NIST has designed deep learning detectors to accurately detect the presence of radar signals for commercial 3.5 GHz sensors. Known as Environmental Sensing Capability (ESC) systems, these sensors are responsible for detecting federal incumbent signals and triggering interference protection mechanisms. NIST is also creating software tools and digital waveforms that can be used by the industry and the regulator to test and certify commercial 3.5 GHz sensors. The digital waveforms are either simulated or derived from radar measurements conducted by NASCTN in the 3.5 GHz band. In addition, NIST has developed algorithms to strategically place the sensors along the coasts effectively and efficiently.
The NIST radar waveform generator software, RF dataset, and the baseline radar detectors are submitted as a use case for IEEE 1900.8 Standard. The use case describes a reference workflow for generating Radio Frequency Machine Learning (RFML) datasets. The IEEE 1900.8 standard aims to standardize the storage format of RFML datasets and the interfaces that connect stages of the RFML model training pipeline. It will also address use cases for RF signal detection, classification, and characterization as well as identification of RF emitters.
More details about ESC sensor detection and placement can be found in the references below:
RF Dataset of Incumbent Radar Signals in the 3.5 GHz CBRS Band (NIST JRES 2019)
Detection of Incumbent Radar in the 3.5 GHz CBRS Band (GlobalSIP 2018)
3.5 GHz ESC Sensor Test Apparatus Using Field-Measured Waveforms (WInnComm 2018)