Mission-Oriented Resource Management Technology In Space Information Networks

Space information network(SIN)utilizes space segment as its carrier,such as geosynchronous orbit satellites,medium-orbiting satellites or low-orbiting satellites,and lift-off platforms,etc.At the same time,the control ground segment,such as network control subsystems,ground stations and mission management centers,plays the control and management role for SIN.Additionally,SIN is one of the top 100 national strategic projects in the “13th Five-Year Plan”.Compared with traditional terrestrial wireless networks,SINs have significant advantages such as global coverage,flexible networking,and long-distance transmission,which facilitate their applications.Specifically,SINs can provide integrated services and support for navigation and positioning,emergency response,space measurement and control,and intelligent transportation.As a result,SIN has gradually become a national strategy.Compared with resource characteristics in traditional terrestrial networks,network resources in SINs are characterized by discrete distribution,heterogeneity,dynamics,limited resources,and weak coordination capabilities.In addition,the resources in the SIN are diverse,and the conflicts between resources are complex and time-varying.At the same time,the missions in the SIN are complex and diverse,and different mission types have different characteristics.For example,the communication mission has the characteristics of small delay and high reliability,while the observation mission has the characteristics of large data volume and observation frequency requirement.Each mission in an SIN often requires multiple resources to be collaboratively completed.Different missions have different resource requirements.For example,an observation mission requires imager resources and a communication mission does not.Therefore,in the mission-oriented network resource management process,it is necessary to satisfy the requirements of missions to resources and the conflict-free resources scheduling.The variety of resources,the diversity of requirements for different missions,the network resources and the time-varying conflicts between resources pose great challenges to the resource management in SINs.To alleviate the contradiction between resource shortage and the increasing mission demand,how to design efficient resource management strategies is essential for the multi-dimensional resources characteristics and the various mission characteristics and requirements in SINs.To this end,this dissertation studies the multi-dimensional resource joint management technology in SINs for dealing with the requirements of the different mission types and the time-varying communication link resources,which achieves efficient matching of mission requirements and multi-dimensional resources,thereby improving resource utilization and improving the amount of downloaded mission data.This dissertation is organized as follows:1.Aiming at the local resource bottleneck problem caused by the mismatch between the distribution of observation missions and transmission resources,a mission-aware resource joint management strategy is proposed to achieve the effective matching of observed missions with link resources,energy resources and storage resources.Specifically,an extended time-evolving graph is first exploited to characterize network resources.Based on the graph,this dissertation formulates the design of mission-aware resource allocation,aiming at maximizing network profit in terms of sum weighted data volume as an MILP(mixed-integer linear programming)problem.Due to its NP-hardness,this dissertation proposes a primal decomposition method to efficiently solve the formulated problem by exploiting its special structure.To further reduce the complexity,this dissertation proposes a link metric considering the issues of residual energy of satellites and the differentiation for missions in the conflict graph.Based on the conflict graph,this dissertation devises a heuristic algorithm to design the mission data delivery strategy.Simulation results demonstrate the efficiency of the proposed algorithms and necessitate the consideration of the differentiation of missions for resource management.2.Aiming at the problem of the constraint relationship between the channel state distribution,the link scheduling and power allocation in mission scheduling,a channel-aware relay satellite mission scheduling scheme is proposed to achieve effective matching between mission requirements and communication links and power resources.Due to the orbiting movements of satellites and atmospheric attenuation,the channel states of the inter-satellite and the satellite downlinks for the relay satellite system are characterized non-uniformity in time and space.The channel state and the available power of links jointly determine the link transmission capability,and the channel state distribution of links restricts the matching between the spatio-temporal distribution of missions and the transmission capability of links.Therefore,in order to improve the matching accuracy,i.e.,to enhance the mission completion rate,this dissertation proposes a strategy of jointly optimizing the inter-satellite links scheduling and downlink power allocation.Specifically,the time-expanded graph is first exploited to formulate the proposed mission scheduling problem with delay constraints as a mixed-integer nonlinear program(MINLP)optimization problem that is challenging to solve.For tractability purpose,this dissertation equivalently decompose the problem into a PA problem and an optimal power allocation-based mission scheduling(OPAMS)problem which is still an MILP.A new two-stage scheme is further devised to efficiently solve the OPAMS problem.Simulation results validate the significant gains of the proposed algorithm in mission completion number and necessitate the consideration of the time-varying and differentiated inter-satellite and downlink contacts.3.Aiming at the resource competition problem between deterministic arrival missions and stochastic arrival missions,a two-stage mission scheduling scheme based on fuzzy information of the stochastic arrival mission is proposed.Furthermore,this dissertation studies the mechanism of the stochastic mission arrival distribution on the coupling relationship between link capacity allocation,storage capacity allocation and energy allocation,thus ensuring the overall network rewards.Considering the dynamic reconstruction characteristics of network resources and the ambiguity of the stochastic arrival missions in the planning periods,a robust two-stage stochastic optimization framework is designed.Specifically,the time-expanded graph is used to describe the dynamic and time-varying resources of the network,and the mission planning problem with delay constraints is formulated as a two-stage stochastic flow optimization problem with the goal of maximizing the total network reward.Then,this dissertation introduces an ambiguity set for the uncertain distribution of the stochastically arrived missions inspired by the idea from the distributionally robust optimization.On the basis of the proposed ambiguity set,this dissertation further proposes a data arrival distribution robust two stage recourse(DADR-TR)algorithm by converting the original stochastic optimization problem into a deterministic cone optimization problem,which is computationally tractable.Extensive simulations have been conducted to evaluate the impact of various network parameters on the algorithm performance and further validate that the proposed DADR-TR algorithm can achieve high data delivery performance without full distribution information of the stochastically arrived missions.