DQN-Based Conditional Handover Algorithm for Leo Satellites Networks

With the development of wireless communication networks, the network has expanded from terrestrial to low space. Non-terrestrial networks (NTN) will play an important role in realizing the ubiquitous connectivity scenario in the vision of 6G networks. Low Earth orbit (LEO) constellations are crucial components of future 6G non-terrestrial networks, enabling global coverage of unserved areas such as oceans and deserts. However, the high-speed mobility of LEO satellites poses challenges to mobility management, especially in managing handovers between satellites for users. Motivated by this fact, this paper designs a conditional handover (CHO) strategy based on deep Q-networks (DQN) to enhance the service continuity of users in LEO satellite networks, which comprehensively considers factors such as signal-to-noise ratio, elevation angle, and distance. The strategy can adaptively adjust handover strategies according to the dynamic changes in the network. Simulation results demonstrate that the proposed handover strategy significantly reduces handover failure rate, decreases interruption delays, and enhances link service time.