90% of All Data Generated in the Last 2 Years

You are going to hear, quite often, that “90 percent of world data has been created in the past two years.” It is a evaluation made by IBM and illustrates the dramatic and exponential growth of largely unstructured data, generated by transactions, logs, records, social media, audio, visual and video consumption.

About 80 percent of all of that data is unstructured. Which is why big data and artificial intelligence now have emerged as strategic assets. AI is just about the only way to wring insight out of unstructured datasets so large.

Estimating a retailer’s sales by examining photos of cars parked in lots is one example of past efforts to correlate data. These days, it likely will make more sense to estimate sales by using location data from smartphones.

source: IBM

source: IBM, MBAskool

source: Kleiner Perkins

Why IoT Requires a Cloud-Based, Virtualized Core Network

Nobody yet knows how many internet of things devices will need to be connected by 2020, using mobile and other local networks. Mobile connections, compared to the 2014 level, could be 22 times to 41 times larger. The total number of IoT connections, including devices using other local connections such as Wi-Fi, could be 12 times the 2014 number, or up to 28 times larger.

source: ACG Research

The other clear observation is that use cases will span a rather wide range of network resource requirements, requirements for mobility, latency, signaling and throughput. That is one reason why cloud-native and virtualized packet core networks are deemed essential for IoT supported by 5G networks. There are simply clear use cases that use different combinations of network-provided resources.

source: ACG Research

40 Years of Differences

In January 1978, when the first Pacific Telecommunications Council conference was held, the world was quite different.

• Fewer than 7% of the world’s people had telephone service

• Telecom was a monopoly and most firms were government owned
• Nobody used a mobile phone
• There was no Internet, no Ethernet, no browsers
• 82 analog voice circuits connected Hawaii and Australia/New Zealand
• Modems were acoustic and operated at 300 bps
• Global telecom revenue and profit was driven by voice, especially long distance
• “Billions” of people had never made a phone call
• The business model was simple: build networks, earn a guaranteed return

Now celebrating its 40th anniversary, we all live in a world where:

• Usage has migrated from voice to data to video
• Bandwidth routinely is measured in terabits per second
• There are 7.9 billion mobile phone accounts, used by 4.8 billion people
• Telecom is part of the internet and computing ecosystems
• Most telecom markets are fiercely competitive
• All legacy revenue streams are under pressure, and new revenue models must be created
• Cloud computing, OTT, 5G, smart cities and internet of things are top of mind
• The business model is anything but certain,and every legacy service is mature or soon to be mature

U.S. Ranks 10th for Mobile Internet Speed: Why That is Not a Problem

Less often than in the past, one hears it said that the United States has a broadband problem. Costs are said to be too high, speeds too low, choice inadequate. That is true, in some locations, to some extent.

At times over the past decade, it has been argued that, where it comes to fixed network internet access, the United States was “behind” in either coverage, usage or speed.

The digital divide these days continues to be an issue in rural areas, but arguably is more complicated an issue since some users prefer mobile-only access and some people say they do not use the internet because they do not wish to do so.

Some might also argue that the way people and nations use the internet also matters, not simply availability, price or speed.

International comparisons can be instructive, though sometimes not for the reasons one suspects. Consider voice adoption, where the best the United States ever ranked was about 15th, among nations of the world, for teledensity.

For the most part, nobody really seemed to think that ranking, rather than higher on the list, was a big problem, for several reasons. Coverage always is tougher for continents than for city states or small countries. Also, coverage always is easier for dense urban areas than rural areas. The United States, like some other countries (Canada, Australia, Russia) have vast areas of low population density where infrastructure is very costly.

On virtually any measure of service adoption (voice or fixed network broadband, for example), it will be difficult for a continent-sized market, with huge rural areas and lower density, to reach the very-highest ranks of coverage.

That remains the case for mobile internet coverage or mobile internet average speeds, where, according to Akamai, the United States ranks about 10th.

source: Akamai

Will Edge Computing and Low-Latency Services Allow ISPs to Move Up the Stack?

Most moves made by most tier-one telcos “up the stack” have not worked well, if at all, and that include early moves into computing, data center operations, app stores, appliances and devices, over the top voice and messaging apps, or even OTT video services.

The jury still is out on moves into banking services, mobile advertising and content, but many telcos have fared rather well in the linear video subscription areas.

And though it is a statement of vision more than a practical reality at the moment, AT&T believes that, with a move to pervasive computing (which is one way to describe what “internet of things” is about), there is an inherent ability to embed higher-value operations into the network.

“The network itself moves from a connection to an experience that can include the compute,” said John Donovan, AT&T chief strategy officer.

In other words, even if data warehouses generally have proven to have modest strategic value for access providers (telcos and other access providers), that might well change as services and apps are created that rely on edge computing support.

As a horizontal business model, edge computing support could emerge as an area where telcos and other access providers might actually have some advantages, such as dense networks, access to power, other real estate and network elements that could play a role in supplying edge compute services to third parties.

Consider other potential advantages. AT&T’s new AirGig platform, for example, offers the promise of affordable trunking anywhere above ground where there are power utility poles and transmission lines.

Even if that is not so crucial for urban areas where access providers already have easements, pole attachment rights and access to power, AirGig might well play an important role in rural areas, where the cost of networking and bandwidth has always been tougher.

“For us it's a game changer on a cost basis because the components are small, sample and plastic performance wise,” said Donovan.

In other words, in addition to the “connections” function, there is a logical role either at the applications layer or in the computing layer.

That is not to say the task will be easy. It will be hard. But it is possible, and could prove to be among the more-successful ways telcos can move up the value chain.

source: Telco 2.0

CBRS Needs Certainty, Firms Say

Telecom and all other firms generally hate uncertainty. So a call for keeping in place rules for Citizens Broadband Radio Service, and it's approach to spectrum sharing, is important, industry suppliers say  

https://techpolicycorner.org/fcc-should-maintain-consistent-spectrum-rules-792c58458ca9?ct=t(PR_LabMD_Amicus_January_20171_4_2017)

How to Move Up the Stack, How Not To

Telcos have been trying to “move up the stack” into application layer businesses for quite some time, with very mixed success. Computing firm NCR was acquired by AT&T in 1991, for example, in an effort to create a vertically-integrated computing capability. That effort failed, and NCR was eventually spun off.

That might be one key to how at least some tier-one telcos might look at their moves up the stack. Consider the different way Comcast has used its NBC Universal assets, and how AT&T must use its Time Warner assets.

Comcast did not try to make NBCUniversal (could not, for legal reasons) an “exclusive” or “vertically-integrated” asset available only to its owned cable TV systems. In other words, NBCUniversal was not about vertically integrating the content and making it proprietary to Comcast.

Instead, the value of NBCUniversal content is that it is sold to all other U.S. linear video subscription providers, even if some of that content is used in a proprietary way at the theme parks.

Likewise, AT&T will find its Time Warner content being sold (by law) to all other linear subscription providers, and eventually, in other ways, to over the top services. Likewise, AT&T would have little to no interest in restricting distribution of its studio content (movies) only through AT&T distribution assets. Instead, it would want continued distribution as widely as possible.

The point is that what has worked in the linear video space is not vertical integration, but rather broad sales to direct competitors, who serve customers that want the content.

In other words, instead of vertical integration that seeks uniqueness, content assets are broadly attractive to all suppliers in the linear video business, though also used in a vertical way--as an input--to support AT&T’s own linear video and OTT video operations.

In the internet of things area, a similar approach might be the right way to operate as well. Instead of acquiring or growing assets that are “captive” to AT&T, a better approach might be to create or acquire assets of broad value to customers and competitors.

The other approach--capturing the benefits internally and uniquely within AT&T--might not prove as successful, ultimately.

Most telco VoIP or OTT messaging efforts have failed. One commonality: those efforts were “branded” alternatives to other OTT or VoIP services. In other words, those were attempts to vertically integrate and restrict use of the services only to customers of telco access services.

The opposite approach is taken by wildly-successful consumer apps and appliances such as Google, Facebook, Amazon, Netflix or Apple. Those apps and devices work on all networks, and are not captive to any single access provider.

All that suggests the fruitfulness of seeking assets in IoT that are valuable using any access network, not specific features of a single provider’s access service.

source: Telco 2.0