CommRad RF: A dataset of communication radio signals for detection, identification and classification

In today’s world, data is one of the most valuable resources. It’s now easier than ever to collect millions of radio frequency (RF) signals, both well-known and unusual. With such a large amount of data, researchers can create advanced deep learning techniques. This dataset focuses on baseband signals from communication radios. It includes RF signals captured from 27 different radios, using various frequencies and recorded in different environments. Altogether, there are over 2,700 signals in this collection. The signals were captured using an RF receiver that automatically detects and records signals until they disappear. To organize these signals into a library, the receiver was connected to a laptop, which processed and saved the unique RF fingerprints from different devices.