It's indicative of the continuing high rate of technological innovation in wireless, but we’re about to see yet another major transition in broadband wide-area wireless, this time from 3G to 4G. This is all the more remarkable when you consider that the history of the field goes back well over a hundred years to Marconi and Tesla, and if you’re a stickler all the way to Benjamin Franklin and his kite.
Yes, it seems as though 3G is barely deployed, but 4G promises, as usual, even higher throughput, and no one will argue that this is always a desirable goal.
And yet it may be that 4G, despite clearly improved specs, isn’t really going to change what us end-users see -- and thus do -- all that much. It may in fact thus be the case that the required huge investment in 4G won't yield improved performance and especially improved price/performance for a number of years yet, which might give the carriers and their CFOs a bit of a pause.
Why? First of all, there’s all that investment. Building out a 4G footprint will cost the carriers tens of billions of additional dollars, and they're still trying to get a return on 3G. As well, it's still early. W e’ve yet to settle even on a definition of 4G.
It's very likely that the ITU, the arbiter of all things definitional, will rule that 4G is appropriate in implementations with a minimum throughput of 100 Mbps. That, to me, is a truly silly number. First of all, every metric we specify in wireless is always an absurd theoretical upper bound that users will never see.
For example, look at the 54 Mbps of 802.11g or the 300 Mbps of .11n. These are signaling rates, not Layer-7 throughput. So, while we might in 4G be able to send 100 million bits over the air, using, by the way, lots of very expensive licensed spectrum, end users won't see anything close to that theoretical maximum.
Add in the variable nature of radio communications and the fact that all that capacity is shared among a growing set of users with equally growing application demands, including support for time-bounded traffic, and there's a real danger that 4G might look a lot like 3G with expectations thus seriously misplaced. That is always bad for business, especially if you're that CFO who's trying to calculate ROI.
Instead, I think it's more likely that extensions to current 3G services will be the direction of choice for mobile broadband users for at least the next five years, and perhaps much longer. HSPA, which AT&T is now upgrading to 7.2 Mbps in response to massive demand from the iPhone community, could eventually, as HPSA+, reach 42 (again, peak) Mbps or even more.
LTE and WiMAX are both starting to appear, both of which have the potential for tens of megabits and beyond. 3G, evolving to what's being called 3.5G -- or, if you’re not too confused at this point, 3.9G, a term sometimes used to describe LTE -- may yet have a lot of life left in it.
Since I brought up the point earlier, I think we should define 4G not by throughput, since that’s confusing anyway, but rather by what it offers the end user -- broadband, mobile, all-IP capacity with support for time-bounded traffic, irrespective of peak throughput, which is exactly what we expect our landline infrastructure to provide. Parity, then, is what 4G is all about, and not necessarily throughput alone. The air looking just like wire to applications; I’d buy that.
So, does 4G belong in your planning? If you accept my definition above, absolutely. LTE is on the way, and let's assume that WiMAX will provide some effective competition, perhaps holding down prices in the bargain. But should you count on vastly improved throughput on the order of 100 Mbps? No, you should not.
A few reliable, available megabits per second ought to make just about everyone happy for the next few years and perhaps a bit longer. And, in case you were wondering, what about 5G, with, say, gigabit performance? Believe it or not, the technology exists. But this one is so far off in the future that it’s suitable at present for discussion only by dweebs -- and analysts.
Craig Mathias is a Principal with Farpoint Group, a wireless and mobile advisory firm based in Ashland, MA. Craig is an internationally recognized expert on wireless communications and mobile computing technologies. He is a well-known industry analyst and frequent speaker at industry conferences and trade shows, and is a member of the Interop advisory board and is program chair for Mobile Business Expo.