Tracking the All-IP Customer Experience
And the network is carrying more and more IP traffic across all access technologies. The latest numbers from Cisco’s Visual Networking Index (VNI) Global IP Traffic Forecast indicate that IP traffic around the world has increased eightfold over the past five years and will increase threefold over the next five. CAGR for that IP traffic varies by region, ranging from the massive (22% in North America) to the gargantuan (57% in the Middle East and Africa).
An all-IP environment offers lots of advantages over the legacy network it is slowly but steadily replacing. Chief among these advantages is the simplified network architecture that IP allows, leading to cost reductions and improved time-to-market.
The Cost of Complexity
However, like anything else worth doing, there are major challenges associated with this exciting new environment.
Take signaling, for instance. In the TDM world, sending messages was relatively expensive, so signaling was based on a 56K model, wherein tiny individual binary messages, just a few bytes in size, were exchanged, encouraging simplicity even at the expense of content. In an all-IP environment, on the other hand, the relative costs of sending a few kilobytes is practically nothing, so VoIP protocols can afford to up the number of messages sent. A few years ago, a VoIP call might have required 14 messages. Now, an average Voice over LTE (VoLTE) call requires 136.
Furthermore, advanced all-IP networks allow for something that TDM networks couldn't easily facilitate: user mobility. Now VoIP users can treat their voice service more like email and less like legacy voice, in that they can make calls to and from their own numbers from anywhere. But here again, this convenience generates a glut of signaling traffic. In fact, the "registration" process, as users access their VoIP accounts from new locations, often accounts for 90% of the signaling traffic involved in VoIP, while call-related traffic only accounts for the remaining 10%. And when telepresence and IM functionality is involved, that signaling traffic spikes even higher.
In addition, existing tools often share the connection-based paradigm of legacy systems, when the optimal view of modern IP events is based on an analysis of packets rather than connections. In addition, physical networks are growing, involving more and more end devices. Network interconnections are growing too, at least as networks remain in the evolutionary phase between legacy and IP environments. Furthermore, even if monitoring data is available in some format for service providers migrating to an all-IP environment, that data often goes unreported across company departments, so even if engineers are aware of faults or system outages, customer-facing departments are often in the dark.
Moreover, in a collaborative environment like today’s communications market, just sharing information across departments is often not enough. Third party developers and partners need that information as well.
And, of course, customer experience is more crucial than ever in high-choice environment. Customers are becoming more demanding, given the high choice environment they currently enjoy. Churn potential is exacerbated by a culture rife with platforms for heavily amplified complaints via social media. Expectations, meanwhile, are often unsustainably high as subscribers assume that their IP-based communications will be as reliable as traditional TDM, even through the transition.
These needs culminate in a new set of demands for service providers attempting to closely monitor the customer experience in a way that accurately captures the view of the end-user and which enables the sharing of data gleaned through such a view across departments and with developers in a meaningful and actionable way. This new customer-centric view means network monitoring must be more proactive and have more capability to drill-down to individual users. And as the move to an all-IP environment is evolutionary, rather than revolutionary, these needs must be met in a way that doesn’t involve widespread rip-and-replace, as existing infrastructure and software rollouts represent significant investment.
