International Association
for Cryptologic Research

Digital predistortion (DPD) is a crucial signal processing technique employed to mitigate the nonlinear distortions introduced by power amplifiers (PAs) in wireless communication systems. These distortions lead to signal degradation, spectral regrowth, and reduced energy efficiency, which are particularly problematic in modern communication systems such as 5G and emerging 6G networks, where stringent linearity requirements coexist with the demand for higher data rates and power efficiency. Traditional DPD algorithms rely on mathematical models and optimization techniques that can struggle to adapt to the increasing complexity of modern communication environments, such as wideband signals, high carrier frequencies, and dynamic operational conditions. Machine learning (ML) offers a transformative approach to DPD by enabling adaptive, efficient, and robust solutions that outperform conventional methods in real-world scenarios.

This project aims to develop AI-optimized digital predistortion algorithms and architectures tailored for modern and future communication systems. The research will focus on designing robust, real-time, and computationally efficient ML-based DPD solutions that adapt to diverse PA characteristics and operational conditions.

Candidate description and eligibility

• A highly motivated candidate with an MEng (or M.Tech)/BEng (or B.Tech) degree or equivalent in electronics and/or electrical engineering, with a strong passion for VLSI for Machine Learning/AI, Communication and Signal Processing is sought herewith.
• Desirable: In addition to above qualifications, expertise and interest in FPGA/ASIC and EDA tools would be advantageous.

To apply please send your motivation letter, CV, and recommendation letters to M.T.Khan@hw.ac.uk. Feel free to reach if you have questions.