Skip to main content

Everything in our world seems to be wireless today.  Cell phones, smart watches, power meters, you name it. Sorting through all of those signals can be hard, let alone determining what shouldn’t be there.

That’s where machine learning and DeepSIG come in.  Our OmniSIG sensors use advanced machine learning to tell you what’s in your environment.  Unlike traditional RF sensors, OmniSIG underlying technology allows it to output large amounts of data for a wide range of signal types. That data can then be analyzed alone or with additional machine learning to look for unusual patterns in what’s in your RF space.   Some examples of these unexpected signals (anomalies) could include:

  • New signals not seen before
  • Signals such as cell phones at unusual times of day
  • Signals using more bandwidth than normal
  • Increased activity of a particular signal type
  • Unusually strong or weak signals of a particular type
  • Drop-offs from normal activity levels
  • An entire signal type that suddenly stops

For instance,  the Virtualitics VIP tool, shows a cell phone present in the middle of the night could indicate someone is in or near your building that shouldn’t be.  Or higher than usual amount of LTE activity near a secure government facility may indicate a threat that needs to be assessed and dealt with accordingly.

DeepSIG’s OmniSIG Sensor can detect and classify these signals and more, out of the box, using any of a variety of readily available software defined radios, and it can be trained using OmniSIG SDK to classify other signals of interest.

OmniSIG is also capable of integrating with traditional cybersecurity and information management systems through ElasticSearch or syslog.  This allows it to become part of a more holistic view of your environment with easy-to-use dashboards like the one below, and take actions as necessary.  Easy integration into tools such as Kibana,  can provide convenient summaries of the real-time RF environment and when combined with the post-processing analytics, mountains of RF data can be filtered down to find hidden anomalies in the RF environment, saving significant time and cost for operators who are trying to find the RF needle in a haystack.

For information about using OmniSIG sensor networks to detect a wide variety of RF anomalies, contact

Leave a Reply