Resource Management Technology With Time Windows In Space Information Networks

Space Information Network(SIN)consists of the space-based network,the air-based network,and the ground-based network,which is a multi-layer heterogeneous interworking and opening network architecture.It supports global seamless coverage and has the abilities of information acquisition,processing and distribution in real-time.Compared with tradi-tional terrestrial networks,SIN has the characteristics of long-distance transmission,wide coverage,fast response,and large capacity,which facilitate broad application prospects.It provides integrated information services for aerospace,aviation,navigation,military,civil and emergency applications.However,the limited and time-varying network resources,high dynamic network nodes,and a great variety of complex network missions in the SIN turn into the shackles of network service capacity improvement.With the applications of SIN in various fields rapidly spread out,network missions experience a growth spurt.The network nodes of the SIN have long launch period,high cost,difficult on-board load maintenance and online upgrade.It is therefore difficult to meet the increasing mission demands simply by increasing the network size and enhancing the capability of network nodes.It is urgent to carry out research on the efficient resource management of SIN,to achieve accurate match-ing of mission and resources,fundamentally improving the network service capabilities of SIN.Compared with terrestrial wireless resource management,the resource management of SIN is distinctly characterized by variable space and time.To be more specific,the high speed movement of network nodes of SIN in a large scale space-time results in the intermit-tent and fast time-varying network resources,therefore increasing the complexity of resource management of SIN.In order to circumvent this challenge,this dissertation models the inter-mittent network resources as time windows to study the dynamic management mechanism of multidimensional resources in SINs,realizing the rapid and accurate matching of tasks and resources under high dynamic conditions,and further comprehensively improving the quick response capacity and load capacity of SIN.The main concerns and contributions are as follows:1.We propose an area target scheduling strategy based on fixed time window model,to overcome the low resource utilization problem caused by the coupling relationship between the observation and transmission resources in actual scheduling,realizing the efficient and fast matching of resources and missions and further effectively improving the coverage capacity of area targets and the quick response capacity of SIN,mean-while revealing the tradeoff between the coverage ratio and system response time.In view of the wide coverage and urgency of observing tasks in the emergency scenarios of SINs,the joint allocation of observation and transmission resources in area target scheduling is modeled as the coverage maximization and response time minimization problem.The weighted sum of coverage rate and system response time is considered as the optimization objective by introducing weight factors,so that the introduced weight differentiation is reflected in the optimization objective,and the coverage rate and response time in area target scheduling are flexibly regulated,thus providing tech-nical support for the proposed strategy in practical applications.On the basis of this,we propose a fast scheduling algorithm of area target with performance guarantee,and prove its performance lower bound 1-e^-1times of the optimal value in theory,thereby providing theoretical guarantee for its efficiency of the proposed scheduling.2.We propose a resource dynamic control strategy basied on non-fixed time window model,to circumvent the challenges brought by the changes of network task proper-ties and resources,alleviating the impact of the existing task planning algorithm on the scheduling of common tasks when dealing with unexpected tasks and further solving the problem of dynamic transmission resource allocation to meet the requirements of multi-user satellites and multi-missions,therefore effectively improving the real-time or near real-time satellite services of SINs.In view of the differentiation of network missions and the time-varying characteristics of network resources in SINs,we take the relay satellite system as an example to study the tradeoff mechanism between the emergency mission response speed of relay satellites and the network mission plan-ning efficiency.In particular,we explore the multi-antenna transmission characteris-tics of SINs to devise a multi-antenna task planning method based on rolling horizon method,therefore realizing the dynamic and efficient matching between transmission resources and missions.The core idea of the proposed planning method is introduced as follows:We dynamically and asynchronously optimize the scheduling period of each transmitting antenna in the SIN to generate a large number of planning starting points in the multi-antenna network scenarios,therefore realizing both the real-time or near real-time updating of network resources and the dynamic programming of tasks.Compared with the traditional dynamic task planning methods,the proposed plan-ning method alleviates the huge signaling overhead in real-timely updating network resources,expecially when a large number of unexpected tasks pour in.3.We propose a joint management strategy of observation and transmission resources based on extended time window model to circumvent the difficulties from the large feasible region of observation resource allocation.It ensures the efficient matching of observation and transmission resources under the complex temporal and spatial temporal constraints,thus realizing the efficient collection and transmission of more imaging data in the SIN.Specifically,we take agile earth observation satellites as an example to study the joint allocation of observation and transmission resources un-der the extended time window model in SINs.The enhanced observation abilities of agility satellites extend the length of the observation windows,thereby leading to the uncernities in observation resource allocation and also exacerbating the complexity of temporal and spatial timing among transmission time windows.To this end,we model the studied problem as an Integer Linear Programming(ILP)to maximize the weighted sum of imaging missions,which accurately characterizes the complex spa-tiotemporal constraints in real systems.On the basis of this,we explore the hidden structure of ILP to propose a joint resource allocation strategy based on Semi-Definite Relaxation(SDR)method.To reduce computational complexity,we combine the ad-vantages of the proposed SDR method with that of genetic algorithm.The simulation results show that the proposed algorithm significantly improves the weighted sum of scheduling tasks.