Job
- Level
- Junior
- Job Field
- Data
- Employment Type
- Full Time
- Contract Type
- Internship / school internship
- Salary
- from 3.000 € Gross/Project
- Location
- Vienna
- Working Model
- Onsite
Job Summary
You will develop and validate innovative approaches to enhance GNSS state estimation and detect spoofing/jamming attacks using Spiking Neural Networks and the analysis of GNSS metadata.
Job Technologies
Your role in the team
- Initial Situation / Problem Statement: Modern GNSS receivers are based on Kalman filters for state estimation. However, conventional approaches often use static or heuristic covariance models, which lead to reduced accuracy and reliability in complex environments such as multipath or non-line-of-sight scenarios.
- Furthermore, GNSS systems are increasingly exposed to spoofing and jamming attacks, which can significantly impair or distort positioning accuracy.
- This work investigates a novel approach in which Spiking Neural Networks are used for adaptive determination of measurement covariance and for detecting anomalies in GNSS metadata that may indicate spoofing or jamming.
- The concept enables increased robustness, security, and energy efficiency and can be implemented as an external extension of existing GNSS receivers.
- The goal is to investigate and validate a novel approach to improve GNSS state estimation as well as to detect spoofing and jamming attacks using Spiking Neural Networks (SNNs).
- The focus is on the adaptive estimation of measurement covariance as well as anomaly detection based on GNSS metadata to enhance robustness, reliability, and security under challenging and adversarial conditions.
- Main tasks: Literature review on GNSS state estimation, spoofing/jamming detection, Kalman filtering, and SNNs (survey of the state of the art).
- Analysis of GNSS metadata (e.g., DOP, innovation, signal quality indicators) with regard to anomaly detection.
- Investigation of characteristic patterns of spoofing and jamming scenarios.
- Development of a concept for simulating GNSS attack scenarios and generating synthetic training data.
- Design of a simplified SNN model for adaptive covariance estimation and/or anomaly detection.
- Implementation of the simulation and processing pipeline in Python or MATLAB.
- Validation of detection capability and robustness enhancement compared to reference methods.
- Documentation of methods, assumptions, and results.
This text has been machine translated. Show original
Our expectations of you
Qualifications
- Strong interest in GNSS, signal processing, and machine learning.
- Basic knowledge of Kalman filtering and estimation theory.
- Knowledge in the field of neural networks (ideally SNNs or neuromorphic approaches).
- Interest in cybersecurity or signal integrity is an advantage.
- Communication skills, reliability, and flexibility.
Experience
- Experience in Python, MATLAB, or C/C++.
This text has been machine translated. Show original
What we offer
- Start: immediately.
- Duration: 3 – 6 months.
- The successful completion of the master's thesis will be rewarded with € 3,000.
This text has been machine translated. Show original
Benefits
Food & Drink
Work-Life-Integration
More net
Health, Fitness & Fun
Topics that you deal with on the job
Job Locations
This is your employer
Kapsch Aktiengesellschaft
Klagenfurt, Graz, Innsbruck, Dornbirn, Salzburg, Wien, Wien, Wien, Leonding
Kapsch is a highly successful technology company with global significance in the future markets of Intelligent Transportation Systems (ITS) and Information and Communication Technology (ICT).
Description
- Company Size
- 250+ Employees
- Founding year
- 1892
- Company Type
- Established Company
- Working Model
- Hybrid, Onsite
- Industry
- Internet, IT, Telecommunication
Master's thesis: Improving GNSS robustness and jamming detection
- Salary
- from 3.000 € Gross/Project
- Location
- Vienna
- Working Model
- Onsite
- Diversity
- Open for all genders
