Tech Insights - The Future of Voice
As U.S. operators begin to accommodate VoIP calling over their current networks and operators worldwide work toward a single LTE VoIP standard, it’s not clear how 4G networks, built for data, will handle voice traffic in the future.
It is clear that the focus of wireless network development in the industrialized world, at least for the foreseeable future, will be data-centric. After all, with market penetration for voice services at essentially saturation levels, network operators are pinning their hopes for continued revenue growth on the still comparatively untapped data segment. But some recent items in the industry press have been a reminder that while wireless data services and applications are today’s hot topics, plain old voice (and text messaging) cannot be overlooked when planning future network strategies.
Two current topics of interest illustrate how voice services may evolve in coming years. One is the way that major operators like AT&T and Verizon Wireless have danced around the issue of allowing voice over IP (VoIP) applications to run on their 3G networks. For users, VoIP apps like Skype are attractive because they provide international phone service at a tiny fraction of what would be charged for conventionally dialed calls. For example, based on AT&T’s “standard” rates, a direct dialed 30-minute call from the U.S. to Singapore will cost more than $100. Using Skype, the charge will be less than one dollar. And since most smartphone users have service plans with “unlimited” data, 3G VoIP applications also can be used to make calls that don’t draw from the plan’s limited voice minutes.
While VoIP applications can provide cost savings to subscribers, they are far less attractive to network operators. The problem isn’t revenues lost from overpriced international long distance calls or even giving away “free” voice minutes, but rather excessive loading on 3G networks. In terms of peak throughput, VoIP isn’t anything like bandwidth-hog applications such as video file downloads, but calls using low cost VoIP services often last a long time while consuming several times the spectrum resources needed for calls on a conventional wireless voice network.
Ideally, network operators would prefer that their subscribers confine voice calls to their voice networks, which have been optimized for that purpose. Unfortunately, market forces won’t allow them to simply prohibit VoIP applications. AT&T announced last fall that it was bowing to customer demands to permit VoIP apps on smartphones. Verizon Wireless recently followed suit, inking an agreement to allow Skype apps on a number of smartphone models.
I suspect that when these VoIP apps become available in the next month or two, they will be very popular, particularly for international calling. That will most likely exacerbate the 3G network loading problems that AT&T, in particular, is already experiencing. It might also hasten a move by operators toward usage-based pricing for data service.
BIGGER QUESTIONS LOOM
VoIP apps on 3G smartphones may be a headache for operators over the next couple of years, but a larger voice service issue looms in the longer term with deployment of 4G technologies. Specifically, the question is this: How will these new networks, which are optimized for broadband data, handle voice traffic? And there is perhaps a more fundamental question: Should they routinely handle voice traffic?
For most major operators, delivery of voice services on nascent 4G networks won’t be an operational issue for a while. 4G deployments initially will be restricted to large urban markets, so providing competitive voice service coverage will require that 4G-capable smartphones also operate on the carrier’s incumbent voice network. But eventually, 4G coverage is expected to rival that of the extensive CDMA and GSM networks that pretty much blanket populated areas of the United States and the rest of the world. What then?
To address the issue of providing voice service on 4G LTE networks, a group of about 40 major operators worldwide has formed an alliance in the GSMA’s Voice over LTE (VoLTE) initiative to explore creation of a single LTE VoIP standard. Their focus is on using a rather extensive (and complex) set of network protocols called IP Multimedia Subsystem (IMS) as the basis for dialed voice and text messaging services. This technical approach is not universally embraced by all in the LTE world, but the recent addition of large, historically CDMA, operators like Verizon and Japan’s KDDI to the VoLTE group may make things difficult for remaining holdouts.
Adoption of a single global voice services standard for LTE would certainly be welcomed by the industry as a whole, because it would, at least on paper, bring with it the advantages of seamless worldwide roaming. And while LTE networks will operate on different frequency bands in different parts of the world, a single comprehensive voice and data standard would greatly simplify technical challenges in building a 4G “world phone.” But that does not necessarily mean that individual operators will rush to migrate large volumes of voice traffic onto 4G networks.
Before taking such a step, operators will want to make sure that quality and coverage for voice service will not be degraded. In that regard, a host of questions remain largely unanswered. For example, seamless cell-to-cell handoff is critical for voice service; much more so than for most data apps. How well LTE will handle handoff in extremely dense urban areas won’t really be known until such networks are up and running. Then there is the whole issue of mobility. Mobility management in 2G and 3G networks is based on a highly sophisticated and refined set of network systems and air interface protocols that were designed precisely for that purpose and that work very well. The corresponding mobility system in the IP world, which is where the “all IP” LTE networks will be operating, is comparatively clumsy and unproven.
But the overriding determinant of how fast and how completely voice services are moved to 4G may well be the issue of spectrum efficiency. Any illusion that 4G wireless technology will provide essentially unlimited bandwidth capacity in the manner of fiber optic networks has been washed away by problems 3G networks are having with todayfs relatively modest levels of data traffic. And while VoIP over LTE may have certain technical advantages, it is doubtful that its spectrum efficiency can match that of networks, particularly CDMA 1X-RTT, that are optimized for voice.
In summary, many questions remain about what might seem like an inevitable migration of voice services to wireless data networks. In the near future, operators may need to find ways to discourage use of 3G VoIP apps even as they start accommodating them. Longer term, some carriers may have less enthusiasm for moving voice traffic to 4G networks than is currently anticipated.
Drucker is president of Drucker Associates. He may be contacted at firstname.lastname@example.org.