An Intelligent Relay Node Selection Scheme in Space-Air-Ground Integrated Networks

Abstract

The Space-air-ground integrated network (SAGIN) has been a valuable architecture for communication support due to its characteristics of wide coverage and low transmission delay. Both low earth orbit (LEO) satellites and UAVs can serve as relay nodes to provide reliable communication services for ground devices. However, the design of relay node selection scheme in SAGIN is not easy, considering different accessing layers and resource usage of network segments. Moreover, network topology, available resources and relative motion need to be analyzed comprehensively. To address these problems, a traffic prediction method based on autoregressive moving average (ARMA) model is utilized firstly to forecast the resource usage of SAGIN segments. After the analysis of link performance, the Adaboost algorithm is used to classify network nodes for optimal relay node selection. Simulation results show that the proposed intelligent relay node selection scheme is feasible and effective.