Research On Dynamic Adaptive Streaming Over Cloud Computing

As the technology of applying a cloud network to cloud multimedia matures,a solution to the present demand for high quality and diversified cloud multimedia can be provided.However,how to meet the users' demand for high quality and diversified cloud multimedia with handheld devices,which have limited power,is an interesting and challenging study,how to reasonably and efficiently utilize the resource of cloud computing and mobile device achieve a cost-efficient and energy efficient resource provisioning(or service provisioning for server provider)with Quality of Service(QoS)guaranteeing is an inevitable and practical problem.In this thesis,based on the combination of cloud computing technology and traditional adaptive video streaming technology,the framework of cloud based adaptive streaming,resolution and bitrate scheduling algorithm with energy-efficient in mobile terminals,QoS-guaranteed and cost-efficient dynamic resource allocation are studied.The detail of our main work is as follows:(1)Research on architechure design of the cloud based dynamic adaptive streaming system.The multimedia computing and storage for DASH system to achieve adaptive video streaming must be be carried out on the terminal,which increases the energy consumption of terminal devices greatly.Therefore,in this paper,a cloud based energy/cost-efficient service architecture for adaptive multimedia streaming is proposed,which can dynamically adjust the multimedia based on the configuration information of the terminal device.(2)Research on resolution and bitrate scheduling algorithm with energy-efficient in mobile terminals based on the proposed streaming architechture.Reference to the existing research,a mathematical model of influential factors of energy consumption of terminal equipment and the multimedia codec parameters has been formulated based on a further analysis of the energy consumption during video playback.Then a scheduling algorithm is adopted to solve the presented model.Compared to the DAMS scheme,theproposed approach can derive more matching target bitrates and resolution for adaptive streaming with lower computational complexity.(3)Research on QoS-guaranteed and cost-efficient dynamic resource allocationalgorithmbased on the proposed streaming architechture.The real-time transcoding is necessary for the proposed architechture to realize adaptive video streamingscheduling,but massive video transcoding demand bring video service providers a huge capital pressure,therefore,QoS-guaranteed and cost-efficient dynamic resource allocationalgorithm is proposed.The dynamic VM provisioning problem for IaaS clouds is formulated as minimizing the number of active VMs subject to a Qos requirement,and then the proposed resource allocationalgorithm is applied to the above optimization problem.Experimental results show that the proposed algorithm reduces the cost of the video service provider and the server overload ratio,improves the average resource utilization and the quality of service.