AI-Based RF Anomaly Detection for Counter-UAS

Preventing drone crashes by detecting failure precursors and initiating corrective actions. The system uses onboard sensors, AI/ML models, and flight control interfaces to monitor drone components like batteries, motors, airframes, and navigation systems for signs of impending failure. If precursors are detected, the system can warn the pilot, force a return home, or safely land the drone before failure. The AI/ML models can also predict failure risks based on historical data and generate synthetic flight scenarios for training and testing.

Source

System and method for detecting anomalies in electric aircraft flight performance using machine learning models to analyze sensor data from various flight phases, identifying discrepancies in electrical system operation, and transmitting alerts to remote devices for maintenance and operational optimization.

Source

Automated aircraft inspection system that validates captured data from unmanned aerial vehicles (UAVs) before analyzing it for anomalies. The system compares captured data against reference data to determine if it meets quality and accuracy standards. If the data is invalid, the system performs corrective actions such as re-capturing data or adjusting camera settings before proceeding with anomaly detection.

Source

Detecting and managing malfunctions in drone flight behavior using a digital twin and neural network analysis. The system compares the flight state of the drone in real time with a digital twin created from sensor data. If the state difference exceeds a threshold, possible sources of the malfunction are analyzed using a neural network trained from past flight data. This provides a more accurate and timely malfunction detection compared to relying solely on onboard sensors.

Source

A method for detecting anomalies in aircraft trajectories using machine learning techniques. The method involves vectorizing flight trajectories into sequences of enumerators, aligning multiple trajectories by shifting sequences, and detecting anomalies through unsupervised classification. The method can also determine trajectory characteristics, identify sub-chains, and align chains using algorithms such as Needleman-Wunsch or Smith-Waterman. Anomaly detection models can be trained using labeled data and can be based on neural networks or decision trees. The method can be implemented onboard or remotely and can communicate anomalies to ground stations or other aircraft.

Source

A method and system for automated, real-time monitoring of environmental parameters through drone-based surveillance. The system integrates thermal imaging, video analytics, and machine learning algorithms to detect anomalies in environmental conditions, such as temperature, humidity, and fire intensity. The system enables automated detection of fire parameters, including temperature, humidity, and fire intensity, from aerial platforms. The system also provides predictive analytics capabilities to anticipate fire spread, enabling proactive intervention. The system integrates with existing infrastructure, including existing drone fleets, to automate monitoring and control operations.

Source

A method for detecting and mitigating heading anomalies in unmanned aerial vehicles (UAVs) that enables precise and reliable navigation. The method employs advanced fusion of magnetic compass, RTK, and gyroscope data to accurately determine drone heading, with automatic intervention when anomalies are detected. The system employs a multi-stage approach: initial magnetic compass and RTK data validation, followed by comprehensive fusion of all three data sources to determine heading accuracy. If any of these sources indicate an anomaly, the system automatically intervenes with safety measures, including mode degradation and control weakening, to prevent continued flight.

Source

A high-altitude platform for safe navigation of unmanned aerial vehicles (UAVs) using machine learning. The platform, equipped with a neural network, monitors UAVs in its airspace and predicts potential flight hazards, such as collisions or communication disruptions. It can automatically adjust UAV flight plans to avoid hazards and optimize operations, while also enabling real-time communication and surveillance services. The platform can detect and classify anomalies, such as passive intermodulation (PIM) on telecommunications structures, and enable autonomous UAV inspection and detection of PIM sources.

Source

A system for detecting and classifying radio frequency signals from drones and other unmanned systems using machine learning. The system generates a custom neural network model through an end-to-end process of data collection, labeling, training, and deployment, enabling identification of various RF signals beyond traditional spectrum monitoring capabilities.

Source

A method for detecting faults in unmanned aerial vehicles (UAVs) that enables the detection of multiple types of failures. The method involves acquiring flight status information from the UAV, jumping to a corresponding fault detection model based on the information type, and performing fault detection operations according to the model. The method can detect faults in sensors, automatic control systems, batteries, and power systems, and can trigger protection operations based on the detected faults.

Source

A method, equipment, and storage medium for monitoring drones. The method involves obtaining parameter and abnormal information from drone components, wirelessly transmitting this information to a ground terminal, and processing the received data. The equipment includes a drone with a communication interface and a data storage module, while the storage medium contains a computer program that executes the monitoring method.

Source

Unmanned aerial vehicle (UAV) detection and mitigation using sensors and AI analysis. The system detects UAVs using sensors and analyzes the UAV characteristics and behavior to classify them and predict future actions. It determines a risk level and compliance category for each UAV, then prescribes corrective actions based on these factors.

Source

A method for inspecting unmanned aerial vehicles (UAVs) using an inspection drone that combines sensor-based inspection with failure modeling. The method identifies potential failure modes based on flight data and system behavior models, generates a detailed inspection plan that incorporates sensor data, and executes the plan using the inspection drone. The inspection drone's onboard sensor model provides real-time sensor data that is used to present failure modes to the inspection drone, enabling targeted inspection of critical components. The method leverages failure models to predict the likelihood of failure events, ensuring accurate and efficient inspections.

Source

A system for detecting aircraft faults using a generative adversarial network (GAN) that learns to generate simulated flight data and identify anomalous patterns. The system trains a GAN on a dataset of flight measurements, where the generator produces synthetic data and the discriminator evaluates its authenticity. The trained GAN identifies key features that capture the underlying data distribution, and an anomaly detection model is built using these features to predict fault states.

Source

Anomaly detection system for aerial inspection using self-supervised learning models. The system generates a data coding model representing a normal state of a target asset category, and uses this model to reconstruct a normal image from a target image. By comparing the original and reconstructed images, the system detects anomalies without requiring historical imagery or pre-defined anomaly models.

Source

Extending the range of detection for detecting and disrupting unauthorized drones using a network of distributed drone disruption units. The method involves monitoring drones and their controllers in multiple areas using local drone disruption units. If a drone is detected in a monitored area, nearby drone disruption units are notified. They then collaboratively disrupt the drone's communication link to force it to land. This leverages local resources to extend the effective range of drone detection and disruption beyond the capabilities of individual systems.

Source

Modular RF systems for locating, identifying, and tracking objects in 3D space using directional antennas. The systems have compact, movable enclosures with integrated cooling and thermal management. The enclosures house modules like RF, power, and processing. Directional antennas are attached to the exterior. The systems leverage machine learning to identify object types from captured RF signals. Based on the type, they generate customized signals and transmit them from the directional antennas. This allows targeted transmissions vs omnidirectional. Multiple systems can share custom signals. The modular design allows swapping antennas for different applications.

Source

UAVs equipped with neural networks to resist electronic attacks that try to hijack or interfere with their flight. The UAVs have onboard neural networks that analyze real-time sensor data to detect cyber attacks and respond appropriately. The networks analyze images, GPS, and communications to identify attacks like spoofing, jamming, or disorientation. They can then autonomously mitigate the attacks, like continuing the flight path or seeking better connectivity, without being compromised.

Source

A smart surveillance network for detecting and tracking non-cooperative and cooperative air and watercraft, enabling safe beyond-visual-line-of-sight (BVLOS) operations of unmanned aircraft systems (UAS) and unmanned surface vessels (USV). The system integrates radar, Automatic Dependent Surveillance-Broadcast (ADS-B), Automatic Identification System (AIS), and counter-UAS sensors to provide real-time situational awareness, automatic sensor performance assessment, and risk-based traffic analysis. It enables authorities to certify airspace or waterways for BVLOS operations, monitor compliance with regulations, and detect potential safety threats.

Source

Passive sensing of unmanned aircrafts (UAs) using forward scatter effect of radio frequency (RF) signals transmitted by satellites and/or aerostats. The system comprises passive RF sensor units that receive and process RF signals to detect and track UAs within a target volume. Multi-sensor analysis units and a central analysis unit process data from multiple sensor units to provide UA situational surveillance, including target detection, tracking, classification, and event prediction.

Source

A method for detecting micro-drones in urban environments using a combination of active and passive sensors. The method involves recording acoustic signals from the drone's propellers and optical signals from the drone's mechanical vibrations, and then correlating these signals to detect the drone. The active sensors include a millimeter-wave radar operating in the range of resonant absorption of radio waves in the atmosphere, and an active acoustic sensor operating in the range of resonant scattering of sound waves on the drone. The passive sensors include an optical sensor operating in the visible and infrared ranges. The method enables reliable detection of micro-drones in dense urban environments with high levels of interference.

Source

Detecting flying objects beyond visual line of sight (BVLOS) using computer vision on radar data to enable safe beyond-line-of-sight drone operation. The method involves converting radar buffer images into radial frames, subtracting a background, denoising, tracking objects, and identifying flying objects using CV techniques like contour detection. This allows detecting and avoiding other aircraft in BVLOS airspace by processing radar data like a radar screen would be seen by an operator.

Source

Unmanned aerial vehicles (UAVs) are becoming more widely used in various industries, raising growing concerns about their safety and reliability. The flight data of UAVs can directly reflect their flight health status; however, the rarity of abnormal flight data and the spatiotemporal characteristics of these data represent a significant challenge for constructing accurate and reliable anomaly detectors. To address this, this study proposes an anomaly detection framework that fully considers the temporal correlations and distribution characteristics of flight data. This framework first combines a one-dimensional convolutional neural network (1DCNN) with an autoencoder (AE) to establish a feature extraction model. This model leverages the feature extraction capabilities of the 1DCNN and the reconstruction capabilities of the AE to thoroughly extract the spatiotemporal features from UAV flight data. Then, to address the challenge of adaptive anomaly detection thresholds, this research proposes a nonlinear model of support vector data description (SVDD) utilizing a 0/1 soft-margin loss,... Read More

Source

A method and system for identifying aerial vehicles using passive electromagnetic scattering structures attached to rotors. The structures generate unique micro-Doppler signatures during rotation, which are analyzed using machine learning algorithms to identify the vehicle. The system enables reliable identification of small aerial vehicles without active transponders, providing a passive and secure solution for friend-or-foe identification and tracking in shared airspace.

Source

Integrated system for low-altitude drone response that enables rapid identification and mitigation of unauthorized drones. The system combines radar, camera, and communication capabilities to detect and analyze drone presence, determine threat classification, and initiate response actions. The system integrates with a central server to manage drone traffic, including automatic threat detection, tracking, and response capabilities.

Source

Advancements in unmanned aircraft systems (UASs) have led to a considerable increase in unlawful operations of these systems over critical infrastructures and public spaces. In this work, a system that is implemented utilizing software-defined radio and signals of opportunity (SOPs) and is based on the fusion of direction of arrival and passive radar sensor measurements is proposed to counter unlawful UAS/drone activities. Specifically, passive radar methodology applied on SOPs (data collection, disturbance cancellation, cross-ambiguity, and constant-false alarm rate detection functions), along with direction of arrival (DOA) measurements obtained via the multiple signal classification algorithm, are combined to detect and track the rogue UAS/target in an area under observation. A prototype implementation of the detection-and-tracking system, with the use of small embedded processing units and the robot operating system is developed and examined in numerous outdoor experiments to thoroughly evaluate its performance.

Source

Small Uncrewed Aerial Systems (sUAS), commonly referred to as drones, have become ubiquitous in many domains. Examples range from drones taking part in search-and-rescue operations to drones being used for delivering medical supplies or packages. As sUAS commonly exhibit safety-critical behavior, ensuring their safe operation has become a top priority. Thus, continuous and rigorous monitoring of sUAS at runtime is essential. However, sUAS generate vast amounts of data, for example, multi-variate time series which need to be analyzed to detect potential emerging issues. This poses a significant challenge, due to resource constraints imposed on the onboard computation capabilities of sUAS. To alleviate this problem, we introduce ADAM, a novel adaptive monitoring anomaly detection framework for sUAS. ADAM selectively monitors a subset of data streams, which serve as indicators of anomalous behavior. In the event of a raised alert, ADAM adjusts its monitoring strategy, enabling additional detectors and taking further mitigation actions. We evaluated the effectiveness of ADAM through simu... Read More

Source

Unmanned Aerial Vehicles (UAVs) gather data efficiently for power line inspection. Anomaly detection is essential for power infrastructure dependability and security. It proposes a Cloud-Enabled Isolation Forest (CEIF) method for UAV-based power line inspection. It improves the isolation forest algorithm's efficiency and scalability in cloud computing. It can process huge UAV inspection datasets by dispersing cloud computing. The technique, which effectively isolates anomalies, is applied to the cloud for fast power line inspection and anomaly identification. It describes the CEIF system's cloud service integration and distributed computing algorithm optimization. Real-world UAV-based power line inspection datasets show it can accurately detect abnormalities with low false-positive rates. It is scalable and robust for improving power infrastructure dependability and security. It allows cloud services to deploy real-world settings to implement different inspection scales.

Source

Unmanned aerial vehicles (UAVs) are predicted to be widely used in both military and civilian sectors in the coming years due to their high mobility, low cost, and enhancement of the line-of-sight (LoS) conditions in non-terrestrial networks.Nevertheless, this raises some security issues if they are manipulated to cause security threats in restricted locations or to breach user privacy.In order to detect malicious UAVs, radio frequency (RF)-based approaches are adopted to detect ambient RF signals, which can be accomplished with inexpensive RF sensors under both LoS and, in particular, non-line-of-sight (NLoS) conditions.In this paper, we propose a passive detection technique based on received signal strength (RSS), and derive analytical expressions on the detection and false alarm probabilities considering realistic airto-ground (A2G) channel conditions.A novel low-complexity suboptimal detector is also proposed and its performance is compared to the optimal detection.Monte Carlo simulations are used to confirm the accuracy of the derived expressions under the aforementioned channel... Read More

Source

Anomaly detection for flight data is crucial in maintaining the safety and stability of unmanned aerial vehicles (UAVs), making it a topic of significant research and attention. However, existing anomaly detection methods often ignore the random noise of UAV flight data and lack effective parameter selection, resulting in inadequate anomaly detection performance. Furthermore, current methods generally face the problem of insufficient feature extraction capability. In this paper, a spatio-temporal correlation based on one-dimensional convolutional neural network (1D CNN), bidirectional long short-term memory (BiLSTM), and attention mechanism (AM) hybrid neural network with residual filtering (STC-1D CBiAM-RF) data-driven multiple regression framework is proposed for anomaly detection and recovery of UAV flight data. First, a correlation analysis method is used for parameter selection to reduce the dependence on expert knowledge. Second, a multiple regression model fusing attention mechanism is designed. It utilizes 1D CNN-BiLSTM as a feature extractor, guided by the attention mechanis... Read More

Source

The use of Unmanned Aerial Vehicle (UAV) in various industries is increasing, which places higher requirements on the reliability of UAV. One of the ways to ensure the safety of UAV flights is by detecting anomalies in their flight. However, traditional UAV anomaly detection models have some shortcomings. First, they fail to integrate data from multiple sensors across time and frequency domains, hampering the anomaly detection model's ability to accurately assess the UAV's status. Second, they apply the same prediction error loss to all classes, which result in excessive false positives in some key classes. Finally, most of them used unimodal classification models to process data from multiple heterogeneous sensors, which makes it difficult for the models to extract targeted features. This paper proposes a UAV anomaly detection model based on sensor information fusion and hybrid multimodal neural network (IF-HMNN). Firstly, facilitated by the newly devised Multi-source Heterogeneous UAV Sensor Information Alignment algorithm (MHSIA), IF-HMNN can realize information fusion from multip... Read More

Source

In recent years, a surging number of unmanned aerial vehicles (UAVs) are pervasively utilized in many areas. However, the increasing number of UAVs may cause privacy and security issues such as voyeurism and espionage. It is critical for individuals or organizations to manage their behaviors and proactively prevent the misbehaved invasion of unauthorized UAVs through effective anomaly detection. The UAV anomaly detection framework needs to cope with complex signals in the noisy-prone environments and to function with very limited labeled samples. This paper proposes <sc xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">BisSiam</small> , a novel framework that is capable of identifying UAV presence, types and operation modes. <sc xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">BisSiam</small> converts UAVs signals to bispectrum as the input and exploits a siamese network based contrastive learning model to learn the vector encoding. A sampling mechanism is proposed for optimizing the sample size involved... Read More

Source

A cognitive electronic warfare (EW) system that learns to counter unknown radar threats in real-time without relying on pre-existing threat databases. The system uses machine learning to analyze radar waveforms, identify emitter characteristics, and determine threat intent. It then generates and optimizes electronic countermeasures (ECMs) based on observed threat behavior, continuously assessing their effectiveness and adapting the ECM strategy as needed. The system operates in a closed-loop fashion, enabling real-time countermeasures against agile and unknown radar threats.

Source

A system and method for detecting and locating unauthorized unmanned aerial systems (UAS) in restricted airspace around wildfires. The system, called a Mobile Emergency Perimeter System (MEPS), uses multilateration and mesh networking to quickly locate the source of radio frequency signals controlling the UAS. MEPS can be rapidly deployed in various terrain conditions and extreme environments to establish a mobile perimeter around wildfires, enabling authorities to identify and halt unauthorized UAS operations.

Source

An information processing apparatus for safe flight of a mobile device, such as a drone, calculates the reliability of self-position estimation based on environmental factors and sensor data. The apparatus determines the accuracy of self-position estimation using GPS signals, camera images, and LiDAR data, and outputs a reliability value to an external display. The reliability value is used to trigger warning displays and control adjustments when the estimated self-position accuracy falls below a predetermined threshold.

Source

Unmanned aerial vehicle (UAV) detection, tracking, and management system that utilizes a combination of radar, video, audio, and camera sensors to monitor airspace. The system employs advanced object detection algorithms to identify and track UAVs while maintaining accurate spatial relationships. It also enables real-time monitoring of UAVs through camera-based tracking, with automatic camera reorientation to maintain object center-of-view. The system maintains a comprehensive database of detected UAVs, including historical trajectories, to enhance object recognition and tracking capabilities.

Source

A method for recovering original radio signals from Doppler-shifted and time-shifted signals received at mobile sensing devices, such as satellites or drones, using machine learning. The method trains a network with labeled time series data representing RF signals and their characteristics, and uses the trained model to classify new signal sequences and recover the original signals. The approach enables computationally efficient transform-stripped signal recovery, emitter identification, and geolocation based on signals received at one or multiple sensing devices.

Source

System and method for disrupting radar threats that are unidentified, ambiguous, or not effectively countered by known countermeasures. The system employs a cognitive electronic warfare (CEW) system that analyzes the radar waveform, identifies potential countermeasures, and optimizes their parameters through machine intelligence and reinforcement learning. The CEW system can apply multiple countermeasures in combination, varying their parameters to achieve optimal disruption of the radar threat.

Source

With the rapid development of artificial intelligence, unmanned aerial vehicle (UAV) has been widely applied in multiple fields for their convenience, low cost, and powerful function. Compared with single-UAV data, the UAV fleet data has more rich features, which can provide greater data collection and rich data information support to achieve more accurate anomaly detection. However, the flight paths of different UAVs vary greatly, which brings many challenges to fleet-level data analysis. And it is difficult for UAV fleet data to get exception labels to accurately build an anomaly detection model. Based on the above problems, this paper presents an anomaly detection method based on multivariate state estimation (MSET) and density-based spatial clustering of applications with noise (DBSCAN) optimization (DBSCAN- MSET). Establishing a normal state representation model from historical UAV data can effectively detect whether the following flight is abnormal. The proposed method is validated against the actual UAV data, and the results show the validity of our method.

Source

A surveillance system for detecting and classifying off-the-shelf drones in a local area using RF signal detection and machine learning. The system analyzes physical layer features of received RF signals to identify drone signatures, which are matched against a known library of drone signatures. Multiple sensors can be used in combination to enhance detection reliability, with global decision-making performed at a fusion center or mesh network.

Source

An autonomous unmanned aerial vehicle (UAV) detecting system that uses a blimp equipped with cameras and sensors to monitor airspace and detect unauthorized UAVs. The system generates an image signature of detected UAVs and compares it to a database of authorized UAV signatures to determine unauthorized presence. The blimp can be controlled to maintain a desired altitude and position, and can generate an alarm when an unauthorized UAV is detected.

Source

Autonomous drone security for secure opening and closing of provider locations like stores. The system uses drones to monitor the location perimeter for threats, defects, or suspicious objects before staff enter or exit. Drones equipped with sensors fly predetermined routes around the building. They scan using ultra-wideband (UWB) and RFID to detect anomalies. If a threat like a package is found, the drone alerts staff. If defects like fallen branches are found, they are reported for repair. The drones can also be dispatched to gather additional data on potential threats like loiterers.

Source

Unmanned aerial vehicles (UAVs) are increasingly adopted to perform various military, civilian, and commercial tasks in recent years. To assure the reliability of UAVs during these tasks, anomaly detection plays an important role in today's UAV system. With the rapid development of AI hardware and algorithms, leveraging AI techniques has become a prevalent trend for UAV anomaly detection. While existing AI-enabled UAV anomaly detection schemes have been demonstrated to be promising, they also raise additional security concerns about the schemes themselves. In this paper, we perform a study to explore and analyze the potential vulnerabilities in state-of-the-art AI-enabled UAV anomaly detection designs. We first validate the existence of security vulnerability and then propose an iterative attack that can effectively exploit the vulnerability and bypass the anomaly detection. We demonstrate the effectiveness of our attack by evaluating it on a state-of-the-art UAV anomaly detection scheme, in which our attack is successfully launched without being detected. Based on the understanding ... Read More

Source

Radar system for detecting aircraft signatures like drones that provides robust detection in challenging environments like bad weather without needing costly multistatic radar arrays. The system has multiple radar clusters, each with co-located transmit and receive antennas. This confined structure improves resolution and coverage. The radar uses sequential or simultaneous signal broadcasting and reception. It detects motion with Doppler shifts and stationary objects with micro-Doppler analysis. Multiple chirp signals with ramps and frequencies avoid aliasing and improve resolution. Co-located antennas, sequential/simultaneous signals, and micro-Doppler analysis enable robust drone detection without multistatic arrays.

Source

A system for detecting UAVs and their operators from an aircraft, comprising an array of RF antennas that sequentially receive and compare the amplitudes of RF signals from the UAV, enabling the calculation of the UAV's direction of flight relative to the aircraft.

Source

A system for detecting and responding to unauthorized unmanned aerial vehicles (UAVs) in airspace. The system includes a surveillance drone that detects breaches of predetermined rules, such as no-fly zones, and communicates the breach to a control center. The control center can then instruct other UAVs to capture evidence of the unauthorized UAV or take action to disable it. The system enables safe and efficient management of UAV traffic in shared airspace.

Source

A method and apparatus for detecting and identifying unmanned aerial vehicles (UAVs) using radio wave measurement. The method generates a spectrogram, determines a region for direction finding, and finds the direction of the UAV based on signal values in that region. It further determines a region for UAV type identification and identifies the UAV type based on signal values in that region. The apparatus includes a processor that generates a spectrogram, determines regions for direction finding and type identification, and performs direction finding and type identification based on signal values in those regions.

Source

Recently, the number of drones has increased, and drones' illegal and malicious use has become prevalent. The dangerous and wasteful effects are substantial, and the probability of attacks is very high. Therefore, an anomaly detection and protection system are needed. This paper aims to design and implement an intelligent anomaly detection system for the security of unmanned aerial vehicles (UAVs)/drones. The proposed system is heavily based on utilizing ICMetric technology to exploit low-level device features for detection. This technology extracts the accelerometer and gyroscope sensors' bias to create a unique number known as the ICMetric number. Hence, ICMetric numbers represent additional features integrated into the dataset used to detect drones. This study performs the classification using a deep neural network (DNN). The experimental results prove that the proposed system achieves high levels of detection and performance metrics.

Source

Distributed airborne acoustic anti-drone intelligence system (DAAADS) that uses acoustic sensors on tethered balloons to detect, track, and predict trajectories of drones approaching a protected site. It alerts an air defense unit to intercept the drones, then tracks the debris to predict if any will hit the site. The system calculates drone speeds and uses regression to predict trajectories. It also uses machine learning to identify threats and debris.

Source

With the rapid development of unmanned aerial vehicle (UAV) technologies, UAVs are now increasingly leveraged to perform military and civilian tasks today. Meanwhile, as a complex cyber-physical system, UAVs are also facing security and reliability concerns raised by internal systems errors and external cyber-attacks from multiple aspects. Recent research has spent efforts on leveraging AI and machine learning techniques to predict the flying status of UAVs using their flight data for anomaly detection. However, these methods often ignore the prediction delay existing in status-changing periods during the UAVs operation, which inevitably causes false alarms and opens a window for malicious adversaries if they are not appropriately addressed. In this paper, we propose a new approach to enable effective anomaly detection and recovery for UAV flight data. Our approach adopts a hybrid design to eliminate false alarms during the status-changing periods while maintaining the high reliability of anomaly detection. We evaluate the proposed approach on flight data collected from multiple UAV... Read More

Source