MORE | Spring 2021

Employing Deep Learning and GPS Outdoor Positioning for Vision-Aided mmWave Beam Prediction

Data icon, disabled. Four grey bars arranged like a vertical bar chart.
This project is aimed at utilizing machine learning frameworks to develop an optimized beam training model, capable of predicting the optimal direction of mmWave signal at a considerable success probability, and with no training overhead. This is achieved by using Vision and GPS positional datasets collected from a real-time measurement Testbed. The outcome from this project could be employed in future mmWave systems so as to maximize the system utility such as average data rate, at lower latency compared to traditional beam training algorithms.

Student researcher

Tawfik Mohammed Osman

Electrical engineering

Hometown: Tamale, Northern Region, Ghana

Graduation date: Fall 2020

FURI Totals

TotalStudents

0

FacultyMentors

0

OnlineSymposia

0

FocusAreas

0

FURIProjects

0

MOREProjects

0

KEENProjects

0

GCSPProjects

0