Comparative analysis of machine learning methods for equalization

• Grant number: 21/01684-7
• Support Opportunities: Scholarships in Brazil - Scientific Initiation
• Effective date (Start): April 01, 2021
• Effective date (End): July 31, 2022
• Field of knowledge: Engineering - Electrical Engineering - Telecommunications
• Principal Investigator: Romis Ribeiro de Faissol Attux
• Grantee: Luan Lopes Fontes
• Host Institution: Faculdade de Engenharia Elétrica e de Computação (FEEC). Universidade Estadual de Campinas (UNICAMP). Campinas , SP, Brazil

Abstract

The problem of channel equalization is a central task in the context of communications, as well as in several practical domains related to signals and systems. Although the use of linear filters in this problem has an important historical / theoretical role, the constant search for performance improvement and the progress of hardware has led to many proposals and applications based on nonlinear filters, in special universal approximators. Having these facts in view, in this project, the student will perform a comparative study of neural networks in the context of equalization. This study will have, initially, a formative character, but will also encompass implementations and analyses in distinct scenarios. This will be important to familiarize the student with programming methods and technical concepts. Our aim is to form a young researcher in two fields of significant practical interest, thus allowing him to contribute in a consistent and original way to an interdisciplinary field.