Variable-target admission control for nonstationary handover traffic in LEO satellite networks

Handover management is of great importance in a low Earth orbit (LEO) satellite system. However, the blocking performance of handover traffic in an LEO satellite system is not easily formulated with an acceptable level of accuracy because of the nonstationary nature of handover traffic, which causes difficulty in quantitatively making optimal decisions for channel assignment to meet a given quality-of-service requirement. Even though many challenging schemes have been proposed to dynamically determine the number of guard channels, most either use heuristic methods or are based on simplifying assumptions, causing invalid decision results. As an alternative, we suggest a quantified method to minimize the fraction of the number of blocked calls out of the number of total calls under nonstationary handover traffic. We develop new mathematical formulations of those fraction and optimization models with efficient exact solution algorithms. Performance analysis shows that the proposed method improves the blocking probabilities of handovers and new calls; this improvement results from a highly adaptive and reactive characteristic to the fluctuating handover traffic condition.