This page is an archive from our previous website. Please check out our new website where you can read new COMMENTARY eNewsletters, TELL IT LIKE IT IS blog posts or Press Releases.
I think those who freeze their existing 3G networks in favor of LTE are more likely to abandon the 3G networks before they abandon the 2G networks

2G and 3G: Which Goes First?

Thursday, July 02, 2009

Analog was first deployed in the United States in 1981 and was not turned off until February of 2008. By then, only cars with OnStar, some home alarm systems, and some machine-to-machine applications were affected. Most people had stopped using analog services years before. The end of analog came when the FCC ordered it turned off, not when network operators had replaced it.


In reality, network operators had been replacing analog over the years, leaving up just enough of these networks to provide services and reusing the rest of the spectrum for 2G or 3G technologies. In many cases where new cells sites were put in for increased capacity, analog was not even considered, and the last few analog-only sites were decommissioned before the FCC mandated date.


Let me state for the record that 2G and 3G services will be with us for a very long time—LTE won’t replace them anytime soon. When that time does come in 2015 or beyond, I think we might be in for a few surprises as to how and why 2G and 3G are taken out of service.


I don’t believe the FCC will mandate an end to 2G or 3G services in the United States as it did with analog and analog TV. 2G and 3G services are digital, and for this reason alone, I think the FCC will leave it up to the operators to decide when to stop deploying more 2G and 3G sites, and when to start decommissioning existing sites in favor of LTE.


It will be a number of years before LTE includes voice, and LTE will continue to be an evolving standard just as 2G and 3G are still evolving today. There are more upgrades for 3G systems in the works, and AT&T is upgrading its HSPA network to higher speeds this year. Depending on the engineers you talk with at the various companies, you may hear that using 10 MHz of spectrum for 3G upgrades can produce the same results as LTE. The advantage to this is that 3G upgrades leverage an existing technology and build-out. Others believe it is advantageous to go directly to LTE whenever possible. In others words, the technology community is divided when it comes to upgraded 3G performance versus LTE performance given the amount of spectrum available. However, they all agree that LTE is the clear choice when more than 10 MHz of spectrum is available.


You can tell that LTE means different things to different operators by watching what AT&T Wireless and Verizon Wireless are doing. AT&T is spending a lot of money upgrading to 5-7 Mbps 3G and will, I believe, begin rolling out LTE hot zones where data traffic warrants it. Meanwhile, Verizon is treating LTE as a completely new network rather than an upgrade. Verizon has stated that it intends to aggressively build out its LTE network across the United States and has publically outlined some very ambitious goals for this build-out. These two scenarios, and others, will be repeated around the world. And several network operators that have not yet moved to 3G have indicated that they might actually skip 3G and go directly to LTE.


What does all of this mean for 2G and 3G services? That depends on your point of view: upgrade first and build LTE later, or freeze 2G and 3G in place and start now with LTE. I think those who freeze their existing 3G networks in favor of LTE are more likely to abandon the 3G networks before they abandon the 2G networks. Why? Because LTE is an all-IP network architecture that is full IP-based and provides robust data that costs network operators less per megabyte to deliver than their 3G network (although this can be debated all day long).


Since voice is still a big part of their revenue, and today it is still less expensive to deliver voice over a 2G network than a 3G network, it makes sense to keep 2G voice services intact. In the CDMA world, especially where the next rev of CDMA 1X will provide 4X of  voice capacity on the systems, LTE will be used for voice services (actually Voice over IP), but it will be a number of years before voice can be delivered on LTE as inexpensively as on a 2G network that has been in the ground for many years, has been amortized over that time, and typically provides the best coverage (today, 92% of the U.S. population has access to data services while 96% has access to wireless voice services).


This prediction will probably stir many debates. Some will argue that I am dead wrong and that the older technology will be the first to go. Others will agree with me, and still others will say that 2G and 3G networks will be phased out at the same time. But I am not talking about tomorrow; I am talking about sometime in the future, probably after 2015 or so. In the meantime, we will all be carrying multi-mode devices that will be capable of 2G, 3G, and soon, 4G services. What we will see first on 4G networks will be USB and PC Card wireless modem solutions, and then LTE embedded into smartphones. But it will be a very long time before mid-tier and entry-level devices have LTE capabilities, or broadband data for that matter, and they will continue to run on 2G networks, providing a nice revenue stream for network operators even as the selling price for voice comes down.


Even as the percentage of revenue from data services continues to climb—and it is growing rapidly in the United States—voice still pays the bills. Typical percentages for data income are now around 20%, having been in the sub-10% range only a few years ago. But in Japan, where wireless data has been used for Internet access for a very long time and home computers are not nearly as prevalent, data revenue is running about 40%. Voice is still keeping the lights on, even in Japan, and delivering voice as inexpensively as possible is a key concern to network operators as is cutting their cost to deliver broadband quality data.


Some will say that today the iPhone is driving more data than ever before on networks, and they would be correct. But what percentage of AT&T’s income for the iPhone is voice and what percentage is data? I don’t think AT&T breaks this out, but I would be very surprised if the data revenue was more than 40% of the total. Many revenue predications I am seeing are based on the belief that we will all surf the web and use our mobile browsers exactly as we do on our desktops and show data generating the bulk of the revenue. I disagree with this. I think we will see more of an entirely new class of application that does not require invocation of a browser. Instead, these applications will use the power and information on the Internet or back at our corporate offices in the background. This means we won’t be using as much data for typical applications. On the other hand, we will probably be using more data for video, gaming, and similar applications.


Once we get to LTE, it will be next to impossible to break out revenue from data versus voice since everything will be packets flying through the ether. A bit will be a bit, regardless of whether it is a video, data, or voice bit. When we finally reach an all-LTE (and beyond) world, voice, text, MMS, data, audio, or video won’t matter to the network or devices. They will all be packets and income will be income.


But we are a long way from this. In the meantime, we are at the beginning of the LTE era, still living with 2G and 3G networks, and still enhancing them and building more sites capable of one or both technologies.


Does all this really matter to consumers or business customers? They will be getting a new computer modem and a new smartphone midrange or entry-level device. They will pay their money and the devices will do what they are supposed to do. They won’t notice whether they are using a 2G, 3G, 4G network or a combination of all three for voice (which will be the case with some networks for a while). But they probably will notice the difference between 3G and 4G data services and speeds.


Now you know why I believe that 2G for voice outlives 3G for data, and that 3G spectrum will be converted to LTE over time beyond 2015. There won’t be any immediate impact on the network operators or the customers, but the industry should keep this in mind as LTE is rolled out around the world differently in different markets and networks.


Whichever of the technologies is phased out by the network operator, it will be the operator’s decision, not regulatory agencies as it was with analog. The network operators’ decisions concerning how and when to reorder their technology offerings will depend on customer uptake and perceived value of services. The operators’ goal is to provide the best possible voice and data service to as many customers as it can—and that means managing their networks. Having 2G, 3G, and 4G technologies gives them more tools to meet this goal.


Andrew M. Seybold

COMMENTS: This is an archived post. Commenting is no longer available.

Lokdeep Singh - 07/02/2009 16:20:30


Good post. I agree that voice over LTE is years away and most LTE deployments will use some type of fall-back ( to 2G) mechanisms for carrying voice. However adoption of LTE overall, will probably be faster than what you indicated as the pressure to control cost per delivered bit will mount in an environment of "all-you-can-eat" data plans.

Also the technology is being developed ( packet inspection and flow based billing) to differentiate between what an IP datagram carries.


Andrew Seybold - 07/02/2009 17:52:04

Lokdeep--thanks for your comments and yes I know that packet inspection is being developed, the question is if it will be used or not and that will depend upon the network operator--some will just manage their networks better than others--as it is today with 3G, for example, I believe that Verizon manages its 3G resources better than Sprint does--Sprint seems to believe that it has unlimited bandwidth so MVNO's and unlimited data are still available, while Verizon has, I believe, recognized the realities of broadband--it must be managed in order to insure that all customers have access to the best possible data rates and capacity, The importance of network management does not go away with LTE, it will still remain a vital part of the network operators requirements.

Syed Hosain - 07/05/2009 23:29:41


I think you are absolutely right in your comments. Clearly, LTE has promise for high-speed data, and there is market pressure to deploy that in select urban regions. However, the fact that Verizon is deploying LTE in 700MHz - new spectrum - should clearly explain how they believe it should fit into their future. As an adjunct to ANSI-2000, not as a replacement - certainly not for many, many years!