ADAPTIVE AND HIERARCHICAL NETWORK AUTHENTICATION FRAMEWORK

    公开(公告)号:US20230188982A1

    公开(公告)日:2023-06-15

    申请号:US17550841

    申请日:2021-12-14

    CPC classification number: H04W12/06 H04W12/79

    Abstract: A non-transitory computer-readable storage medium stores instructions to configure a base station for user equipment (UE) authentication in a wireless network and to cause the base station to perform an operation comprising decoding configuration signaling received from a PHY security function (PSF) of the wireless network. The configuration signaling includes a request for collection of a plurality of signal samples from the UE, the UE authenticated based on successful completion of a first authentication process. A response message is encoded for transmission to the PSF. The response message includes the plurality of UE signal samples. A trained machine learning model received from the PSF is decoded. The trained machine learning model associates the authenticated UE with an RF signature of the UE. The RF signature is based on the plurality of signal samples. A second authentication process of the UE is performed based on the trained model.

    PHYSICAL LAYER TECHNIQUES TO MITIGATE THE HANDOVER PROCESS VULNERABILITIES

    公开(公告)号:US20230095401A1

    公开(公告)日:2023-03-30

    申请号:US17483912

    申请日:2021-09-24

    Abstract: An apparatus and system to mitigate non-genuine handovers are described. The handovers include handovers based on fake measurements and handovers to malicious cells. To mitigate these, a mitigation procedure is initiated when excessive handovers are detected. Location information obtained from the UE, estimation of PHY layer properties by the serving and/or target cell, or AI modeling of the best serving cell at the UE location is used to determine whether the handover is valid. If not, the handover is canceled and the UE is stopped from initiating new handovers for a specified time, the UE may be instructed to perform re-authentication with the network, and/or the serving cell recommends to the network authentication entity to revoke the UE authentication. To ensure that the target cell is legitimate, an AI model is used to classify the target cell as known/unknown and the result sent to the network in NAS signaling.

Patent Agency Ranking