Resource Management Based On Resource Mobility In Space Information Network

Space Information Network(SIN)consists of space platforms(synchronous satellites,low earth orbit,or medium earth orbit satellites,manned or unmanned aircraft,etc.,)and terrestrial networks,which acquires,transmits and processes spatial information in real time.Compared with the traditional terrestrial network,SIN has the characteristics of changeable but predictable topology,diverse but limited resources,and complex task execution process involves multiple resources.It is especially important that due to the high-speed mobility of the space platforms,the resources it carries are moving.On the one hand,this feature increases the complexity of task planning and resource scheduling in SIN.On the other hand,it also brings great challenges to the network's service quality assurance,which becomes an important factor affecting the network performance.Although resource mobility poses a challenge for resource management,it also provides opportunities for network performance improvement.However,when resource mobility is utilized through resource transfer,resource exchange,and resource aggregation,it is often at the expense of time to use resources originally belonging to other regions,thus it is difficult to guarantee the quality of service(QoS)requirements of tasks.In order to make full use of the resource mobility to improve network service capabilities under the condition of guaranteeing QoS requirements of tasks,this thesis studies the gain of resource mobility on task execution,and proposes a resource management method that can effectively utilize resource mobility to improve the network service capabilities.The main contents and results are listed as follows: 1.Aiming at the trade-off relationship between resource mobility utilization and QoS requirements of tasks in SINs,a metric quantifying the resource mobility gain for network service capability brought by resource mobility is analyzed.Specifically,the time-expanded resource relationship graph is firstly used to describe the moving behavior of multidimensional resources and the collaborative relationship between different resources.Then by jointly considering the number of disjoint resource combinations k,the length of moving time window L,and the parameter of resource independence n,we propose(7)k,n,L(8)degree of freedom on resource combination as a metric measuring performance gain resulted from resource mobility.Furthermore,the analysis of resource mobility is transformed into the problem of finding disjoint paths in the graph.Finally,the tradeoff relationship between the gain brought by resource mobility utilization and the payment of delay is revealed through simulation.2.Aiming at the difference of QoS requirements between common tasks and emergency tasks,a hybrid resource-oriented resource management method based on resource mobility is proposed to maximize the sum of task priority while guaranteeing QoS requirements of tasks.Specifically,in the common task planning phase,based on the parallel ant colony optimization model,a parallel scheduling method is designed for solving the conflict between different feasible resource combinations of common tasks.This algorithm performs multi-dimensional resource scheduling for multiple common tasks without conflicts.At the same time,the convergence speed is improved by constructing a feasible initial solution based on the priority before the start of the iteration and directly allocating the resource combination with the smallest number of slots when the ant fails in the middle of the search process.In the emergency tasks planning phase,the influence factors based on resource mobility and task priority is designed to quantify the impact of candidate resource combinations on existing resource management schemes.Then we select the least affected resource combination for the emergency task and reschedule the tasks been conflicted.At last,the superiority of the proposed algorithms in terms of sum priority and completion rate is verified by simulation.