RF TX impairment minimization through the optimal use of configurable analog and digital equalization in RF DACs

RF power amplifiers (PAs) typically consume a significant proportion of the power budget in wireless communication systems. Consequently, substantial effort is devoted to maximizing their efficiency while ensuring compliance with stringent system performance requirements. A fundamental tradeoff exists between amplifier efficiency and linearity, a challenge that is further exacerbated by the use of efficiency-enhancement techniques such as envelope tracking, Chireix outphasing, and Doherty architectures.

Digital-to-analog converters (DACs) are traditionally optimized for maximum linearity, with their architectures and calibration schemes designed to maximize metrics such as signal-to-noise-and-distortion ratio (SNDR). Compensation of transmitter nonlinearity is then typically performed in the digital domain using techniques such as digital predistortion (DPD). However, these approaches are computationally intensive, consume significant power, and do not scale efficiently to very large MIMO arrays.  This is especially problematic in future 6G applications or NTN satellite architectures which explore very large arrays to increase antenna gain.

The goal of this PhD project is to investigate a joint analog-digital linearization approach in which a controlled nonlinearity is intentionally introduced at the front end of the DAC to approximate the inverse transfer characteristic of the transmitter chain. The candidate will explore the optimal partitioning of correction functions between the analog and digital domains, evaluating the associated tradeoffs in complexity, power consumption, and performance. The research will further examine how these techniques can be adapted to different RF front-end architectures and signal bandwidths, with the objective of enabling more power-efficient transmitter implementations for future wireless systems.