Unmanned aerial vehicle with immunity to hijacking, jamming, and spoofing attacks
AssigneeDrobotics, LLC
InventorEyal Stein
Life of this patent
Estimating the term…
Abstract
An unmanned aerial vehicle (UAV) or “drone” executes a neural network to assist with detecting and responding to attacks. The neural network may monitor, in real time, the data stream from a plurality of onboard sensors during navigation and may communicate with a high-altitude pseudosatellite (“HAPS”) platform. For example, if the neural network detects a cyber-attack but determines that it does not interfere with external communications, it may shift navigation control of the drone to the HAPS.
Claims26
An unmanned aerial vehicle (UAV) comprising:
- a flight package;
- a UAV navigation system;
- a UAV image-acquisition device;
- a UAV communications array including a plurality of agile transceivers;
- a computer memory including a plurality of pre-stored waypoints and flight images of a pre programmed flight path;
- and
- a drone inspection neural network including a neural compute engine and a predictor,
- the neural compute engine including a processor and instructions, stored in the computer memory and executable by the processor, for using real time data received from one or more of the UAV image-acquisition device, the UAV communications array or the UAV navigation system as input to reconstruct a real time 3D map of the UAV operating environment and to position the UAV in real time in the 3D map;
- the neural network-based predictor computationally trained to:
- receive the real time 3D map and the real time UAV data and
- identify and classify (a) that an attack on the UAV is occurring, (b) characteristics and type of attack, and (c) if the attack is new and uncharacterized,
- where the neural compute engine is configured to:
- send attack data to a HAPS or cloud based neural compute engine if the attack is classified as uncharacterized,
- request updated predictor weight and configuration files identifying the characteristics and type of attack, and
- receive updated weight and configuration files with classifications for the type of attack. and mitigation instructions in real time.
A method for a UAV including a flight package, a UAV navigation system, a UAV image-acquisition device, a UAV communications array including a plurality of agile transceivers, a computer memory including a plurality of pre-stored waypoints and flight images of a pre programmed flight path, a drone inspection neural network including a neural compute engine and a predictor, the neural compute engine including a processor and instructions, stored in the computer memory and executable by the processor, and the neural network-based predictor, the method comprising:
- using, by the neural compute engine, real time data received from one or more of the UAV image-acquisition device, the UAV communications array or the UAV navigation system as input to reconstruct a real time 3D map of the UAV operating environment and to position the UAV in real time in the 3D map;
- receiving, using the neural network-based predictor, the real time 3D map and the real time UAV data;
- identifying and classifying, using the neural network-based predictor, (a) that an attack on the UAV is occurring, (b) characteristics and type of attack, and (c) if the attack is new and uncharacterized
- sending, using the neural compute engine, attack data to a HAPS or cloud based neural compute engine if the attack is classified as uncharacterized,
- requesting, using the neural compute engine, updated predictor weight and configuration files identifying the characteristics and type of attack, and
- receiving, using the neural compute engine, updated weight and configuration files with classifications for the type of attack, and mitigation instructions in real time.
Dependencies are inferred from the claim text.