Machine learning-based beam management algorithms for space communications.

The spectrum environment for the satellite communications industry is becoming congested, contested, and complex due to increasingly massive constellation deployments. Newly activated terminals can take a long time to find and establish a link with their desired satellite, resulting in significant downtime. Beamforming is a key method for producing high-data rates and efficient communication links. However, a drawback is that the beam management procedures incur latencies and can result in radio link failures. As such, the interns of this project will explore and design low-latency and efficient beam management techniques that leverage machine learning (ML). The intent is to produce models that could be deployed to satellites and terminals which would automatically and passively detect its counterpart and configure for optimal operation. The resulting research is expected be used by Qoherent to develop technologies that achieve efficient, low-latency, and high throughput space communications.

Faculty Supervisor:

Hatem Abou-Zeid

Student:

Partner:

Qoherent

Discipline:

Engineering

Sector:

Manufacturing

University:

University of Calgary

Program: