Design And Implementation Of Cloud Resource Management System Based On Dynamic Network

With the development of mobile internet of things system and increasingly popular unmanned system,the traditional cloud resource management system faced new technology requirement,that is to say the centralized resource management mode which supported by the traditional static internet network architecture is changed to the distributed resource management mode which supported by the dynamic network architecture.This thesis took the mobile network with autonomous control and unmanned system as the research subject,to explore the core difficulties of cloud resource management technology based on dynamic self-organizing network,that is: to break through the resource management mode that based on the original logic-centered mode,but emphasized the distributed and peer-to-peer characteristics of unmanaged system as resource management entities;the resource scheduling method under the original statically known network topology architecture not only maintained the original resource scheduling capability and interface,but also can form distributed resource information sharing and cross-node dynamic management that oriented to dynamic changes in dynamic network topology and transmission capability caused by unmanned system mobility.Based on above demand analysis,this thesis designed and realized the mobile cloud resource management system that based on unmanned system under dynamic network environment constraints,through the research of distributed and flat resource scheduling methods,realized the efficient synchronization and sharing mechanism of distributed resource information among unmanned system,and formed the rapid response capability of unmanned system to cross-platform resource requests.Firstly,the cloud resource management system realized the monitoring to dynamic network environment,and realized the self-organizing and self-healing networks by neighbor discovery,dynamic network topology maintenance,and multi-hop,formed a local distributed resource tree directory representation of adjacent platforms on this basis,and finally formed a network resource pool view oriented to different service quality through a cross-platform resource service method.Secondly,the cloud resource management system realized cross-domain interoperability of resource data,cross-platform resource invocation was completed through cross-platform resource request and response,cross-platform resource subscription and notification services.The cloud resource management system with resource integration service,can integrated the similar or related resource,reduced the transmission of redundant data by unmanned system.On this basis,through the distribution planning service of resource data,according to the priority of resources,realized the priority transmission of key data under the condition of limited bandwidth.At last,this thesis established a simulation system under the circumstance of distributed cooperative sensing of unmanned vehicles,realized the cross-platform dynamic resource management and active information sharing between unmanned vehicles,verified the feasibility and effectiveness of the cloud resource management system proposed in this thesis.