This Industry Viewpoint was authored by Cam Cullen, VP of Global Marketing at Procera Networks.
In 1996, Bill Gates wrote a famous article titled, “Content is King,” and the thesis of the article was that content sites, not ecommerce sites, would generate the most revenue on the Internet. Although Gates had an ulterior motive at the time— Microsoft had just created the MSNBC news channel—his assertion is becoming a reality today. The key driver of bandwidth consumption is desirable content. High-bandwidth content such as streaming videos and online games are not just bandwidth hogs, for suppliers of broadband connectivity they are the drivers of future traffic and potential new revenue streams.
A more granular understanding of content consumption is becoming critical, as the types of content that consumers enjoy is driving their usage patterns and broadband demand. Content intelligence provides the user context needed to understand the various types of web content and the way consumers are reacting with it to drive even more broadband usage. Granular content intelligence is critical to improve policy decisions for more efficient network management, and to create more sticky and profitable services in next generation networks.
Internet sites are normally ranked only by “hits” or visits, and not by the bandwidth that they consume. When looking at an average Fixed Line North American customer, some DPI systems would show the network as 90% HTTP traffic, typically with about 60% video and 30% text. While technically true, this information is useless in understanding bandwidth consumption trends and which sites are driving it. For example, a look at CNN.com would simply result in a category of “news” on most policy management systems. However, CNN.com contains various popular sub-sites with lots of videos, advertising, social network buttons, and active discussion boards. These features indicate that the site is sticky and will generate more traffic than a news site containing simple text.
Next generation networks require a solution that enables more granular content information such as whether the streaming video is HTTPS, HTML5 or Flash and what content and which sites are driving high bandwidth traffic so network operators can plan and manage traffic accordingly. In addition to this, more granular content intelligence analytics can not only be used to set policy for network optimization and to deliver high QoE, it can be used in the marketing of new sticky services that drive revenue growth for operators. For example, with a high percentage of bandwidth consumption dedicated to online video, service providers could offer consumers a video “happy hour” where they-zero rate such traffic, or those who show an affinity for Facebook could be offered a bundled “unlimited Facebook” service at a small additional rate.
Content Categorization and Filtering
Many countries around the world require that certain sites be blocked to protect children or for other regulatory concerns. For example, The Internet Watch foundation in the UK, CyberTip in Canada, and the National Center for Missing and Exploited Children in the US all provide blacklists that are commonly enforced on broadband networks. However, current content categorization and filtering solutions have extremely limited functionality, don’t scale well and lack the subscriber awareness required to enforce policy. In general, today’s content categorization solutions also face a variety of obstacles related to cost and scalability amongst others: the numerous racks of costly equipment and space required; high costs of power consumption for both operating and cooling; an excessive compute and networking resources footprint; low performance and scalability; and singular functionality. Service providers need to meet various regulatory requirements throughout the world while also delivering differentiated network-based services to consumers that drive revenue growth.
Most current DPI based solutions are not quite ready to provide both content categorization and policy enforcement at the cost and throughput levels required for next generation networks. Operators demand a solution with an order of magnitude performance improvement at a reduced cost and with lower ongoing operating expense. What is needed is a cost-effective, highly scalable content signature database that can deliver intelligence on tens of millions of URLs with an extremely minimal impact on performance.
Service providers need a solution that combines both policy enforcement and content intelligence while delivering high throughput and scalability, offering actionable insight into URLs, their users, and the applications and bandwidth they run. Then, service providers can offer demand-driven services while delivering both user- and content-based policy management and control capabilities at the speeds, price and scalability that next generation networks require. Providing granular insight into what content is driving bandwidth demand for the creation of desirable new services, while maintaining control over its utilization of network resources is the key to new revenue generation.