4G Mobile Services Perpetuate 3G's Big Flaws

With all the frustration over 3G networks unable to handle the traffic generated by the iPhone, not to mention the slew of Google Android devices now in the works, pinning your hopes on a 4G network is understandable. Sprint, in particular, is playing off these aspirations, advertising 4G networks, while several handset manufacturers are claiming to offer 4G devices as well.

Not only does 4G not exist, but when it does come along several years from now, it won't solve many of the issues that bedevil customers, such as spectrum shortage and lack of device portability across carriers.

Just What Is 4G Exactly?
Despite not yet existing, 4G is cropping up increasingly in advertisements. But what carriers such as Sprint mean by the term is the high-speed WiMax wireless data service that it and its partner Clearwire are deploying in dozens of U.S. cities. Other carriers are slapping the 4G label on a 3G-based technology, LTE, an extension of the top-speed HSPA 3G technology that carriers are only now starting to deploy and one that has not yet been fully standardized.

But as Phil Redman, Gartner's longtime mobile analyst, points out, 4G is still being defined by the international telecom standards bodies. For now, it's a meaningless term that carriers slap on to whatever they want you to think is the next big thing.

Redman notes that a main attribute of the 4G definition that's likely to emerge involves its theoretical maximum throughput: 1Gbps when used in a fixed location, such as from a laptop in a café. (Throughput rates in a mobile context, such as from a smartphone while walking, are usually a tenth of the fixed rate.) Of course, real-world throughput will be much lower -- as is always true of wireless technologies -- than the theoretical maximum. Still, the 4G target represents a 10-fold speed improvement over today's top 3G networks.

The term "4G" won't refer to any specific wireless technology (if specific technologies deliver on the specs, they'll count as 4G), but in practice 4G will likely be offered through the LTE Advanced standard now being defined, Redman notes. The 802.16m standard, currently under development as a successor to the current WiMax's 802.16d and 802.16e standards, could also fit the likely 4G requirements, though Redman expects most carriers will go the LTE Advanced route.

We won't see real 4G until between 2015 and 2020, Redman says -- with full deployment to follow a decade later, if history is any guide. In other words, don't hold your breath for 4G anytime soon.

Why 4G Won't Solve Lock-in and Spectrum Shortages
Even when real 4G exists, it won't cure the two most conspicuous problems that users face today with 3G: spectrum shortages and the inability to use your device of choice with any carrier's network.

AT&T has made a lot of noise that its iPhone users are already straining its 3G network in the United States, and carriers elsewhere in the world have said the same thing. That's just one device used by 3 percent of its customers -- imagine what happens when smartphones, 3G-enabled laptops, and wireless e-readers such as the Amazom.com Kindle and the rumored Apple media slate are commonplace.

Even as carriers bolster their local cell stations' capability to relieve local congestion, the FCC has warned that there is not enough spectrum to go around -- despite the fact that just two years ago a big block of analog-TV spectrum was auctioned off to carriers for use in "3G-plus" LTE and WiMax networks. WiMax is available in about a dozen cities now, while the first-generation LTE network should start coming online in 2011 -- but don't confuse these 3G LTE and 802.16d/e WiMax deployments with the 4G LTE Advanced and 802.16m WiMax networks expected in the next decade or so.

Exponential mobile device growth is one culprit in this presumed shortage of spectrum. The other is the quiltwork of wireless spectrum allocated to carriers. Each country has set aside spectrum for various purposes, but the spectrum used in one country for, say, 3G may be different than the spectrum used in another. Because of this, four spectrum bands comprise the GSM cellular technology used by most carriers. Even within a country, carriers may use different bandwidths for the same technology, based on what became available when they put in their bid.

Thus, in the United States, AT&T uses a different spectrum than T-Mobile does for its GSM-based voice and data services. Each carrier also uses different spectrum for 3G, even though both run the same network technology (UMTS). Many GSM-capable phones work with all four GSM 2G bands by having tunable radios -- but they are rarely designed to work with all the UMTS 3G bands -- and that's why Google's Nexus One, designed for T-Mobile, can't do 3G on AT&T's network.

Throw in multiple bandwidths for the CDMA 2G/3G technology used by Sprint and Verizon Wireless, and you quickly get so many technology and frequency variations that the phones can't easily be designed to support them all. Adding the circuitry and multiple radio tuners to support every possibility quickly causes space, power usage, and heat issues -- and higher costs.

Avoiding these issues is a big reason why smartphones are tied to specific carriers. But the fractured networks that result are subject to oversaturation, as carriers cannot divert excess traffic to one another to better balance the load. AT&T says that most iPhone users' complaints about poor 3G coverage center around New York and San Francisco, where the local networks are saturated -- but AT&T can't divert users to T-Mobile's network, though both are based on the same technology. (AT&T and T-Mobile do divert voice traffic to each other when the 2G network is overloaded.)

T-Mobile's late entry into 3G services via Android smartphones this past year also raises concerns, Redman notes, since T-Mobile doesn't have a lot of spectrum available, a consequence of sitting out several spectrum auctions in the past decade due to economic woes at its parent company Deutsche Telekom. Worse, T-Mobile doesn't have any spectrum that could be used with the expected LTE devices; Verizon and AT&T locked up most of that spectrum in 2008.

When 4G is available, there will be the same split in technologies -- likely LTE Advanced and 802.16m WiMax -- and in frequencies, Redman notes. And the patchwork of spectrum based on who successfully bids as it becomes available will continue. Devices will likely continue to be tied to specific carriers, who won't be able to balance usage across each other's networks.

For example, in the United States, Verizon and AT&T have licensed the 700MHz spectrum for LTE, while European carriers will use the 2100MHz spectrum. Sprint plans to offer WiMax service instead, as that technology was available sooner and Sprint wanted to beat its competitors to the "4G" market -- but Redman notes that very few other carriers have WiMax plans, so Sprint's WiMax smartphones may not be usable in most parts of the world. Also, there may not be enough demand for device makers to bother with dual LTE/WiMax models.

4G's Glimmers of Hope
If there's any hope, it's that the 4G networks will be less fragmented than today's 2G and 3G networks, so device makers might be able to create cost-effective and power-efficient global smartphones, at least for LTE Advanced networks.

Redman expects carriers to offer dual 3G/4G smartphones, using the 4G networks for data services and the 3G networks for voice traffic. Doing so could decrease the congestion by essentially having phones use two networks simultaneously. Plus, this approach would let the carriers load-balance across their own 3G and 4G networks, since 3G supports both data and voice traffic, while 4G can support voice if it's converted to VoIP, à la Skype.

And the near-standardization on LTE Advanced, coupled with the smaller number of frequencies involved, might let carriers perform mutual load-balancing across one anothers' networks to help smooth out capacity gaps even more. That would be particularly useful when one carrier has only a small amount of spectrum in certain geographies and thus can't serve all the demand in those specific areas.

Of course, it's in the carriers' interest to lock in customers by tying devices to their networks, so even if technologically the ability to use your desired smartphone on most carrier networks comes to pass, the carriers may not play ball. After all, that would let customers easily change carriers, leading to price wars and unstable revenues. What's good for users may not be good for the carriers -- and the carriers call the shots, at least within the mosaic of spectrum available to them.


Subscribe to the Best of PCWorld Newsletter