Abstract
This chapter provides an overview on the recent advances and perspectives on content delivery networks (CDNs). After a reminder on the definition and core features of CDNs, the first section highlights their importance with quantitative illustrations. The second section identifies the various types of CDNs which have been deployed to address different markets. The growth of the CDNs has been driven primarily by video streaming. Next to media content, CDNs have evolved to deliver always more demanding social networks and applications. Security solutions are now fully integrated into CDNs and marketed as flagship products. The third, fourth, and fifth sections outline the challenges and technical evolutions of the CDNs to keep up with their customers' hunger for media content, web performance, and security. The sixth section focuses on the convergence of CDNs and clouds. The seventh section reviews the status and perspectives of different approaches for using multiple CDNs. The last section presents the current positioning and future perspectives of the CDNs in the mobile domain.
TopIntroduction
A Content Delivery Network (CDN) is a set of servers specifically designed and deployed over one or several networks for optimizing the storage and delivery of content (e.g., web objects, audiovisual live or on-demand content, large files). From a high-level and functional perspective (Figure 1) the main components of a CDN include: request routing servers that handle and redirect content requests towards cache node servers, cache node servers (a.k.a., CDN delivery servers) that deliver the requested content, content ingestion servers that ingest content in the CDN, analytics and accounting servers, and management servers.
CDNs are expected to handle 72% of the global Internet traffic by 2022 while it handled up to 56% in 2017 (Cisco, 2019a). The outstanding development of CDNs since the late 90s has been driven by their intrinsic strengths: reliability, scalability, improved service latency thereby leading to a better quality of experience for end users, and better network resource utilization leading to a reduction of congestion risks and costs of interconnection links.
From a business perspective, there is an increasing trend for “commoditization” in the CDN industry with strong market competition and price decline. In such challenging context, CDNs must evolve to meet the (new) requirements of the supported applications. CDN players concentrate their technical and business strategies on fast evolving and promising areas such as web acceleration, web security, “serverless” edge computing, convergence with clouds, multi-CDN systems, or IoT and mobile networks.
This chapter aims to provide an overview of these recent advances and perspectives on CDNs.
This chapter is organized as follows. First, the introduction sets the context. After a short reminder of the definition, role and core features of CDNs, this section highlights their importance in today’s Internet and provides some quantitative illustrations. The second section identifies the different types of CDNs currently deployed and the market segments targeted. The growth of CDNs has been driven primarily by video streaming in the last decade. Next to media content, CDNs have evolved to deliver content from always more demanding websites, social networks, and web applications. Besides, web security solutions are now fully integrated into CDNs’ infrastructures and marketed as flagship products. The third, fourth, and fifth sections outline the challenges and technical evolutions of CDNs to keep up with the intensifying needs for media content, web performance, and security, respectively. The sixth section focuses on the convergence of CDNs and clouds. The converging and competitive landscape of online content delivery and cloud services opens new business opportunities for the market players, as well as technical challenges and innovations, introduced in this section. The increase in the number of CDNs and related players stimulate the development of multi-CDN strategies. The seventh section reviews the status and perspectives of different approaches for using multiple CDNs. The last section discusses CDNs with a focus on cellular networks. It particularly outlines the current positioning and future perspectives of CDNs in the mobile domain (a.k.a., cellular networks).
TopCdn Market Ecosystem: An Overview
This section introduces CDN positioning within a service delivery ecosystem. Then, it provides an overview of the main CDN market players.
Key Terms in this Chapter
Over-the-Top (OTT): Designate an Internet market player, application or service that does not rely upon Telcos or internet service providers, except for just getting connected to the end-users via the internet.
Cloud: Set of servers as well as software and databases that run on them, accessed remotely, possibly via Internet, to manage, collect, manipulate, store, and deliver data.
Content Delivery Network Interconnection (CDNI): Relationship between CDNs that enables one CDN to provide content delivery services on behalf of other CDNs.
Adaptive Bit Rate (ABR) Streaming: Content delivery scheme based on segmenting content into small objects called segments encoded at multiple bitrates. The client accesses the content by requesting one after the other the segments that best fit the terminal and network conditions.
Content Delivery Network (CDN): Set of servers specifically designed and deployed over one or several networks in order to optimize the storage and delivery of content objects (e.g., web objects, audiovisual live or on-demand content, large files).
Home Network: The private local network inside the end-user’s premise, including all the terminal devices connected to it. The home network is connected to the internet via a home gateway.
Caching: Special form of memory deployed in networks (on specific servers) as well as on computer architectures and web browsers. When the same object is requested several times, caching aims at avoiding this object to be accessed as many times from its origin location so as to minimize resource utilization and latency.