LEO Satellite Handover Using Advantage Actor-Critic Algorithm in Giant Constellation Network

Low Earth orbit (LEO) satellite network is a crucial component of the space-air-ground integrated network (SAGIN) architecture and nonterrestrial networks. It can provide global, real-time, and seamless broadband data services to many users, playing a critical role in emergencies such as disaster relief. This article proposes an advantage actor-critic (A2C) algorithm-based handover strategy for LEO satellite networks to address high handover frequency, data transmission blockage, and optimal access selection issues. The strategy considers five handover factors: available satellite bandwidth, elevation angle, signal reception strength, potential service time, and service quality. At first, the handover problem is modeled as a combinatorial optimization problem, subsequently transformed into a Markovian action-value function-based multiagent Markov decision process, and solved in the framework of the A2C algorithm. The results of simulations illustrate the superior performance of the proposed handover strategy compared to five handover strategies, namely, random handover, maximum number of free channels, maximum service time, multiagent deep Q-learning, and maximum instantaneous signal strength scheme, in terms of both convergence and practicality. The proposed strategy reduces the number of average handovers by up to 58% and improves signal stability, with enhancements reaching up to 39% . In addition, this strategy significantly lowers the probability of data transmission blockage and the number of handover failures (handover failure rate), effectively ensuring user communication quality.