[2024] OpenWiFi-DPD: An Open-Source Digital Pre-distortion In-the-loop Wi-Fi Demonstrator

Introduction

The rapid advancements in wireless communication systems, particularly towards 6G and Wi-Fi 7, necessitate improved digital pre-distortion (DPD) techniques to handle increasing communication capacities and the demand for data accuracy. Our OpenDPD project offered a novel framework utilizing Deep Neural Networks (DNN) for DPD to correct non-linearity in wideband Power Amplifiers (PAs). This Master's thesis proposes integrating the OpenDPD framework and OpenWiFi into an Analog Device ADRV9364-Z7020 FPGA-RF platform, creating an open-source Wi-Fi demonstrator that directly incorporates digital pre-distortion into the signal processing chain.

Project Aims and Objectives

This thesis aims to develop and implement a comprehensive system that leverages the OpenDPD framework for DPD within a Wi-Fi demonstrator environment. The objectives include:

1. Integration: Embedding the OpenDPD framework into an FPGA-RF platform to establish an in-the-loop DPD system for Wi-Fi signals.
2. Enhancement: Utilizing the digital pre-distortion module developed by OpenDPD to improve the linearity and efficiency of the Wi-Fi transmission system.
3. Evaluation: Assessing the performance of the OpenWiFi-DPD system in terms of Adjacent Channel Power Ratio (ACPR), Error Vector Magnitude (EVM), and bandwidth efficiency on wideband signals.

Opportunity for Advanced Research

This project offers a unique opportunity to contribute to the cutting-edge field of wireless communications, paving the way for future research and potential Ph.D. projects focused on digital signal processing and integrated circuit design for advanced wireless technologies.

Assignment

The project will entail:

• Literature Review: A thorough examination of current DPD techniques, focusing on DNN-based approaches and their application in wireless communication.
• System Design and Development: Designing and developing the OpenWiFi-DPD system, incorporating the OpenDPD framework into an FPGA-RF platform using Verilog for digital integrated circuit design.
• Simulation and Performance Analysis: Testing the system with simulated Wi-Fi signals to evaluate improvements in signal linearity, efficiency, and overall performance metrics (ACPR, EVM).

Expected Outcomes

This project aims to deliver:

• An open-source Wi-Fi demonstrator that integrates a state-of-the-art DPD module showcases signal quality and transmission efficiency improvements.
• A comprehensive analysis of the system's performance, contributing valuable insights into DNN-based DPD techniques for wireless communications.

Requirements

Requirements

Ideal candidates for this project should have:

• A strong background in digital signal processing and wireless communications.
• Experience in digital integrated circuit design, preferably with FPGA platforms using Verilog.
• Proficiency in programming languages relevant to machine learning and DNN frameworks, such as Python and PyTorch.

Contact

Last modified: 2024-02-20