Friday, January 24, 2014

Is there a Spectrum Crunch?

It might seem odd to question whether there actually is a spectrum or capacity problem in the mobile communications business. 

After all, “globally, we are seeing continued significant (Internet data consumption) growth at 50 percent to 90 percent compounded rates over five years,” said Cisco VP Robert Pepper.


And consumption is increasing, with each new generation of mobile networks. “An LTE user consumes 19 to 25 times as much data as a 3G user,” said Pepper.


And that happens under circumstances where “about 66 percent of mobile data is offloaded to Wi-Fi,” Pepper added. So, to nobody’s surprise, Cisco argues that the gap between available spectrum to support growth and what is needed shows a widening gap.

Still, some might argue the future spectrum gap is the issue, not any immediate current bandwidth shortages. Forecasts of mobile Internet user growth illustrate the potential size of the challenge.


To be sure, there are ways to increase the supply of spectrum. “Possibly 50 percent of the frequencies useful for communications are licensed to government entities,” said Rob Frieden, a professor of telecom law at Pennsylvania State University. “The U.S. government can help free up spectrum, or keep it.”
That will not be an easy process. “In much of the (Asia-Pacific) region, there still is a PTT mindset,” said David Satola, The World Bank USA Lead Counsel and VP. “Equity sometimes is viewed more importantly than efficiency.”


Still, there is “huge demand for mobile Internet,” meaning there will be “lots of battles over who will get to use it,” Satola noted.



Releasing spectrum also will be costly. As always, that raises issues about how to encourage new entrants. 

“How do you encourage entities other than the carriers to acquire the spectrum?” asked Tim Hewitt. KVH Co.senior international legal counsel.


Issuing unlicensed spectrum, of course, is one way of encouraging new application, service and product providers to enter the market.


In that regard, the industry will “need more licensed and unlicensed spectrum, especially short range (5 GHz) for unlicensed” spectrum, in part because “we will be using 80 MHz to 100 MHz wide channels,” said Pepper.


One thing is clear: we should not underestimate the amount of growth. We have done that in the past, though it also is fair to note there have been times when mobile bandwidth consumption forecasts have been overstated.

Should Content Delivery Networks be Illegal?

Have you ever heard anybody argue that Akamai and all other content delivery networks should be outlawed?


After all, a content delivery network gets paid money by some app providers--not all--to speed up delivery of packets over the backbone, providing a better quality of experience for end users.


That isn’t a violation of the notion of an open Internet. Everybody can get all lawful content, either best effort or enhanced by use of a content delivery network.


So, seriously, have you ever heard anybody argue that CDNs should be outlawed, as a violation of the open Internet?


The reason one asks the question is that “network neutrality” rules are the same thing. The only difference is that the CDN function extends not only from data center to data center across the backbone, but extends to the end user location.


By outlawing anything but “best effort” access at the end user location, one also prohibits optimizing a Skype video session while that session is underway, even if that is what the end user wants.


“Best effort only” rules likewise prevent a user from specifying that voice calls get priority when calls are in progress, or that a Netflix viewing gets priority when the family is watching a movie.


So it is disheartening that Reed Hastings, who has done a marvelous job at Netflix, confounding his critics time after time, makes statements that are, to be polite, untruthful.


Hastings, like many application providers, has a position on network neutrality opposite that of Verizon.


So it is unfortunate that his latest letter to shareholders literally distorts and misrepresents existing U.S. Federal Communications Commission policy on impeding or blocking lawful content.


Since 2004, it has been clear that consumers have the right to use all lawful content. The FCC reiterated the policy in 2005.


In 2010 the FCC again clarified that content cannot be blocked. The FCC also said that “A person engaged in the provision of fixed broadband Internet access service, insofar as such person is so engaged, shall not unreasonably discriminate in transmitting lawful network traffic over a consumer’s broadband Internet access service.”


As a practical matter, that means no lawful applications can be blocked or impeded, the exception being measures taken to manage the network.


In practice, that means any ISP that actively tried to “slow down” a competing service’s packets would quickly run afoul of the FCC, just as much as if it had tried to block a lawful application. Some skeptics might argue the FCC would not act, but many would find little incentive or precedent for that position.


Hastings argues in the letter that, “ In principle, a domestic ISP now can legally impede the video streams that members request from Netflix, degrading the experience we jointly provide.”


To put matters politely, that seems a clear misreading of the FCC rules. The FCC’s are based on three principles, transparency, “no blocking,” and “no unreasonable discrimination.”


The “no blocking” principle is that “fixed broadband providers (such as DSL, cable modem, or fixed wireless providers) may not block lawful content, applications, services, or non-harmful devices.”


“Mobile broadband providers may not block lawful websites, or applications that compete with their voice or video telephony services.”


The “no unreasonable discrimination” rule says that “fixed broadband providers may not unreasonably discriminate in transmitting lawful network traffic over a consumer’s broadband Internet access service.”


“Unreasonable discrimination of network traffic could take the form of particular services or websites appearing slower or degraded in quality.”

With all due respect, the way Hastings characterizes what now is possible--”legally impeding video streams”--is in no way accurate or truthful.

The irony is that Netflix runs its own content delivery network!

Wi-Fi Offload Means "Don't Use Our Product"

What does one make of a business that encourages people not to use the key growth product the industry sells?

Perhaps the key growth product has a huge business model problem: if people really used it, the supplier couldn’t keep up with demand, and customers would become massively dissatisfied with both prices and experience.

Some might say that is precisely why mobile service providers encourage Wi-Fi offload: neither networks nor business models would survive actual usage of the product.

Consider the growing consumption of video on mobile networks. At some point, adding more capacity, in the form of spectrum, does not work. To be sure, mobile operators have other tools, including better air interfaces and more-intensive coding and small cell architectures.

Still, current mobile operator behavior continues to suggest there is a huge strategic issue here: the high-margin apps, such as messaging and voice, are being displaced by medium-margin Web browsing and transaction apps and a huge amount of low-margin video apps.

And there is almost a direct inverse correlation between bandwidth consumption and profit margin.



The problem actually is intensified as mobile networks introduce next generation networks. One constant in the mobile or fixed Internet access business is that when a faster network is made available, users consume more data.

Most observers would likely agree that the reason is better user experience. When content can be retrieved more quickly, people spend more time interacting with apps. That is the reason Google is so focused on encouraging faster Internet access.

And a new study conducted by JDSU’s Location Intelligence Business Unit suggests not only that LTE encourages more data consumption, but also dramatically increases the consumption of the heaviest users. Overall, heavy LTE users consume 10 times more data than similar heavy 3G users.

iPhone 5s users demand seven times as much data as the benchmark iPhone 3G users in developed markets, and twenty times as much data in developing markets, the study suggests.

Apple iPhone 5s users demand 20 percent more data than iPhone 5 users  in developed markets and 50 percent more data in developing markets, the study suggests.

Samsung Galaxy S4 users generate five times as much data as iPhone 3G users in developed markets, and eleven times as much data in developing markets.
The extreme one percent of all users consume over half of the downlink data in both developing and developed markets.
LTE users are ten-times more extreme: 0.1 percent of all users consume over half of
the LTE downlink data.

In 2011, one percent of 3G users consumed half of the entire downlink data. In 2012 the findings were the same: one percent of 3G users still consumed about half of the data.

On the 4G LTE networks, 0.1 percent of 4G users consume more than half of the entire LTE downlink data: an order of magnitude smaller group of users.

“The faster the speeds that mobile operators provide, the more consumers swallow it up and demand more,” JDSU says.

Perhaps significantly, the study revealed that users in developed and developing markets had similar overall rankings in terms of smartphone data consumption, but also little use of data cards and dongles or tablet connections.


Thursday, January 23, 2014

South Korea to Build 5G network 1,000 Times Faster Than LTE

Though there is no formal fifth generation network standard, South Korea plans to move ahead anyhow, investing 1.6 trillion won (US$1.49 billion) through 2020 to build a fifth generation network (5G), the communications technology ministry says. 

The 5G network is expected to be 1,000 times faster than the existing Long Term Evolution network.

A trial 5G network is due to be rolled out in 2017, with full commercialization in 2020.

Perhaps one way to look at the effort, aside from Korea's determination to lead mobile network evolution, is that 2G networks were lead by voice, 3G by data (messaging, email, Internet access), 4G possibly by video. Will 5G be lead by the "Internet of Things" or machine-to-machine applications. We'll see. 

The initiative also seems a way for Korean firms to attempt what Huawei has achieved, namely getting a significant share of global telecom infrastructure share. 

Huawei recall, had by 2012 gotten about 26 percent share of global telecom infrastructure sales. Today, Korea mostly serves its domestic market, and has about four percent share of infrastructure sales. 

Wednesday, January 22, 2014

Okay, There are a Few Places Where There Really is No Spectrum Crunch

Is there really a spectrum “crunch?” Actually, the answer is more nuanced than you might think.

Certainly, there are places--such as the South Pacific--where spectrum seems not to be an issue, and is not considered to be a major concern even in the future. But that likely is an exception to the rule.

There are some who argue there really is not generally a problem, though that view seems a distinctively small minority opinion. And "when?" is a key qualifier.



Whether there is a spectrum problem today--in larger nations-- is a different question than whether there will be a problem later, unless multiple approaches are used to supply more spectrum.

Most observers agree that demand for mobile and untethered Internet access is growing fast, and will grow even faster as people consume more full motion video on untethered and mobile  devices.

Perhaps the better question is whether ISPs will need much more spectrum in the future to meet demand, even if they aggressively re-farm existing spectrum, redesign networks and have access to better network gear.

There is widespread agreement that mobile will become the primary way most people use the Internet, perhaps as early as 2017.

And if people start to use mobile Internet connections as they do fixed Internet connections, demand for spectrum will qualitatively change our notions of how much spectrum will be required.

Reasonable people can agree that spectrum ought to be “hoarded” only to deprive potential competitors of the assets. But reasonable people also can agree that present trends point to a clear need for more spectrum.




source: KPCB

U.S. Fixed Line Market Illustrates Mobile Importance

If you believe this access line forecast by JSI Capital Advisors is close to accurate, you will grasp the importance of mobile and wireless access technologies and revenue streams. 

Though up to this point firms such as Verizon Communications and AT&T, Comcast and Cox Communications have been able to maintain aggregate fixed line revenues by selling more services to a smaller number of accounts, one would be on safe ground in arguing that continued declines will, at some point, undermine that strategy.



The other observation is the important role now played by cable TV operators in the Internet access and voice businesses. Once upon a time it might have been thought new "telco" providers would be the big winners from the Telecommunications Act of 1996, which legalized local access competition.

In the early going, in fact, it was AT&T and MCI, then the dominant long distance companies, that represented most of the competitive local exchange carrier industry's leaders. Since then, both AT&T as an independent long distance company, and MCI, have disappeared, absorbed into other firms (AT&T was bought by SBC, which rebranded itself), while Verizon acquired the MCI assets. 

In voice or Internet access segments of the local access market, cable TV operators are the single most important category of "new" providers. 

The Internet in 1969

Embedded image permalinkSometimes it is helpful to remember just how far one has come. 

Mobile communications was a concept, before it became a commercial service. 

In 1985, when the CTIA  first began keeping statistics, there were about 200,000 mobile phone customers in the whole United States.

Consider mobile adoption rates now. 

Earlier, the big problem had been access to "telephone service" at all. In 1996, phone phone penetration was about 1.9 percent in Africa, for example. 

Today, mobile adoption in Africa is at least 67 percent. 

Consider the size of the Internet in 1969: four nodes. Today, two billion people, at minimum, use the Internet. 

To twist a commonplace phrase, "progress happens."

Embedded image permalink


Many Winners and Losers from Generative AI

Perhaps there is no contradiction between low historical total factor annual productivity gains and high expected generative artificial inte...