Role of Cloud Computing in On-Line Social Networking and In-Depth Analysis of Cloud-Sim Simulator

Role of Cloud Computing in On-Line Social Networking and In-Depth Analysis of Cloud-Sim Simulator

Suresh Annamalai (Nehru Institute of Engineering and Technology, India) and Udendhran R. (Bharathidasan University, India)
Copyright: © 2019 |Pages: 13
DOI: 10.4018/978-1-5225-9023-1.ch003

Abstract

In this chapter, the authors introduced cloudsim simulator and cloud computing role in online social networking. The communication incurred by other activities such as management jobs is negligible. Social relationships can be established for numerous reasons. For example, family members, colleagues, or classmates often have strong social interactions resulting in large communication load. Cloud computing as well as social network-based applications will become dominant in many aspects of life in the next few decades. The performance of such large-scale systems is characterized by system capacity in terms of number of users/clients, flexibility, scalability, and effective cost of operation, etc. Popular social networks have hundreds of millions of users and continue to grow.
Chapter Preview
Top

A Brief Introduction To On-Line Social Networking

For the purpose of proper understanding in this chapter, an on-line social network application is defined as a system in which communication load is generated mostly by the interactions or social relationships among users or clients. There is no single server or even cluster (a group of servers) can handle such large amount of data. These social networking applications are served by datacenters which consist of hundreds of thousands or even more than one million networked servers in total. The goals of resource allocation problems are to decide how resources are allocated to deliver specific services or to process computation requirements of clients/users. Specifically, solving resource allocation problems is to assign clients/users to servers in order to minimize operational costs or maximize resource utilization. The dependencies among clients/servers might incur inter-server communications, which increases the operational costs. The total operational costs are comprised of some components such as energy consumption, network bandwidth, load balancing, etc. Therefore, the resource allocation problems are often formulated as multi-objective optimization problems.

Figure 1.

Service layers of cloud computing

978-1-5225-9023-1.ch003.f01

Among many technical challenges, resource allocation becomes one of the most important factors determining the viability of systems. Good resource allocation schemes help increase the availability and scalability of the systems as well as reducing operational costs significantly as shown in figure 1. With the number of users of social networking applications increasing quickly during last few years, the data generated has grown dramatically accordingly. Schoolmates and/or transitive friends often interact infrequently, thus result in weak connection. In general, a social network can be represented as a weighted graph of vertices and edges representing users and their social relationships respectively. The weights of edges represent frequencies of interactions among users. Depending on the interests of analysis, edges of social network graphs can be undirected (symmetric) or directed (asymmetric). Some people are talkative, continually posting updated news or information with their friends; others might just enjoy receiving news from friends. Understanding the behaviours of users as well as types of information in each social network might help reveal the network structure and the growth model which support the designing of systems to operate efficiently.

Figure 2.

Social aware cloud computing/Storage

978-1-5225-9023-1.ch003.f02

The weights of edges represent frequencies of interactions among users. Depending on the interests of analysis, edges of social network graphs can be undirected (symmetric) or directed (asymmetric). Some people are talkative, continually posting updated news or information with their friends; others might just enjoy receiving news from friends. Understanding the behaviours of users as well as types of information in each social network might help reveal the network structure and the growth model which support the designing of systems to operate efficiently as shown in figure 2.

Top

Social Network Analysis

There are various methods to characterize and analyze social networks. This section discusses some metrics which are often used in social network analysis. First, let G(V, E) be the graph representing a considered social network, where V and E are the set of vertices (users) and edges (social links) respectively.

Complete Chapter List

Search this Book:
Reset