Nexo Defence & Cybersecurity

Every emitter has its fingerprint

We classify radar emitters from their pulse parameters and help keep emitter libraries up to date.

ForDefence integrators and signals intelligence units

Request information See Defence & Cybersecurity →

The challenge

Why it matters

A receiver picks up thousands of pulses per second from different sources. Separating them and recognising which emitter is behind them, quickly and reliably, is a big-data problem.

How we solve it

Data, model, solution

01
Data
Pulse descriptor words (PDW): time of arrival (TOA), frequency, width, amplitude and direction of arrival (DOA). The pulse repetition interval (PRI) is derived from the arrival times.
02
Model
Clustering and classification with machine learning, with a confidence level.
03
Solution
Emitter identification and suggestions to update the library.

What's included

In detail

Machine learning
Signal processing
Clustering
Synthetic data

De-interleaving of pulse trains from several sources
Classification with an explainable confidence level
Management and versioning of emitter libraries
Training with realistic synthetic data

FAQ

Questions we often get

What is an emitter library?

A database with the characteristics of known radars that makes it possible to recognise them when detected.

What does AI add over fixed rules?

It adapts to emitters with variable parameters and to dense environments where manual rules fail, and it states how confident it is in each identification.

Request information