Research On High Energy Efficient Computation Offloading In Space Information Network

With the growing demand for computing power in emerging mobile services in recent years,resource-constrained terminal devices offload tasks to ground base stations for processing through multi-access edge computing technology,which can meet their requirements in terms of latency and energy consumption.But in remote areas(such as oceans,mountains,and polar regions)that lack base station coverage,the edge computing model based on terrestrial cellular networks is no longer applicable.The space information network is composed of various air,space and ground platforms,and low earth orbit(LEO)satellites and UAVs are important components of the space information network.The LEO satellite constellation has the characteristics of wider network coverage and less environmental dependence,and UAVs can be flexibly arranged and can establish a line-of-sight connection with the user.Therefore,deploying the MEC server on the LEO satellite/UAV in the space information network can provide computation offloading service for equipment nodes in remote areas.In this offloading mode,different offloading strategies have different effects on node energy consumption and benefits.For example,a large proportion of nodes offloading will lead to high transmission energy consumption and large processing burden on the MEC server.Therefore,this paper studies high energy-efficient computation offloading scheme in space information network.The specific research content is summarized as follows:1、Facing the Io T production operation scenario in remote areas,aiming at the problem of high transmission energy consumption caused by the offloading task of the ground node set to the LEO satellite MEC server,in order to reduce the energy consumption of the node set,and at the same time make reasonable use of the computing resources in the network,combined with the device-to-device communication proposes a D2D-assisted satelliteground bilateral cooperation partial computation offloading scheme.Considering that the task is limited by the delay and the load of the MEC server,the problem is modeled as minimizing the total energy consumption of the ground node set during the offloading process,using the node offloading strategy based on improved genetic algorithm(NOS-IGA)to solve the node set minimum energy consumption.The experimental results show that NOS-IGA has good convergence under different delay requirements and the number of task nodes,and effectively reduces the energy consumption of the ground node set during the offloading process.2、In addition to the energy consumption of node sets during computation offloading,this paper also studies the problem of node set offloading benefits according to the different requirements of different tasks on latency and energy consumption.The offloading benefit of nodes is composed of energy consumption benefit and time benefit.The weight coefficient represents the preference of nodes’ benefit on energy benefit and time benefit.A three-layer MEC computation offloading scheme of ground node-UAV-LEO satellite is proposed,and the tasks can be processed in three places: nodes,UAV and LEO satellite MEC server.The maximum benefit during node set offloading is solved by the UAV relay offloading strategy based on improved genetic algorithm(UOS-IGA).The experimental results show that the UOS-IGA can effectively improve the benefits of nodes set under different weight preferences.