Machine learning for passive RF signal processing

Qoherent makes data-driven software for autonomously recognizing wireless activity in a complex spectral environment. Qoherent provides datasets and models for passive sensing of the RF spectrum to identify and analyze signals such LTE, 5GNR, and more. Qoherent also provides design tools for the full workflow required to deploy machine learning for RF applications. Qoherent is not restricted to any toolchain, and can offer services using open source tools (e.g. GNU Radio, PyTorch) or proprietary ones (e.g. MATLAB).


Research Activities

  • Reviewing and consulting on prior-art
  • Replicating and advancing peer-reviewed research
  • Implementation of original research activities

Engineering Activities

  • Dataset creation
  • Model design and training
  • Model integration and deployment into existing software

Qoherent is currently building a development automation platform for creating intelligent radio solutions.

The platform, called the Wireless AI Sandbox will be available for private-beta in early 2022.