2020.11.17
Overview
An RF fingerprint signal processing device which is suitable for executing a machine learning algorithm on a plurality of input signals includes a receiver-feature determination circuit and a classifying circuit. The receiver-feature determination circuit is configured to compute the plurality of input signals on a neural network. The classifying circuit is coupled with the receiver-feature determination circuit. The classifying circuit is configured to feedback information of a receiver-feature component to the receiver-feature determination circuit. The receiver-feature determination circuit decreases a receiver-feature weight of the neural network, which the receiver-feature weight is associated with the receiver-feature component, to use the receiver-feature weight which is decreased to compute an output of the neural network.Category
發明
Patented
17/099,796
US 11,985,508 B2
Filing Date
2020.11.17Expired Date
2042.07.20Notification
2022.10.24