A Trust Evaluation Method for Cloud Service with Fluctuant QoS and Flexible SLA

ICWS(2014)

引用 19|浏览17
暂无评分
摘要
The QoS (quality of service) of a cloud service is not always trusted as advertised, due to the variable network environment or fake advertisement reasons. Therefore, to promote the trusted service selection, we should first evaluate the trust of each cloud service, based on its historical QoS records from past invocations. However, different from the traditional web service whose historical QoS record is a fixed value (e.g., a historical latency record of a web service is 2s), the historical QoS record of a cloud service is usually not fixed, but fluctuant during the long-running period of a single service invocation. For example, an virtual organization O rents cloud service WeatherInquiry between 8:00 am and 8:00 pm so that the employees of O can access WeatherInquiry. In this scenario, the latency of WeatherInquiry is fluctuant from the perspective of organization O, during the served 12 hours. In this situation, it is a challenge to evaluate the trust of service WeatherInquiry, based on WeatherInquiry's multiple historical QoS records that fluctuate continuously. In view of this challenge, we introduce a novel concept of flexible SLA, to better accommodate the fluctuant QoS of cloud service, and further put forward a trust evaluation method based on fluctuant QoS and flexible SLA, i.e., FL-FL (FLuctuant QoS-FLexible SLA-based trust evaluation method, FL-FL). Finally, a set of experiments are designed and deployed to validate the feasibility of our proposal, in terms of effectiveness and efficiency.
更多
查看译文
关键词
trust,cloud service,quality-of-service,quality of service,trusted computing,cloud services,cloud service, trust, historical record, fluctuant qos, flexible sla, long-running period,trusted service selection,qos record,trust evaluation method,fluctuant qos,virtual organization,cloud computing,web service,historical record,web services,long-running period,flexible sla,organizations
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要