Qoherent
Toronto, Ontario
Qoherent develops software and engineering tools for applying machine learning to radio signals and software-defined communications. Its work connects the practical stages of an RF application: generating or recording signal data, organizing datasets, training and evaluating models, and integrating inference into radio equipment. RIA Toolkit supplies Python libraries for this workflow; RIA Hub adds a collaborative environment for datasets, code, packages and experiment development. RIA RAN applies that approach to open 5G radio access networks. The documented ESA project places Qoherent's inference engine alongside the distributed-unit processor in a modified srsRAN architecture, with USRP radios, ARM or x86 computing and an open5GS core. The objective is to adapt radio-resource assignment using local inference. This is an integration component within a wider radio network, whose behaviour also depends on the scheduler, radio hardware, core network and deployment configuration. The company also participates in NATO DIANA's 2026 cohort with an adaptive 5G payload for mobile ad hoc networking. That programme describes a related development direction; it does not establish that the DIANA payload and the ESA non-terrestrial-network configuration are the same deliverable.
What they offer
Explore Radio Intelligence Applications for 5G RAN