IN Brief:
- Anritsu EMEA and YOTASYS have combined RF analyser hardware with AI and GPU processing for spectrum monitoring.
- The platform supports signal classification, interference detection, anomaly identification, modulation recognition, and geopositioning workflows.
- RF monitoring is moving toward distributed, software-defined, and AI-assisted analysis as spectrum environments become denser.
Anritsu EMEA and YOTASYS have formed a collaboration to deliver AI-enabled spectrum monitoring systems for RF analysis, signal classification, interference detection, anomaly identification, and distributed monitoring.
The collaboration brings Anritsu spectrum analyser hardware together with YOTASYS software, AI, and GPU-based processing. YOTASYS is acting as system integrator, combining Anritsu RF hardware with the Y9827A Inceptron Spectrum Analyzer, the Y9220A Agentic Spectrum Analyzer Server, and its Cognitelligent RF and Microwave framework.
The Y9827A Inceptron Spectrum Analyzer is built around Anritsu analyser technology and enhanced with GPU-accelerated processing and AI software. The system supports wideband monitoring from 9kHz to 54GHz across commercial, satellite, and defence-related applications.
High-performance NVIDIA Jetson Orin CPU/GPU computing boards are integrated to support machine-learning analysis inside the measurement workflow. The system is designed to detect, analyse, and classify RF signals, while also supporting automated anomaly detection, modulation recognition, and geopositioning.
The Y9220A Agentic Spectrum Analyzer Server adds orchestration for distributed RF monitoring. Multiple analysers and monitoring tasks can be coordinated across different locations, spanning portable field systems and centralised monitoring environments.
Wider frequency coverage, signal routing, and system-level RF integration have been moving quickly across adjacent designs, including Qorvo’s wideband RF switches for multi-band radio systems. Spectrum monitoring sits on the other side of that same density problem: more signals, more bands, more operating modes, and less tolerance for slow manual interpretation.
Wireless environments are becoming increasingly crowded as 5G, private networks, satellite communications, industrial wireless systems, drones, defence electronics, and test infrastructure all operate across complex RF conditions. Manual monitoring and post-event analysis become less effective when signal activity changes rapidly or when anomalous behaviour needs to be identified while it is happening.
AI-assisted monitoring still depends on the quality of the RF capture. Sensitivity, dynamic range, frequency coverage, and front-end integrity determine the quality of the signal data entering the analysis chain. The processing layer then has to classify activity, detect patterns, flag unusual signals, and coordinate decisions across distributed measurement environments.
Defence, satellite, regulatory, and wireless network optimisation applications often require persistent monitoring over wide frequency ranges. Operators need to distinguish legitimate traffic, interference, unplanned emissions, and potentially hostile or anomalous signals, frequently across multiple sites. Automated classification and workflow orchestration can reduce the burden of reviewing large volumes of spectral data manually.
The collaboration also reflects a wider shift in test and measurement. Instruments are increasingly becoming nodes in software-defined measurement systems, where local processing, remote orchestration, AI models, and workflow automation sit alongside RF performance. In spectrum engineering, the analyser is moving from a measurement endpoint toward a connected intelligence layer within the RF monitoring chain.
