MACHINE-LEARNING BASED RADIOFREQUENCY IMPAIRMENT CLASSIFIERS FOR SATELLITE COMMUNICATIONS

Competitiveness & Growth

Status

Ongoing
Status date

2023-10-15
Activity Code

6B.126

Objectives

The objective of this project is to create a solution for detecting RF impairments with machine learning, integrated into a passive RF receiver system, and achieve TRL7 for this scope.

Challenges

Key challenges include the design and development of suitable datasets for the project, selecting appropriate algorithmic solutions, then building a testbed for classifying and quantifying impairments in real-time. Technical challenges include handling of large datasets, automated training, orchestrating multiple training targets, and implementing the model in an inference solution that is suitable for a range of commercially available software radio technologies.

Benefits

The solution that is developed in this project will demonstrate that a learned methods can be produced in short development cycles with the appropriate tooling in place. Such tooling, leveraging Qoherent’s intelligent radio design suite, can help produce solutions that perform comparably to traditional analytical methods.

Features

The ability to classify and quantify a range of impairments, backed by solutions for creating datasets, training models, testing them, then deploying them into an inference solution.

System Architecture

The project is built and demonstrated with software radio testbed that consists of a broadcast transmitter, an channel emulation transmitter, and a receiver which is integrated with AI inference hardware. Inference solutions are deployed within a docker container. The overall system is controlled with a management application.

Plan

The project follows 5 core steps: (1) Project tooling such as development of signal generation and synthesis software (2) Recording capture and synthesis, then curation into datasets (3) Model training. (4) Model training and testing in software environments. (5) Model training and testing in an over-the-air capture environment.

Current status

Qoherent is currently part-way through the first step of the project, primarily focused on developing appropriate solutions for signal capture via the testbed and signal synthesis using our dataset synthesis products.

Ongoing