Over the Top – VoIP
Wireless operators have become pipes. We still rely on many of the services they provide, but at the end of it all, they are internet pipes to an always connected world. Data only plans which have long been predicted by CEOs such as Randall Stephenson of AT&T are really beginning to take form. For instance, a user could get a Google Voice account in order to have phone number, place calls over Google Voice. Use the Google Voice number to sign up for Viber or Vine and still have friends available to voice chat and text with on Facebook, Skype, Google Hangouts, and Facetime (among many others). Over the top (OTT) applications have become a separate standard of how we communicate with friends, family and even in workplace. Rarely does someone directly dial people in another country through traditional telephony. OTT applications have made the world closer, at our fingertips, and for the most part free.
Operators have long been devising plans of how to deal with this lost revenue by focusing on data networks, low cost capacity (backhaul cost is typically 10% of what it was just 5 years ago for operators), and deploying that rarer commodity, spectrum. VoLTE/ViLTE roll-outs are to over simplify it, just OTT applications with higher network priority. The typical open LTE networks of today support most of these features well, how they work in time as capacity and spectrum dwindle will remain to see. The race will come between the VoLTE quality with the higher prioritization and the engineers who’s coding schemes will create the potential for higher quality with less packets muting some of the gains from prioritization.
Testing OTT applications will become more and more critical for operators in the coming years. Audio/Video Quality, network load impact, device drain, WiFi vs. Network, Technology performance difference (LTE, HPSA+, EVDO, etc). OTT applications provide new challenges in how to check network performance KPIs on systems not under their control. Essentially how does a Skype or Viber call peg a dropped call that can be looked at and resolved by engineering teams? New focus will need to be brought on existing network KPIs that indicate problems in OTT applications. Full engineering studies will need to be put in place to measure what KPIs/issues on the network correlate to OTT performance degradation and issues.
In studies by NSN Smart Labs, they show the early versions of the OTT applications being similar in quality, but having more signalling overhead. This is something that every company does differently, but they also know their survival is based on not clogging up the mobile networks they depend on. This simple requirement will conitnue to push them to develop efficient applications in terms of required bandwidth, signalling, and device battery drain. The top level people know it is coming, the question is whether to work with them and embrace the technology or only stand behind the operators VoLTE alternative and take harsh measures on limiting OTT applications as some operators have begun to do.