Research On Interest-based Resource Management Techniques In Cloud VR Environment

In recent years,with the rapid development of cloud-based virtual reality(VR)technology,traditional VR technology is no longer limited by expensive hardware costs,which has solved the problem of the development of traditional VR technology.Currently,it has become a hot topic in the field of computer applications.As the virtual scene construction,real-time rendering,and massive data storage of cloud VR systems are all implemented in the cloud,the amount of data generated by the system is much larger than that of traditional applications.Therefore,how to properly store and schedule resources in the cloud VR environment,and improve the operating efficiency and service quality of cloud VR systems,has received widespread attention and research from more and more scholars.Currently,resource storage processes are chaotic,storage space utilization is low,and improper resource scheduling leads to poor performance in cloud-based VR systems.Based on research and analysis of existing resource management technologies in cloud-based VR environments,this thesis proposes an interest-based resource management technology.This technology not only achieves cloud resource storage based on resource interest according to the correlation between cloud resources but also implements resource scheduling based on user interest according to the correlation between virtual users and virtual resources,effectively improving the operating efficiency and user experience of the cloud VR system.The main work of this thesis is as follows:(1)Addressing the problem of chaotic storage processes and low storage space utilization in current cloud resource storage technologies,this thesis proposes an interest-based cloud resource storage technology.This technology firstly considers the correlation between resources in the cloud-based VR environment,defines and calculates resource interest to quantify the degree of correlation between resources,groups the resources based on the calculated interest,and then uses reinforcement learning to place the resource groups in storage devices rationally.Experimental results show that this technology can eliminate performance bottlenecks caused by uneven resource distribution,improve overall storage space utilization,and increase the efficiency of resource storage in cloud-based VR environments.(2)Addressing the problem of poor performance caused by improper resource scheduling in current cloud resource scheduling technologies,this thesis proposes a user interest-based cloud resource scheduling technology.This technology firstly considers the correlation between virtual users and virtual resources,defines and calculates user interest to quantify the degree of correlation between users and resources,introduces a heuristic Q-learning algorithm,combines user interest with the algorithm to propose a resource-adaptive heuristic action selection strategy,and finally applies this strategy to a multi-objective cloud resource scheduling model to achieve the best action selection for resource scheduling.Experiments show that this technology can effectively reduce the completion time of resource scheduling tasks,improve resource scheduling efficiency,and reduce the overall scheduling overhead of cloud-based VR systems.(3)This thesis designs and implements a virtual farm application system based on cloud resource management.The system consists of three modules: a comprehensive resource monitoring module for the cloud platform,a cloud resource storage module,and a cloud resource scheduling module.The comprehensive resource monitoring module is responsible for monitoring resource usage throughout the cloud-based VR system and conducting data analysis.The cloud resource storage module runs interest-based cloud resource storage technology to store cloud resources in the cloud-based VR system.The cloud resource scheduling module runs user interest-based cloud resource scheduling technology to schedule resources in the cloud-based VR system.The implementation of the system shows that it effectively improves the operating efficiency and user experience of cloud VR systems,and improve the quality of service of the cloud VR systems.