<p>The slice admission control is used to ensure efficient allocation of resources in the space-air-ground integrated network (SAGIN). The heterogeneity and resource constraints present significant challenges in meeting the increasing diversity of service requests. This paper proposes a slice admission control algorithm named Pre-Authorization Considering Priority and Fairness (PACPF) to address the challenge. Specifically, this paper introduces a network slice model that redefines priority and fairness representation to accommodate heterogeneous service goals in SAGIN. Then, the pre-authorization algorithm is proposed to maximize service level and resource utilization for network slicing providers while ensuring service priority and fairness through comprehensive multi-factor consideration. The reward function, prioritized experience replay, and adaptive noise mechanisms are designed to address the challenges of resource constraints and environmental complexities in SAGIN. Finally, the simulation results show that PACPF improves the service level by 19% and the resource utilization by 6% on average.</p>

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

PACPF: a network slicing admission control approach in space-air-ground integrated network

  • Hongxia Zhang,
  • Xiangxu Zhao,
  • Shiyu Xi,
  • Gao Ning,
  • Peiying Zhang

摘要

The slice admission control is used to ensure efficient allocation of resources in the space-air-ground integrated network (SAGIN). The heterogeneity and resource constraints present significant challenges in meeting the increasing diversity of service requests. This paper proposes a slice admission control algorithm named Pre-Authorization Considering Priority and Fairness (PACPF) to address the challenge. Specifically, this paper introduces a network slice model that redefines priority and fairness representation to accommodate heterogeneous service goals in SAGIN. Then, the pre-authorization algorithm is proposed to maximize service level and resource utilization for network slicing providers while ensuring service priority and fairness through comprehensive multi-factor consideration. The reward function, prioritized experience replay, and adaptive noise mechanisms are designed to address the challenges of resource constraints and environmental complexities in SAGIN. Finally, the simulation results show that PACPF improves the service level by 19% and the resource utilization by 6% on average.