Distributed hybrid cloud for profit driven content provisioning using user requirements and content popularity

作者：Daniel D; Raviraj P

来源：Cluster Computing, 2017, 20(1): 525-538.

DOI：10.1007/s10586-017-0778-7

• Data center management
• Multimedia cloud service provider (MCSP)
• Hybrid cloud
• Fast quadratic Lyapunov algorithms (FQLA)
• Quality of service (QoS)-aware requests scheduling
• Kernel support vector machine (KSVM)
• Energy

Growing use of internet turn people to use hand held devices such as smart phone to download/view the multimedia contents at any time with more flexibility. Due to limited resources of handheld devices, multimedia content handling becomes more complex task which can be resolved by using the popular field called cloud computing. The resource handling can be optimized by maintaining both private (permanent storage) and public (temporary and emergency storage) clouds, so that user can respond with required multimedia contents without interruption. However the multimedia provisioning such as video, audio and games downloading through cloud affects the profit of both cloud users and cloud providers if it is not handled with more concern. This problem is resolved in the proposed research methodology by constructing the novel frameworks that can leads to profit aware multimedia contents handling. In the proposed research work, profit of multimedia cloud service providers are increased by constructing the user profile that includes information about user services using kernel support vector machine, so that resource utilization can be optimized in case of current resource failure. And then scheduling and rescheduling of multimedia contents storage level is done by using the Fast Quadratic Lyapunov Algorithms in different time granularities. Finally, popularity based cache management is introduced to reduce the unwanted costwastage while downloading same multimedia content multiple time. The overall results of the work is simulated in the multimedia cloud environment and it is proved that the proposed research methodology tends to provide profit aware multimedia provisioning to the cloud users with increased performance.

• 出版日期2017-3