زمان بندی خدمات چندرسانه ای در محیط محاسبات ابری Scheduling multimedia services in cloud computing environment

1. Introduction To meet the requirements of users, many of multimedia service providers (MSPs) provision various multimedia services such as, WebTV, VoD, mobile instant message, voice and video services delivered over Internet (Tan and Su 2011). Multimedia services demand power computational capacity and large storage space to process rich multimedia content. Recently, cloud computing has broad prospect to provide desired resources and provisioning quality of service (QoS) for multimedia services due to its features of on-demand supply and cost advantage (Tan and Su 2011; Zhu et al. 2011). Many of multimedia services in cloud computing environment spring up, such as video surveillance (Ren and Van Der Schaar 2013), cloud-based healthcare (Shini, Thomas, and Chithraranjan 2012), real-time access of medical images and video (An et al. 2014; Yang et al. 2015), cloud based photo and video sharing and so on. Currently, security is a critical factor for multimedia services running in cloud computing environment (Lu, Varna, and Wu 2010; Wang et al. 2014b; Zhu and Lu 2009). Due to inherent distributed deployment of cloud computing environment, it is difficult to ensure each user can recognize a huge number of MSPs and select appropriate and trustworthy MSPs for trade. Furthermore, uncertainty and unreliability of multimedia cloud pose threats to the operation of multimedia services. For example, multimedia cloud services sometimes may be offline unexpectedly due to power outage or other reasons. And the abuses of service resources by malicious users will greatly deplete MSP’s resources and decrease its service capabilities (Zhou et al. 2011). On the other hand, some of selfish MSPs focus on augmenting their benefit and do not worry the benefit of users. In the context, it has contributed to the lack of confidence in the information discovered by the MSPs (Fard, Prodan, and Fahringer 2013). For instance, the MSPs might overstate the capacity of their resources in order to achieve more profits. As an effective mechanism, trust has recently been suggested to enhance security and alleviate attacks in cloud computing environments (Dong and Dai 2008). Executing the jobs on the trusty MSPs can guarantee the security of sensitive data, reduce the services completion time and drop the ratio of failure execution. Unfortunately, most existing scheduling algorithms for multimedia services in cloud computing environment are merely devised to guarantee QoS demands while possibly ignoring the trust. In this paper, we propose a novel scheduling and provisioning architecture for multimedia services in cloud computing environment. Within the proposed architecture, we build a trust model including both subjective trust and objective trust to evaluate the trust degree of a MSP. By employing Bayesian theory, the subjective trust degree between multimedia service providers and users is obtained. According to the attributes of QoS, the objective trust degree of a MSP is calculated. Considering QoS demands in terms of cost, complete time and the requirement of trust for multimedia services in cloud environment, we propose Trust Multimedia Scheduling Strategy (TMSS), a novel trust mechanism that counters the sensitive security of multimedia services. TMSS can heuristically find trusted resource allocations and achieve minimum completion time while meeting the constraints of cost and trust.