AI-Augmented L1/L2 Triggered Mobility: Enabling 6G-Native Predictive Mobility

3GPP has been actively exploring new approaches to enhance the performance of traditional handovers triggered by Radio Resource Control (RRC) signaling. Introduced in 3GPP Release 18, L1/L2 Triggered Mobility (LTM) enables mobility control through Medium Access Control (MAC) layer signaling based on physical layer measurements, effectively reducing signaling overhead and handover interruption time. Since 3GPP Release 19, AI/ML techniques have been investigated to further improve Layer 3 (L3) handover performance, with standardization expected in Release 20. Motivated by the promising applications of AI/ML in wireless communication, this article extends their use to address the challenges of timing advance (TA) acquisition during LTM uplink (UL) early synchronization. Two AI-augmented LTM schemes are proposed. The first is an early TA acquisition occasion prediction scheme, which enables the user equipment (UE) to perform early TA acquisition at the optimal time, thereby avoiding LTM failures caused by outdated TA values and reducing the need for candidate cells to continuously reserve resources. The second is a TA value prediction scheme, where the network directly predicts TA values for candidate cells, eliminating the need for the UE to perform early TA acquisition procedures. This further reduces service interruption prior to LTM cell switching, which otherwise occurs when the UE redirects its transmitter (Tx) to candidate cells for UL early synchronization. The proposed AI-augmented LTM schemes pave the way for 6G-native predictive mobility with enhanced efficiency and reliability.