Sunday, February 3, 2019

Will 5G Change Fixed Network Business Model?

Do industry executives increasingly believe 5G will change the economics of the fixed networks business? This diagram suggests they increasingly believe that is possible. As 5G networks are built, today’s fixed network access (copper, fiber to home or hybrid fiber coax) will be faced with new competition from fixed wireless and mobile wireless alternatives.

That will happen as deep fiber architectures enable use of lower-power microcells and small cells with access to an order of magnitude more spectrum, in addition to better radios and using the frequency reuse principle (cell site division) to historically-dense levels.



By some estimates, 4G has lead to cell tower density of perhaps 2 kilometer (1.25 miles) spacing. Some believe the 5G network using millimeter wave spectrum will require small cells placed about every one-third of a mile. In other words, 5G using millimeter wave spectrum might require about eight times the number of transmitting sites, compared to 4G using towers spaced at about 2 km (1.25 mile distances).

That is probably a high-end forecast, assuming an equally-dense network in all deployment scenarios  in dense urban and suburban settings. Virtual nobody believes that is possible in rural areas.

Some might note that such densities, while perhaps common in more-dense urban areas, are not so common in suburban settings, and nonexistent in rural areas, for the most part. Also, the simple assumption here is that optical backhaul to one macrocell in point to point fashion also applies to more-dense networks.

In other words, if one site requires a single point to point connection, then 25 small cell sites require 25 total point-to-point connections. But density itself changes the topology, leading to “more tree and branch off a ring” topology that does require lots more fiber, but not as much as point-to-point links would require.

Still, the business model impact on fixed line network operators will possibly be significant. Think of 5G as a new overbuild operation. Where today a cable operator and telco compete for the consumer internet access customer, tomorrow it could well be the case that those two competitors face a third or maybe even fourth competitor for a market that is very close to saturated.

That means market share losses will happen. The degree of business model disruption then turns on the amount of share the new competitors can take, and from which current suppliers.

3G to 4G to 5G: What is Common?

In one sense, it might be easier to envision 5G primarily as a way to supply increased bandwidth for consumer mobility, as one might see 4G as a way to increase bandwidth supply for mobile internet access demand.

Ignore for the moment the lower costs per unit 4G offered over 3G, or the similar benefits 5G will offer over 4G. Ignore for the moment the lower latency 4G offered over 3G, or the lower latency 5G will supply, compared to 4G.

Look only at 5G as a platform for supplying increased bandwidth at lower costs per bit, as was true for 4G as well. So long as there is a transparent fallback from 5G to 4G (as was true of 4G fallback to 3G), and so long as 5G experience and 4G experience are matched closely enough, then 5G can be added more gracefully than some expect.

In fact, one good reason for marketing and supplying bandwidth more incrementally when adding 5G is precisely so the 4G fallback is graceful. In my own experience, the fallback from 200 Mbps or 100 Mbps to 20 Mbps or 14 Mbps is quite graceful, per concurrent user.

In fact, unless there are some latency effects, I typically cannot tell you whether my current connection is running at 14 Mbps, 20 Mbps, 100 Mbps, 150 Mbps or some higher figure. Granted, I do not download big files very often.

My typical use cases range from web surfing to cloud-based communications to streaming video. None of those apps is overwhelmed or unpleasant if my connection is anywhere from 14 Mbps to some other three-digit rate. In other words, if I switch from a cable modem connection to mobile 4G, my experience does not suffer.

The point is that, even if I buy a 5G service with higher headline speeds, defaulting back to 4G is not going to be a problem.

Saturday, February 2, 2019

AIOps Now Shows how AI is Applied to Network Operations

I just got back from chairing the new AIOps Expo, sponsored by TMCnet, a three-day event looking at the advantages and challenges of artificial reality as applied to management of information technology systems (both enterprise and communications service provider).

These days, it always is clear that value and opportunity for connectivity providers largely has shifted up the stack to applications and platforms.

One big question asked by enterprise buyers at the event is how “real” AIOps is, where and how it can be used today, and what the roadmap looks like.


As with all new buzzwords and trends, we have to define what we are talking about. According to Gartner researchers, “AIOps platforms combine big data and machine learning functionality to support all primary IT operations functions through the scalable ingestion and analysis of the ever-increasing volume, variety and velocity of data generated by IT.”

Putting the AIOps moniker into context, one might argue it is the latest way to describe the use of artificial intelligence (machine learning and deep learning more than neural networking at this point) to improve IT operations.


You can get an argument about where “automation” ends and “AI” begins; that was clear from discussions at the event. You will not get much argument that, at the present state of the art, it is much easier to apply AI to specific functions and processes than to integrate and correlate all functions and processes.

Compared to a decade ago, when the ability to analyze “big data” was the buzzword, insight still is the desired outcome. What seems different now is that Moore’s Law makes a difference.

Analysis that might have been just as valuable 10 years ago--or two decades ago--now is feasible because the costs of computation and storage as so much less.

But one new emphasis is on machine learning: allowing AI to work autonomously--with human approval--to modify system behavior based on what has been learned, without human intervention.

One concrete difference is the role of scripting and code writing. Compared to present practice, the goal is to allow machines to modify their own behavior without direct coding labor. That obviously raises clear issues about bias in the coding systems and security and privacy issues.

But a big strategic change is the shift to allowing machines to discover patterns that would be prohibitively expensive we were to attempt to discover patterns using human agents only.

One illustration of the potential benefit can be glimpsed if you ask why network operations centers have so many screens, as technologist Frank Yue, KEMP Technologies solution architect. The reason is that each subsystem has its own management and monitoring system.

That means fault isolation is more complex than it would be if all systems were correlated, if all the data could be analyzed and understood in ways that reduce the total number of alarms, for example. The cascading alerts NOCs have to deal with is itself a problem, as many of the alerts from different systems actually refer to common events, noted Bhanu Singh, OpsRamp VP.

Frank Yue, Kemp Technologies

In fact, as noted by Taly Dunevich, Ayehu global VP, AIOps is, in many ways, AIOps is the latest way to automate IT processes, without scripting or coding.

All of which lead Wayne Parker, Northrop Grumman technical advisor to quip that “in 10 years we’ll be calling it something else.” Indeed.


The Business Case for FTTH in the 5G Era

Could U.S. fixed network service providers--other than cable operators--make higher profits if they chose to dramatically increase investment in their access networks? Some might argue the answer is “yes,” while others will maintain the answer is largely “no.”

The essential argument is that telcos could deploy more fiber and thereby grow revenues, in addition to providing better customer experience. Naysayers might point to several reasons why that is likely untrue.

First, U.S. cable operators already lead in market share for internet access, have been adding more than 100 percent of net new additions for years--no matter what telcos have done--and already know how they will migrate the platform from gigabit to 10 gigabit capabilities, at costs far lower than the telcos can manage.

In that, that has been true in the entire internet access era: cable operator simply have been able to upgrade internet access speeds faster, and at much lower cost, than telcos have been able to do. Nor is there any particular reason to think that will change, in the fixed network arena.

That actually has been a key and ever-present issue in the competitive era.

Also, total fixed network revenues (telco and cable) are flattish to falling, as demand shifts to products that do not actually require use of the fixed network at all. Voice and linear video revenues are falling in both cable and telco segments of the fixed network business.

And there is every reason to believe that mobile substitution now is going to make itself felt in internet access and video as it has in voice.

Also, alternatives seem to keep popping up, ranging from overbuilders such as Google Fiber and other independent ISPs to coming use of low earth orbit satellite constellations and possibly other unorthodox platforms.

All of that means the potential return from a major investment in optical fiber access by a telco is getting harder, not easier. It will not help that mobile access now appears poised to erode yet more share from the fixed networks.

Verizon total wireline revenues, for example,  decreased 3.2 percent year over year in fourth-quarter 2018 and 3.0 percent for the full year, compared with 2017, to $7.4 billion and $29.8 billion, respectively.

AT&T fixed network revenues in the fourth quarter of 2018 likewise were down. That stagnating trend has been true for both firms, in the fixed network segment, for some years.
So to the argument that telcos could make more money if they invested more, once must look at Verizon, which already has heavily moved to fiber-to-home facilities. It has not helped much, apparently, though losses would have been greater had it not done so.
AT&T has a far-smaller optical access footprint, and might arguably benefit, in some of its markets, by upgrading to optical fiber. The issue is how much that would help. AT&T cannot expect much revenue lift from video, as it already is the largest U.S. linear video distributor. It is losing fixed network voice lines, as are all providers.
So AT&T would have to risk the value of the optical fiber upgrade, in the consumer segment, on internet access alone. Most who have looked at that model would agree it is likely not going to produce a positive financial result. The incremental gain in internet access share alone does not justify fiber to home investment on a ubiquitous scale.
Some of us believe the financial return from new optical deployments is not from fiber-to-home, but from from deep fiber to support 5G and business customers. That is the new twist as we move into the 5G era. At least in the U.S. market, the financial upside from FTTH increasingly gets worse as use of the fixed network by consumers declines, and as cable operators dominate what remains of revenue and market share in that market.

A new era is coming.

Why AT&T Changed its Thinking about 5G Fixed Wireless

Over the last couple of years, AT&T has seemingly waned and waxed about commercial upside for 5G fixed wireless. For most of 2018, AT&T officials had expressed skepticism about the business model.

That thinking might have been based in large part on use of millimeter wave frequencies (28 GHz, 39 GHz) AT&T owns.

But the ability to use spectrum in the 3.6-GHz region (Citizens Broadband Radio Service) should require less-dense cells and backhaul, perhaps four to 16 times less dense.

Also, some of us would argue the business case for 5G increasingly looks like an enterprise play. Sure, consumers eventually will upgrade to 5G. But that is mostly substitution of one source for another, and does not necessarily increase total revenue much.

On the other hand, there long has been thinking that the latest generations of optical fiber for communications (NG-PON 2, for example) mostly are useful for business applications and mobile backhaul, not consumer mobility.

Combine that with the need for optical fiber densification and the logic is that 5G produces incremental new revenues in the business customer segment and provides internal value for 5G backhaul. Consumer revenues are mostly an afterthought.

All that might have brought a change of views about use of fixed wireless.

“I will say over time three to five year time horizon unequivocally 5G will serve as a broadband, a fixed broadband replacement product,” says AT&T CEO Randall Stephenson. “I am very convinced that that will be the case.”

“You know, we back in the 90s everybody was saying wireless would never serve as a substitute for fixed line voice because there wasn't sufficient capacity,” Stephenson said. “Well it is a substitute for voice.”

Right now, AT&T says it has 11 million fiber to home locations and eight million business locations. AT&T also expects to reach 14 million consumer fiber-to-home locations soon. It probably is worth noting that AT&T’s fixed network passes--is able to sell services to--as many as 62 million U.S. homes.

In other words, AT&T might soon pass 22.5 percent of its consumer locations with optical fiber drops.

Even without quantifying the matter, if AT&T has managed to build optical fiber to less than a quarter of its U.S. homes, and also believes 5G will provide a workable substitute within three to five years, it is hard to see the logic of continuing to build consumer optical fiber connections, at a time when consumer fixed line accounts are shrinking overall.

Phoenix Center Analysis is an Issue in Court Review of Net Neturality

Not often does a single academic paper become a material issue when a major communications policy is adopted or reviewed by courts. But that seems to be the case for an analysis of capital investment levels written by George Ford, chief economist at the Phoenix Center.

The U.S. Court of Appeals for the District of Columbia Circuit is about to hear a challenge to the Federal Communications Commission’s 2018 Restoring Internet Freedom Order. The court review may turn on the FCC’s use of Ford’s analysis of the impact of Title II common carrier regulation on U.S. internet service provider capital investment.

Ford's analysis and the latest defense of his findings on capital investment in the wake of the common carrier classification seem once again to be a part of the policy review.

How Much New Connectivity Spending Does Consumer IoT Generate?

The perhaps common belief is that the growing use of connected sensors and devices is going to drive additional connectivity revenues, and that is true to some extent. But it is likely that most of the connections will not create additional connectivity account demand.

Consider consumer and household use of connected appliances and devices. Today, most smartphone users connect to Wi-Fi when indoors. That does not create demand for new mobile connections in a direct sense.

Likewise, our internet TV devices use Wi-Fi. Our smartwatches use Bluetooth. Most of the other devices, ranging from security cameras to appliance sensors, rely on Wi-Fi.

Only the more “industrial” sensing functions in our homes tend to create demand for additional communications circuits. Gas and power meters provide good examples.

In fact, though the details are likely to change as IoT actually becomes commercialized on a wide scale, though some new mobile connections are created, a fair number of other local connection technology platforms might emerge as significant pathways.

In many cases, those local connections will be to a local edge server most of the time, with external wide area communications on a less-frequent basis. So there is some additional impact on WAN data traffic, but not necessarily a direct increase in the number of connections because of additional IoT sensor or device use.


By some estimates, about 22 percent of IoT spending will be for connectivity. But some estimates suggest connectivity will be far less than that, as a percentage of total spending on IoT.
source: Bain

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...