ISAC-Enabled Handover Design in LEO Satellite Networks

Mega-constellations of low Earth orbit (LEO) satellites are envisioned to deliver global broadband and direct-to-cell services, requiring seamless handovers (HOs) to ensure uninterrupted connectivity. Conventional break-before-make HO protocols, although supported by inter-satellite links, suffer from beamforming delays due to the high orbital velocities of LEO satellites. To address these limitations, we propose an integrated sensing and communication (ISAC)-assisted HO (ISAC-HO) protocol that enables make-before-break HOs via ISAC-enabled ground terminals (GTs). A novel ISAC design capable of generating a three-dimensional (3D) beampattern under realistic conditions allows GTs to sense approaching LEO satellites without significantly compromising communication with the currently serving LEO satellite. The design problem is highly challenging due to its non-convexity and mixed-integer nature. To tackle these challenges, we propose an approach that leverages Riemannian manifold optimization and closed-form solutions. For multi-GT scenarios, we introduce a multi-agent deep reinforcement learning framework that mitigates sensing collisions and ensures quality-of-service under shared spectrum constraints. Numerical results confirm that the proposed ISAC design significantly improves the communication–sensing trade-off, enables smooth HO, and remains robust in the presence of imperfect channel state information.