Article Preview
TopIntroduction
Background
A lot of researches are going on different aspects of Cloud throughout the globe. At this moment, Cloud is one of the hottest topic in the world of science and technology. It involves everything of IT world such as Web services, Web technology, network activities, cluster design, network security, Client-server connectivity, and so on. Distributed computing has been utilized to establish Cloud system environment. Parallel and concurrent activities have been used in communication purposes (Andrews, 2000). Client-server techniques are used in service oriented networking activities (http://en.wikipedia.org/ cloud_computing). A typical way of managing software to run on distributed systems is to separate functions into “client” and “server”. A “client” is a program which utilizes services that other programs offer. Further, there are more Cloud services introduced by the researchers (Kundu & Banerjee, 2010; Kundu, Banerjee, & Saha, 2010). Resource selection and allocation techniques have been recommended using specific framework in Cloud (Banerjee, Kundu, Bhaumik, Sinha Babu, & Dattagupta, 2012). Cloud is typically based on the pattern oriented architecture using various software in real-time to execute users' queries (Buschmann, Henney, & Schmidt, 2007).
Main Goal
Main goal is to achieve a classifier based on Cloud structure exploiting the connectivity between clusters. To achieve the target, Cloud manager is responsible to broadcast according to the prescribed protocol set in the paper.
Motivation
Cloud computing in a network environment has become the ultimate necessity for the service providers to maintain huge data in a distributed manner with full or partial transparency. It is helpful for delivering high performances on a range of predetermined services. Cloud is situated on distributed network using local-area and wide-area network. Existing packages use typical classification techniques to handle the situations. The paper proposes a classifier having structural differences providing high performances along with mathematical expressions to understand the viability of the system network.
Past Researches
In the past performance modeling of architectures having distributed nature has been demonstrated in (Spinnato, Albada, & Sloot, 2004). Different designs of embedded systems are being carried out by the researchers in heterogeneous network environment (Radojevic, Salcic, & Roop, 2011). Location is one of the major factor for performance measurement of any network system. Therefore, location awareness in unstructured peer-to-peer systems has been shown in (Liu, Xiao, Liu, Ni, & Zhang, 2005). Synchronization is another important factor in distributed network systems. Time should be adjusted between several server machines at distinct geographical locations for maintaining the synchronization perfectly (Xu, Tang, & Lee, 2006). Better performance typically requires fast data transactions in a dynamic sense (Eckart, He, Wu, & Xie, 2010). Network performance testing is demonstrated in (Pakin, 2007). A lot of synchronization based approaches have shown us the performance implications of memory reclamation and transactional locking procedures (Hart, McKenney, & Brown, 2006; Dice, Shalev, & Shavit, 2006; Herlihy, 2005).