Research On Computing Offloading And Inter Satellite Routing Strategy In Leo Satellite Network Scenario

The Low Earth Orbit(LEO)network is an important part of the spaceground integration network,and with the advancement of on-satellite processing capacity,LEO satellites can take on more and more computing services.However,there are differences in population distribution due to geographical reasons,the problem of uneven distribution of satellite network services has been caused,and a single satellite may not be able to independently undertake the service requests received locally.In this case,satellites can offload part of the service traffic to other satellites through inter-satellite links,and other satellites use idle resources to assist processing.At present,there is a lack of a joint optimization mechanism for computing offloading and resource allocation suitable for multi-type service performance optimization.In addition,when the idle resources of neighboring satellites are insufficient,it is also worth studying how to use inter-satellite routing to assist computing offload and optimize service performance.Based on this,this thesis conducts the following research on the above two problems.Firstly,aiming at the scenario of highly dynamic satellite network with uneven distribution of service requests,a traffic management and resource allocating joint optimization mechanism for dynamic task offloading is proposed.The proposed mechanism adopts the Lyapunov dynamic optimization framework,establishes a general service requirement model and affect service traffic and performance by adjusting application parameters,which can be applied to multi-type service performance optimization in highly dynamic network scenarios.This thesis takes the secure access authentication and remote sensing image recognition for LEO satellite networks as examples to verify the versatility and effectiveness of the proposed mechanism.Secondly,when neighboring satellites cannot provide sufficient idle resources to assist in processing task requests,the source satellite needs to find a farther auxiliary satellite through intersatellite routing for offloading.Considering that the rapidly changing dynamic topology of LEO satellite networks will affect the choice of inter-satellite routing,this thesis introduces time expanded graph(TEG)to describe the dynamic topology of satellite networks,and proposes a routing algorithm based on time expanded network and multi-commodity flow(MCF)with the minimum task delay as the optimization goal.Simulation results show that the proposed algorithm can effectively reduce the task delay.