On the Impact of Online Social Networks in Content Delivery
Advanced Content Delivery, Streaming, and Cloud Services
As Content Delivery Networks (CDNs) task is the improvement of Internet service quality via replication of the content from the origin to surrogate servers scattered over the Internet, the area of CDNs faces three major issues concerning the maximization of their overall efficiency , : (i) the best efficient placement of surrogate servers with maximum performance and minimum infrastructure cost, (ii) the best content diffusion placement, either in a global or in a local scale, i.e.,
... h content will be copied in the surrogate servers and to which extend, since this requires memory, time and computational cost, and (iii) the temporal diffusion, related with the most efficient timing of the content placement. The increasing popularity of Online Social Networks (OSNs) , ,  and the growing popularity of streaming media have been noted as being the primary causes behind the recent increases in HTTP traffic observed in measurement studies . The amount of Internet traffic generated every day by online multimedia streaming providers such as YouTube has reached huge numbers. Although it is difficult to estimate the page 2 proportion of traffic generated by OSNs, it is observed that there are more than 400 tweets per minute with a YouTube link  . These providers often rely on CDNs to distribute their content from storage servers to multiple locations over the planet. Towards this direction we can exploit information diffusion analysing the user activity extracted from OSNs. Thus, the improvement of user experience through scaling bandwidth-demanding content largely depends on the exploitation of usage patterns found in OSNs, and can be conducted either through the selective prefetching of content, also taking into account timing issues, or through the strategic placement of surrogate servers. Furthermore, the cost of scaling such content in CDNs can be expressed in different ways. For example, it might be the number of replicas needed for a specific source or it may take into account the optimal use of memory and processing time of a social-aware built system. Thus, it is crucial to support social network analysis tasks that accommodate large volumes of data requirements for the improvement of user experience (e.g., through prefetching via a CDN infrastructure). The goal of this chapter is to present existing approaches that can be leveraged for the scaling of rich media content in CDNs using information from OSNs. Specifically, we present a taxonomy of the relative research (outlined in Figure 22 .3 in the next section), taking into account phenomena related with rich media content and its outspread via OSNs, and measurement studies on OSNs that could provide valuable insight into CDN infrastructure decisions for the replication of the content, as well as systems built with the leverage of OSNs' data.