CBL - Campus del Baix Llobregat

Projecte llegit

Títol: GNSS Jamming detection using ADS-B data

Estudiants que han llegit aquest projecte:

Director/a: DE LA TORRE SANGRÀ, DAVID

Departament: FIS

Títol: GNSS Jamming detection using ADS-B data

Data inici oferta: 16-06-2025     Data finalització oferta: 16-02-2026


Estudis d'assignació del projecte:
    GR ENG SIST AEROESP
Tipus: Individual
 
Lloc de realització: EETAC
 
Paraules clau:
GNSS jamming, ADS-B, Mode-S, Integrity, Accuracy
 
Descripció del contingut i pla d'activitats:
Ground-based radionavigation aids are reliable, but their associated costs due to maintenance and limited coverage have led the aviation industry to seek alternatives.
Nowadays international organizations such as ICAO and the FAA are promoting satellite navigation as the primary means of navigation, not only for en-route segments but also for approach procedures. Performance-Based Navigation (PBN) enables more direct routing and efficient use of airspace, resulting in reduced fuel consumption, lower emissions, and decreased congestion.
However, satellite navigation is vulnerable to GNSS jamming and spoofing, which can significantly degrade system performance and even compromise safety during critical phases, such as the approach.
Automatic Dependent Surveillance Broadcast (ADSB) messages are automatically transmitted through an aircraft's Mode S transponder. These messages contain key indicators closely related to the quality of the satellite signal received, as they quantify the aircraft's protection levels when flying under PBN.
Motivated by this context, the student will design an algorithm capable of detecting and mapping GNSS jamming and/or spoofing events, based on the exploitation of information contained in ADS-B messages.
The core of this project is to create and validate an algorithm that fulfils this purpose. Depending on the student's time and motivation, a complementary application or web interface could also be developed to visualize the results.
 
Overview (resum en anglès):
Global Navigation Satellite System (GNSS) radiofrequency interference (RFI) poses a growing threat to civil aviation safety. This project proposes a method to detect and characterise GNSS RFI areas using Automatic Dependent Surveillance-Broadcast (ADS-B) messages.

The main objective is to overcome the limitations of existing monitoring platforms, which aggregate quality indicators (mostly NIC) over a global worldwide grid, an approach that is susceptible to false positives and requires a high data density. Additionally, this work aims to characterise the behaviour of the Airborne Velocity message and its NACv indicator under interference, an aspect not previously addressed in the scientific literature, and to confirm if the Position message with Type Code Zero is transmitted when no valid GNSS solution is available due to interference.

Methodologically, a software tool has been developed in Java 21, composed of a custom ADS-B decoder, an ingestion module that stores the relevant parameters of each message in PostgreSQL, and a detection module that analyses, for each individual flight, the temporal evolution of the NIC, NACp, SIL and NACv indicators using sliding-window algorithms to identify downgrades and recoveries relative to previous values. This analysis is coupled with a Position Gap detector algorithm, detecting GNSS signal loss from coverage loss by correlating the absence of position messages with the continued transmission of operational status messages. Data was obtained from the OpenSky Network for two datasets: one over the Iberian Peninsula, used as a baseline, and another over the Eastern Mediterranean, a known jamming hotspot. Results were analysed through Python scripts, generating bar charts of anomaly combinations, temporal multiplots and KDE heatmaps.

The results confirm the degradation-recovery pattern of quality indicators when crossing RFI areas and show that isolated Position Gaps are not reliable stand-alone indicators, but become highly significant when correlated with other anomalies. The software proves that the correlation between anomalies is the most reliable way of detecting GNSS RFI, rather than focusing in one indicator. Thanks to this correlation, the software is capable of filtering possible false positives, and does not have any limitation towards the density of the data, as the algorithms are applied per flight track.

The Airborne Velocity message behaviour is practically identical to the the Position message, once the aircraft enters into a high RFI area, the ADS-B system of the aircraft ceases to transmit both types of messages at virtually the same timestamp.
NACv indicator was found to behave slightly different from the rest, remaining consistently low rather than following the same sudden degradation pattern. The transmission of Type Code Zero messages could not be confirmed, likely due to prior filtering by the data provider.

© CBLTIC Campus del Baix Llobregat - UPC