What it Takes to Assure QoE and Succeed with VoLTE

Posted by Mae Kowalke on Thursday, February 11, 2016 with No comments

Senzi Fili and RCRWireless recently collaborated on a report (Getting the Best QoE: Trends in Traffic Management and Mobile Core Optimization) examining the challenges mobile operators now face around assuring service quality. The interviews on which the report is based (mostly with companies specializing in mobile network performance assurance, including Alepo, Antritsu, Ascom, Guavus, and Viavi) paint a useful picture of market forces pushing network virtualization and automation. 

The report's conclusions are especially telling as they relate to voice over LTE (VoLTE):

  • Operators are taking their time deploying VoLTE because they get only one chance to make a good impression.
  • QoE is highly subjective, and in some cases the subscribers themselves are the best judge. Bringing insight from subscribers into the mix is therefore necessary. Example: for VoLTE, getting the MOS for every call and every user is a good starting point.
  • With the move to all-IP and VoLTE, QoE assurance is no longer a matter of determining whether a service is working (on) or not (off); operators must also be able to understand how well services are working for individual users.
  • Finding root causes of QoE issues is crucial, and also complicated. End-to-end visibility (needed to account for non-uniformity in mobile networks)  involves applications, end devices, signaling, backhaul, RAN, and core.
  • Although it's involved to do so, it is now possible (and affordable) to collect data from many sources and glean actionable QoE information down to the end device and application level.
Other conclusions from the report include:

  • Mobile network management is now hugely complex, and it is difficult to extract actionable insights from the many data sets available.
  • Automation speeds up the troubleshooting process significantly, making it possible to analyze data as soon as it becomes available.
  • For operators, "real-time" basically means the ability to conserve the expected customer experience, by preventing a problem or fixing it very quickly. Self-organizing networks (SONs) are the only way to be really proactive.
  • Policy control is now a continuous feedback loop, with the operator setting policies based on real-time network information, and then using real-time analytics and insights from subscribers and the network to make business decisions and define future policies.
  • When operators talk about personalization for services they offer, what's meant is understanding the QoE for every data consumer on the network. This means having the ability to identify that an issue exists and then drill down into the details--even as far as a specific device or application.
  • Traditionally, deep packet inspection (DPI) has been used to peer inside packets and determine how the customer is using them. But, DPI has limitations: it's sometimes prohibited because of data privacy concerns, and doesn't work for encrypted sessions.