Non-Linear Components for High-Efficiency, Wide-Band, and Reconfigurable Transmitters

About This Grant

This project aims to significantly advance wireless communications by developing AI-enhanced energy-efficient high-performance RF transmitters to address the growing energy consumption and efficiency challenges for next-generation wireless systems, which have significant environmental and economic impacts. Radio-frequency (RF) power amplifiers are responsible for 60% of the energy consumption in radio units, making them a critical area for improvement. The project introduces a novel mixed-domain pre-distortion (MPD) approach that combines AI/machine-learning (ML) algorithms and digital/analog/RF nonlinear components to create high-efficiency, wide-band, and reconfigurable transmitters. By reducing power consumption, improving bandwidth, and enhancing linearity, this research has the potential to transform the design of RF transmitters, enabling more sustainable and cost-effective wireless networks. The project also includes a robust educational and outreach plan, which will modernize RF/microwave courses, engage undergraduate and graduate students in hands-on research, and inspire high-school students to pursue STEM careers through practical design projects. The outcomes of this research will directly benefit U.S. industries such as telecommunications, defense, and energy, while contributing to the development of a globally competitive U.S. STEM workforce. The research of this project seeks to develop a unified theory and practical implementation of mixed-domain pre-distortion, leveraging the strengths of digital, analog, and RF nonlinearities. The research will involve creating a unified theoretical framework for MPD, designing and prototyping transmitter systems, and validating their performance through simulation and experimental testing. Advanced semiconductor technologies, such as GaN and GaAs, will be utilized to fabricate MPD prototypes, which are expected to achieve significant reductions in power consumption, wider bandwidths, and improved linearity compared to existing solutions. The project will utilize state-of-the-art facilities including advanced measurement setups and prototyping tools. The MPD approach aims to achieve a 70% reduction in power consumption compared to digital-only solutions while maintaining high linearity and enabling GHz-level bandwidth. The research will explore innovative nonlinear shapes, AI/ML algorithms, and circuit architectures, with direct applications in RF components such as power amplifiers, low-noise amplifiers, and converters. The findings will be shared with industry partners and the broader research community, contributing to the development of more efficient and scalable wireless communication systems. By addressing the efficiency-linearity-bandwidth tradeoff, this project will advance the state-of-the-art RF communication systems and provide transformative solutions for future wireless technologies, including 5G-Advanced and 6G networks. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.