Today’s mobile operator customer spends the majority of her data services quota on video. We know this because we can measure the amount of video traffic moving across mobile networks and can see that it accounts for over 60% of a typical subscriber’s mobile data consumption. It’s thus fair to say that quality of experience (QoE) for a typical subscriber is most closely tied to her experience watching video content. Get that wrong and the subscriber starts looking for alternatives.
Many mobile operators proactively manage data traffic in order to optimise both video and web QoE. By observing the changing behaviour or traffic on their network, they can forecast and establish management and optimization policies based on that expectation. For example, when a new video streaming app such as Meerkat or Periscope takes off, it can disrupt the network and subscriber QoE. If operators can see these changes coming, they can act in advance to maintain the best possible service.
A fundamental shift in the nature of mobile data traffic is making it harder for operators to see these changes coming. That shift is encrypted content. With most of this encrypted content being served by YouTube, Google (non-YouTube) and Facebook, it’s no surprise that nearly half of that content is video; this fact has significant implications for operators seeking to proactively manage subscriber QoE.
The first of those implications is content “blackout.” When content can’t be identified, both content-specific forecasting and proactive policy control and optimization become challenging. That’s a problem because, among other reasons, the wasted bandwidth associated with ever more popular adaptive bitrate (ABR) video protocols leads to a higher incidence of congestion, not to mention increased infrastructure cost necessary to accommodate that excessive bandwidth consumption. To top it off, the presence of encryption increases latency. All this adds up to degraded QoE.
Why are content providers encrypting their content if doing so degrades QoE? Certainly it would make sense for content that needs to be secure such as banking transactions. But cat videos? The likely answer is that content providers want exclusive access to consumers’ surfing information. Such exclusive insight into browsing habits is a competitive advantage in their core business of advertising. Protecting that advantage is paramount and encryption is an excellent tool to achieve that end.
Operators have a different end in mind: ensuring subscriber QoE. Thus, the operator and the encrypting content provider are at odds. For the subscriber’s part, she’s typically unaware that the content she’s watching is encrypted and, were she aware, the operator would likely receive no “credit” for a poor experience. A subscriber frustrated by video stalling and slow-loading webpages will blame the network, not the content provider, and may well seek to change.
It is thus up to operators to take control of encrypted traffic on their networks, as they’re unlikely, for now, to get much sympathy from content providers. It’s reasonable to assume that encryption is here to stay. Architecturally speaking, operators are the only ones in the position to effectively manage traffic, encrypted or otherwise, as only they can take into account the collective subscriber experience owing to their all-seeing view of the network. The good news is that non-intrusive tools already exist to allow operators to characterize and manage encrypted traffic. Operators should ask their vendors what they can do to help.
By Mark Davis, Sr. Director, ByteMobile Product Marketing, Citrix
Mark Davis is the Sr. Director of Product Marketing for the ByteMobile BU within the Citrix Delivery Networks business, where he leads worldwide marketing and product marketing for ByteMobile traffic management, analytics and application delivery products to communications service providers. Mr. Davis has twenty years of wireless and telecommunications industry experience ranging from start-ups to AT&T and Lucent Technologies. He earned a B.S. in Aerospace Engineering and an MBA, both from the University of Virginia.
