Advances in technology have led to innovation in UAVs, often called drones. As a result, the use of a drone, which can threaten civilians, has increased. In order to prevent this, a decentralized UAV detection system was devised. Detection model was established using three models, RNN, CNN, and Dense, with two features including MFCC and Mel-spectrogram. RNN showed the best result with 96% accuracy. As part of a defense system, it was designed decentralized using the ad-hoc network.

UAV detection, RNN, CNN, dense, ad-hoc network.