You are on page 1of 3

QOS Ranking Prediction for Cloud Services

Abstract: Cloud computing is becoming popular. Building high-quality cloud applications is a critical research problem. QoS rankings provide valuable information for making optimal cloud service selection from a set of functionally equivalent service candidates. To obtain QoS values, real- orld invocations on the service candidates are usually required. To avoid the timeconsuming and e!pensive real- orld service invocations, this paper proposes a QoS ranking prediction frame ork for cloud services by taking advantage of the past service usage e!periences of other consumers. "ur proposed frame ork requires no additional invocations of cloud services hen making QoS ranking prediction. T o personali#ed QoS ranking prediction approaches are proposed to predict the QoS rankings directly. Comprehensive e!periments are conducted employing real- orld QoS data, including $%% distributed users and &%% real orld eb services all over the orld. The e!perimental results sho that our approaches outperform other competing approaches. Objective:C'"() computing is *nternet-based computing, hereby shared configurable resources +e.g., infrastructure, platform, and soft are, are provided to computers and other devices as services. Strongly promoted by the leading industrial companies +e.g., -ma#on, .oogle, /icrosoft, *B/, etc.,, cloud computing is quickly becoming popular in recent years. -pplications deployed in the cloud environment +named cloud applications in this paper, are typically large scale and comple!. 0ith the rising popularity of cloud computing, ho to build high-quality cloud applications becomes an urgently required research problem. Problem Definition: E isting S!stem: QoS is an important research topic in cloud computing. 0hen making optimal cloud service selection from a set of functionally equivalent services, QoS values of cloud services provide

e!ecuting a large number of service invocations is time consuming and resource consuming.cloud application1..  4!tensive real. Pro"osed S!stem *n this paper. e propose a personali#ed ranking prediction frame ork.cloud application2.. soft are components are invoked locally.. since the locations of the cloud applications are quite different. Advantages:  This paper identifies the critical problem of personali#ed QoS ranking for cloud services and proposes a QoS ranking prediction frame ork to address the problem.  0hen the number of candidate services is large. different cloud applications may receive different levels of quality for the same cloud service.orld e!periments are conducted to study the ranking prediction accuracy of our ranking prediction algorithms compared ith other competing ranking algorithms . 3ersonali#ed cloud service QoS ranking is thus required for different cloud applications.valuable information to assist decision making.orld service predict the QoS ranking of a set of cloud services invocations from the intended users. the QoS ranking of cloud services for a user +e. cannot be transferred directly to another user +e. and some service invocations may produce irreversible effects in the real orld.g.g. since invocations of cloud services may be charged. to ithout requiring additional real. "ur approach takes advantage of the past usage e!periences of other users for making personali#ed ranking prediction for the current user. *n traditional component-based systems. hile in cloud applications. it is difficult for the cloud application designers to evaluate all the cloud services efficiently.  4ven if the invocations are free. Client-side performance of cloud services is thus greatly influenced by the unpredictable *nternet connections. Disadvantages:  This approach is impractical in reality. *n other ords. Therefore. named Cloud 5ank. cloud services are invoked remotely by *nternet connections.

Servlets. Cloud 5ank2 . some of these components invoke other cloud services +e. 1.ard )isk 5-/ 7 3entium *> 7 ?%. and hotel booking services in 6ig.T/'. 6ind Similar (sers &.B 7 &12/B or more 7 0indo s 8392%%: or 'inu! 7 ..Sco"e of t#e "roject: The business process of this cloud application is composed by a number of soft are components. are provided and deployed in the cloud by other 'odules: 1. <S3 7 =etbeans :.1 7 Tomcat : eb services. CSS 7 <avaScript 7 <ava 7 <)BC.. car rental services. 'ogin and 5egistration 2. >ie Cloud Services ?. here each component fulfills a specified functionality. airplane ticket services.g. -dd Cloud Services $. Cloud 5ank1 @. To outsource part of business to other companies. These cloud services +can be implemented as companies.$ 7 /ySql &. Soft$are Re%uirements : "perating System (ser *nterface Client-side Scripting 3rogramming 'anguage 0eb -pplications *)490orkbench )atabase Server )eployment &ard$are Re%uirements: 3rocessor .