OmniSIG™ Sensor

OmniSIG™ Sensor

Using DeepSig’s pioneering application of Artificial Intelligence (AI) to radio systems, the OmniSIG™ sensor provides a new class of RF sensing and awareness. Going beyond the capabilities of existing spectrum monitoring solutions, OmniSIG is able to detect and classify signals with low latency using features learned directly from spectrum recordings. OmniSIG also understands the spectrum environment to inform contextual analysis and decision-making. Compared to traditional approaches, OmniSIG provides higher sensitivity and accuracy and is more robust to harsh impairments and dynamic spectrum environments. It also requires fewer computational resources and a lower dynamic range, allowing for an exponential improvement in speed and device requirements. 

The OmniSIG software can be deployed and scaled on a wide range of target devices, from low-SWaP mobile and embedded systems to mobile personal computers to cloud and datacenter environments. Its web-based UI, open low-latency streaming interface, and control API provide seamless integration into customer systems and applications.



The OmniSIG sensor performs detection and classification of RF emissions across large spectrum bandwidths in milliseconds. This allows it to report anomalies, changes, or threats in near real-time. OmniSIG delivers accurate results in harsh and congested environments, detecting wide- and narrow-band signals over a large instantaneous bandwidth. 

Detection and recognition have been validated across many signal types, including:

  • Cellular and infrastructure signals such as GSM, LTE, CDMA2000, WCDMA, and WiMAX
  • ISM-band signals such as WiFi and BlueTooth
  • Mobile radio services such as P25, GMR, DMR
  • IoT signals such as LoRA
  • Commercial aircraft RADAR signals

It can also be extended to include other signals and protocols based on customer requirements and applications using the OmniSIG SDK. It is robust to intentional and unintentional interference and to other impairments, including those caused by receiver hardware.


The OmniSIG sensor is RF signal-detection software that integrates into customers’ own systems or third-party platforms. The OmniSIG software is flexible and customizable to various processing platforms and elements. It supports numerous standard and custom radio interfaces and can be deployed, scaled, and updated using containerization such as Docker. We can work with customers to identify and configure off-the-shelf hardware items that can host the software to accommodate existing radio receivers and compute platforms.

The OmniSIG software deploys on a general-purpose processor such as an x86 or ARM core. All software components can run on general-purpose processors or GPUs can be used to accelerate the internal signal processing and machine-learned neural networks. Both discrete (e.g., NVIDIA RTX, Quadro, or V100/A100) and integrated (e.g., NVIDIA Jetson TX2 or Xavier AGX) GPUs are supported.

The OmniSIG sensor supports streams of radio samples from a radio receiver interface and file-based playback. These interfaces include standardized packet formats such as VITA-49, drivers for leading radios such as NI’s UHD & Epiq’s Sidekiq, and standardized file formats such as SigMF & Midas Blue.  

The OmniSIG software includes a remote socket-based API which allows for remote control of the radio interface, inference settings, and output formats. The socket-based API is available over ZMQ for automated integration and a command-line interface. A WebSocket API is available which enables monitoring live inference results and controlling the radio & sensor from any device with a browser including mobile phones.

The OmniSIG sensor publishes signal detection & parameter estimates in near real-time to standard protocols using JSON which allows for autonomous edge sensing and edge sensing with cloud fusion. The supported output methods include:

  • SigMF-based files
  • ZeroMQ sockets
  • ElasticStack
  • Web page


Deep Learning RF Signal Processing DeepSig

DeepSig’s revolutionary approach to RF signal processing applies machine learning to time-series radio samples and channel measurements, allowing it to learn from data. By creating algorithms that learn signals and effects from the I/Q representation, DeepSig’s systems achieve better performance than traditional simplified analytic models or feature-based methods.

Our approach allows us to optimize the whole system for a set of performance requirements rather than optimizing individual components. This allows us to rapidly adapt to new signal and interference types, obtain computationally efficient inference engines, and quickly account for various effects and environments.


The OmniSIG sensor’s sensing techniques demonstrate an improvement of 4 to 10 dB higher sensitivity over existing state-of-the-art methods. In some instances, it can provide reductions of 10x or more in computational complexity and throughput. These reductions are achieved through:

  • Algorithmic efficiency
  • Increased parallelism
  • Reduced sample precision and dynamic range

The increased ability to adapt to hardware impairments can lower receiver efficiency requirements and cost. OmniSIG also produces accurate sensing results with reduced data requirements and dwell times, providing power reductions, reduced latency, and increased throughput on equivalent processors. Scanning and detection are also faster. Sensing can be parallelized, scaled to application needs, and deployed on everything from cloud compute clusters to mobile ARM processors.



Download the Datasheet

OmniSIG provides a new class of RF sensing using DeepSig’s pioneering application of Artificial Intelligence (AI) to radio systems.