An O-RAN-native xApp framework for jammer detection and geolocation, published in Sensors

The paper An O-RAN-Native xApp Framework for Jammer Detection and Geolocation (S. A. Ratti, A. Piccioni, F. Graziosi, D. Cassioli, A. Marotta) has been published in Sensors (MDPI), vol. 26, no. 21, article 6859. It is also the first journal paper for PhD student Steve Azarsh Ratti — congratulations.

The work proposes an O-RAN-native architecture in which coordinated xApps on the Near-Real-Time RAN Intelligent Controller detect radio jamming attacks, identify the affected sector and localize the jammer in real time, using only the standard E2 Key Performance Measurements already reported by the network and without dedicated direction-finding hardware.

The framework was evaluated in ns-O-RAN, with 30 Monte Carlo runs per configuration. It localizes the jammer with a median error of 9–24 m, four to ten times lower than centroid-based and path-loss baselines, and processes 2000 UEs in 0.89 s on a single core, within the 1 s control loop of the Near-RT RIC.

The dataset, the ns-O-RAN downlink jamming model and the extended FlexRIC dashboard are released as open source.

• Paper: Sensors (MDPI)
• DOI: 10.3390/s26216859
• Code and data: github.com/steveratti/oran-jammer-detection-geolocation-dataset