UAV & Drone Security
Project Overview
UAV Security Analytics creates an invisible firewall of defense that's both effective and safe. The DroneFox system gives the power of automatic detection, threat identification, and safe, surgical, one-click mitigation (with appropriate authorization). Ensure the operator of the drone in flight has no idea that you can see what they are seeing and provide protection from spying eyes.
Research through PolySec Lab helps to identify the various frequencies that drones use to identify drones in flight better.
HACKRF is a Software Defined Radio peripheral capable of transmission or reception of arbitrary radio signals from 10 MHz to 6 GHz. HACKRF is an open-source hardware platform designed to enable education, experimentation, and deployment of Software Defined Radio (SDR) technology. HACKRF makes cutting-edge SDR technology available to everyone.
The PolySec Lab is committed to using open-source software to identify, modify and develop a stronger radio signal that will help in the use of a device further away from its base, such as identifying a drone's frequency and increasing the operational distance from its base location.
A spectrum/signal analyzer measures the magnitude of an input signal versus frequency within the full frequency range of the instrument. The primary use is to measure the power of the spectrum of known and unknown signals. Given the challenge of characterizing the behavior of today's RF devices, it is necessary to understand how frequency, amplitude, and modulation parameters behave over short and long intervals of time.
The PolySec lab uses the funtionality of the Fieldfox spectrum analyzer to clean up a low-level signal so it could be a useful signal that people could find or to extend the range of the signal so that you can control an object further from the base station.
Explore the Research
Comparative Analysis of Experimental Methodology in RF-Based Drone Detection and Classification Datasets A Research Study on Dataset Evaluation for Drone RF Signal Classification Abstract: This paper presents a comparative analysis of four RF-based drone detection datasets: DroneRF, CardRF, DroneSET, and Drone-Remote-Controller-RF-Signal-Dataset. We evaluate the experimental methodologies, including the number and types of drones used, signal features extracted, the data collection environment, RF chamber utilization, noise mitigation techniques, and classification models employed. We also investigate the datasets' robustness concerning different RF channel conditions, variations in training and testing environments, and overall classification accuracy. Our study provides insights into the impact of these methodological differences on real-world drone detection and classification performance.
Performance Analysis of Classical Machine Learning and Deep Learning Methods for Drone Classification Using RF Signal Characteristics A Research Study on RF-Based Aerial Threat Detection with ML & DL Models Abstract: This paper evaluates classical machine learning (ML) and deep learning (DL) approaches to classify drone RF signals. We systematically compare classical algorithms—Support Vector Machines (SVM), Random Forest (RF), Gradient Boosting (XGBoost), and K-Nearest Neighbors (KNN)—with advanced deep learning architectures including Convolutional Neural Networks (CNNs), Long Short-Term Memory (LSTM), and Transformer models. Experiments are conducted using an RF dataset specifically collected under diverse environments to emulate real-world scenarios. Performance metrics such as accuracy, precision, recall, F1-score, and computational efficiency are reported and discussed. Results indicate the superior accuracy and robustness of DL methods, whereas classical ML algorithms are more computationally efficient, highlighting a clear trade-off valuable for practical deployments.
Cybersecurity Evaluation of Boston Dynamics SPOT: A Penetration Testing Approach A Research Study on Security Assessment of Autonomous Robotics Abstract: This paper presents a cybersecurity assessment of the Boston Dynamics SPOT robot using a black-box penetration testing methodology. Following structured information gathering, vulnerability analysis, and proof-of-concept (PoC) exploitation, several security risks are identified and scored using the CVSS v3 framework. Implications for physical safety and operational reliability are discussed, and mitigation strategies are proposed.