Showing posts with label ITU. Show all posts
Showing posts with label ITU. Show all posts

Sunday, December 19, 2010

ITU Says Some 3G Networks are 4G, Pre-4G is 4G, and 4G is 4G

The International Telecommunications Union recently defined  “LTE-Advanced” and “WirelessMAN-Advanced” as the only "official definitiions of "fourth generation" networks, automatically making networks operated by Sprint, Clearwire, Verizon, MetroPCS and all other operators of WiMAX and Long Term Evolution networks something other than standards-based "4G" networks.

Now the ITU has muddied the waters even more, saying that some "3G" networks are "4G," while the formal "pre-4G" networks in existence, or about to be built, also are "4G."

"As the most advanced technologies currently defined for global wireless mobile broadband communications, IMT-Advanced is considered as “4G”, although it is recognized that this term, while undefined, may also be applied to the forerunners of these technologies, LTE and WiMax, and to other evolved 3G technologies providing a substantial level of improvement in performance and capabilities with respect to the initial third generation systems now deployed," the ITU says in a new statement.

Huh? Some of us have had no issue with T-Mobile USA saying its new HSPA+ network offers "speeds equivalent to 4G," because the WiMAX and HSPA+ networks do offer comparable access speeds. But it does create a definitional muddle. It's one thing for marketplace contestants to position their networks in one way or another.

It might be quite another for a "standards" body to argue that 3G is 4G, existing 4G is 4G, and other possible networks might also be 4G.

What's the point of a standard when it isn't a standard any longer? In this case, it might mean that the "non-standard" standards will grow organically to the point that the newly-minted "4G" standard simply ceases to be relevant, much as adherence to the supposedly-"legacy" TCP/IP completely killed the shift to new protocols for layers one through four of the data communications protocols.

One might say the ITU flip flop is merely embarassing, and yet another example of standards bodies attempting to define "next generation" networks. It might result in something far more substantial than that. One might suggest that the whole effort now is questionable, in terms of helping shape the development of 4G.

Once critical mass developments around the real-world 4G and advanced 3G networks, services, revenue elements and devices, evolution will happen based on those factors. That doesn't mean operators will abandon the effort to keep developing more-capable networks. But as we have seen with TCP/IP and other data "standards," the market often decides what a standard is.

So far, the markets, and end users, have decided the path for next-generation networks, in large part. That could well happen here as well. No matter what the ITU thinks, if voluntary groups such as the GSM decide to evolve LTE in some other direction, the existence of a formal standard will not deter them.

That is not to fault the well-intentioned hard work of the technologists working on the standard. The point is simply that the global telecommunications industry has yet to prove it can devise a "next-generation" network standard that real-world operators actually embrace obviously, and with great commercial success. Instead, the pattern so far has been that network operators and end users sort of grope towards better solutions as best they can.

But it is equally true that, up to this point, real-world commercial success has not been driven so much by the standards as by solutions that users believe are workable and useful.

For a discussion f the ITU standards, read this: http://www.itu.int/itunews/manager/display.asp?lang=en&year=2008&issue=10&ipage=39&ext=html and this http://www.networkworld.com/news/2010/121710-itu-softens-on-the-definition.html.

For a discussion of the change, arguing that the ITU now has erred twice on the same subject, see http://www.abiresearch.com/research_blog/1520.

It's a Good Thing App Developers Aren't at the Mercy of the ITU

If you want an excellent example of why it is a very good thing that software development now is viewed as occurring in "layers," where underlying communications protocols are abstracted, consider the situation facing mobile software developers.

"If you’re realistic you’ll admit that mobile “strategy” is a 12 month window into the future at best, where very few decisions are robust," argues Nick Jones, Gartner distinguished analyst.

Devices, tools, vendors, network contracts, business requirements, customer attitudes and competitors will change rapidly, probably invalidating some aspect of a strategy every couple of months.

For that reason, it is a very good thing that among the complications an application developer does not have to worry about is uncertainty about communications protocols in layers one through four of the "stack."

If there were tight linkage between layer-seven apps and layers one to four, developers would be in a pickle, now that the International Telecommunications Union has first created WiMAX and Long Term Evolution standards no real-world network uses, and further has complicated matters by saying that existing advanced 3G and pre-standard 4G networks are, in fact, 4G networks.

See this  and this this  for more details.

It's one thing to create standards that allow global networks to communicate. It's a good thing to have an evolution plan for networks that support greater functionality. It isn't so clear how useful it is to create a well-intentioned standard that first defines all existing networks of that type out of existence, before backtracking and declaring all of them to be "standards-compliant," and then to stretch the definition to include some advanced 3G networks as well.

App developers would face much greater uncertainty were they forced to create tools and products that had to track those sorts of changes.

"Tokens" are the New "FLOPS," "MIPS" or "Gbps"

Modern computing has some virtually-universal reference metrics. For Gemini 1.5 and other large language models, tokens are a basic measure...