US 12,444,241 B2
Anomaly detection model using message id sequence on unmanned moving objects
Huy Kang Kim, Seoul (KR); Jeong Do Yoo, Seoul (KR); Seonghoon Jeong, Seoul (KR); Eunji Park, Seoul (KR); Kang Uk Seo, Seoul (KR); and Minsoo Ryu, Seoul (KR)
Assigned to Korea University Research and Business Foundation, Seoul (KR)
Filed by Korea University Research and Business Foundation, Seoul (KR)
Filed on Nov. 22, 2021, as Appl. No. 17/532,272.
Claims priority of application No. 10-2020-0158745 (KR), filed on Nov. 24, 2020; and application No. 10-2021-0007599 (KR), filed on Jan. 19, 2021.
Prior Publication US 2022/0165101 A1, May 26, 2022
Int. Cl. G07C 5/00 (2006.01); G06N 3/08 (2023.01); G07C 5/08 (2006.01)
CPC G07C 5/008 (2013.01) [G06N 3/08 (2013.01); G07C 5/0808 (2013.01)] 19 Claims
OG exemplary drawing
 
1. A processor-implemented method of detecting anomaly of an unmanned moving object, the method comprising:
collecting packet data generated in the unmanned moving object, the packet data being transceived via a communication protocol from the unmanned moving object;
pre-processing the collected packet data;
detecting an anomaly of the unmanned moving object based on a message ID pattern of the packet data by inputting the pre-processed packet data to a pre-trained neural network model, the pre-trained neural network model being pre-trained and configured to predict a message ID following a current message ID sequence of the packet data input to the pre-trained neural network model; and
performing pre-defined missions, including at least one of a courier service, driving, surveillance, and military-related tasks, by controlling operations of the unmanned moving object based on a result of the detecting of the anomaly,
wherein the pre-training of the neural network model includes:
collecting normal packet data generated in a normal communication state of the unmanned moving object;
pre-processing the collected normal packet data; and
training the neural network model in a way of predicting a message ID following a current message ID sequence of the normal packet data by inputting the pre-processed normal packet data to the neural network model.