Open Projects for Master Students

This project aims to develop a 12 nm ASIC hardware accelerator for real-time audio generation using advanced generative AI models like WaveNet or diffusion models. By integrating customized arithmetic units and optimizing AI algorithms for low-latency, low-power audio processing, this project will enhance applications in audio GenAI. The deliverable will be a prototype chip demonstrating fast high-quality audio synthesis and processing capabilities.

Preferred Skills: Verilog, Python

This project aims to harness the potential of the OpenDPD framework together with OpenWiFi to build an open-source Wi-Fi demonstrator on an Analog Device ADRV9364-Z7020 FPGA-RF platform. Integrating a digital pre-distortion module developed through advanced machine learning techniques, this project will tackle the power amplifier (PA) non-linearity correction challenge in a practical Wi-Fi system—a critical aspect for next-generation wireless systems. Through the practical application of deep neural networks (DNNs) in digital pre-distortion, the project will elevate the efficiency and reliability of Wi-Fi systems. Check out this demo: https://youtu.be/9hzc7aHR-Bw?si=eSF30oqaawlwgDE7

Preferred Skills: Verilog/VHDL, Embedded C/C++ Programming

Deliverable: Your homemade Wi-Fi router!

This project aims to innovate bio-inspired energy-efficient deep neural networks for 6G radio frequency (RF) signal processing tasks, namely crest factor reduction and digital predistortion (DPD); the project seeks to enhance signal efficiency, reduce power consumption, and improve overall transmission quality in the upcoming 6G networks.

Preferred Skills: Python; it would be helpful if you also know MATLAB.