Task Offloading And Resource Allocation In Mobile Edge Computing-Enabled Satellite-Terrestrial Networks

With the integration of satellite networks and terrestrial networks,users in areas not covered by terrestrial networks or in disaster areas can get services through satellite-terrestrial networks.In mobile edge computing(MEC)-enabled satellite-terrestrial networks,MEC servers are deployed at low earth orbit(LEO)satellites,which can provide computing services to delay-sensitive and computation-intensive applications of ground equipment.Although the existing literature has studied MECenabled satellite-terrestrial networks,some problems remain.To solve these problems,this dissertation studies task offloading and resource allocation in MEC-enabled satellite-terrestrial networks.The main research content and innovations of this dissertation are summarized as follows:Aiming at the problem that the computing resources of ground equipment are not utilized effectively in MEC-enabled satellite-terrestrial networks,the device to device(D2D)assistance-based task offloading in MEC-enabled satellite-terrestrial networks is studied.Service equipment(SE)assists the LEO satellite in computing tasks from user equipment(UE)by D2D offloading.The utility maximization problem for all UE,SE,and the LEO satellite is modeled and decomposed into two sub-problems,i.e.,task offloading ratio optimization and task offloading decision.Aiming at the decomposed sub-problems,the analytical solution of the optimal task offloading ratio of D2D offloading and MEC offloading is obtained by the theoretical derivation,respectively,and a matching game-based task offloading decision algorithm is designed.A joint optimization scheme of task offloading ratio and task offloading decision is proposed.The simulation results demonstrate that the proposed scheme has better performance than the reference schemes.For example,when the number of UE changes,the utility increases by 11 to 30 percent.Aiming at the task offloading decision and computing resource allocation problem in the multi-satellite scenario in MEC-enabled satelliteterrestrial networks,the offloading decision and resource allocation based on the inter-satellite cooperation in MEC-enabled satellite-terrestrial networks is studied.Multiple neighboring satellites assist the local satellite in computing tasks from Internet of Things(IoT)devices within its coverage area through inter-satellite cooperation.The task completion delay minimization problem for all IoT devices is modeled and decomposed into two sub-problems,i.e.,task offloading decision and computing resource allocation.For the decomposed sub-problems,the Lagrange multiplier method is used to obtain the optimal computing resource allocation for the local satellite and neighboring satellites,respectively,and a grey wolf optimizer algorithm-based task offloading decision algorithm is designed.A joint optimization scheme of offloading decision and resource allocation is proposed.The simulation results show that the proposed scheme has better performance than the reference schemes.For example,when the number of IoT devices changes,the delay decreases by 4 to 23 percent.Aiming at the problem of insufficient computing resources of a single satellite in the presence of ground infrastructure in adjacent areas in MECenabled satellite-terrestrial networks,the offloading decision and resource allocation based on the satellite-terrestrial cooperation in MEC-enabled satellite-terrestrial networks is studied.A base station(BS)in an adjacent area assists a LEO satellite in computing tasks from UE through the satellite-terrestrial cooperation.The task completion delay minimization problem for all UE is modeled and decomposed into two sub-problems,i.e.,task offloading decision and computing resource allocation.For the decomposed sub-problems,the Lagrange multiplier method is used to obtain the optimal computing resource allocation of the LEO satellite and the BS,respectively,and a potential game-based task offloading decision algorithm is designed.A joint optimization scheme of offloading decision and resource allocation is proposed.The simulation results validate that the proposed scheme can obtain better performance compared with the reference schemes.For example,when the number of UE changes,the delay decreases by 7 percent.Aiming at the task offloading decision and computing resource allocation problem when the network state information is unknown in MEC-enabled satellite-terrestrial networks,the online learning-based offloading decision and resource allocation in MEC-enabled satelliteterrestrial networks is studied.Multiple IoT devices select the same LEO satellite from multiple LEO satellites for task offloading in each time period.The problem of minimizing the average sum task completion delay of all IoT devices over all time periods is modeled and decomposed into two sub-problems,i.e.,task offloading decision and computing resource allocation.For the decomposed sub-problems,a task offloading decision algorithm based on the device cooperation aided upper confidence bound algorithm is designed,and the Lagrange multiplier method is used to obtain the optimal computing resource allocation of the selected LEO satellite.A joint optimization scheme of offloading decision and resource allocation is proposed.The simulation results verify that the proposed scheme performs better than the reference schemes.For example,when the time period changes,the average sum task completion delay decreases by 6 to 7 percent.