The Migration to an All-IP Network

Last month the FCC recommended that carriers adopt a number of security measures to help block against hacking in the SS7 Signaling System 7). Anybody with telephone network experience is familiar with the SS7 network. It has provided a second communication path that has been used to improve call routing and to implement the various calling features such as caller ID.

Last year it became public that the SS7 network has some serious vulnerabilities. In Germany hackers were able to use the SS7 network to connect to and empty bank accounts. Those specific flaws have been addressed, but security experts look at the old technology and realize that it’s open to attack in numerous ways.

It’s interesting to see the FCC make this recommendation because there was a time when it looked like SS7 would be retired and replaced. I remember reading articles over a decade ago that forecast the pending end of SS7. At that time everybody thought that our legacy telephone network was going to be quickly migrated to all-IP network and that older technologies like SS7 and TDM would retired from the telecom network.

This big push to convert to an IP voice network was referred by the FCC as the IP transition. The original goal of the transition was to replace the nationwide networks that connect voice providers. This nationwide network is referred to as the interconnection network and every telco, CLEC and cable company that is in the voice business is connected to it.

But somewhere along the line AT&T and Verizon high-jacked the IP transition. All of a sudden the transition was talking about converting last-mile TDM networks to digital. Verizon and AT&T want to tear down rural copper and largely replace it with cellular. This was not the intention of the original FCC plans. The agency wanted to require an orderly transition of the interconnection network, not the last-mile customer network. The idea was to design a new network that would better support an all-digital world while also still connecting to older legacy copper networks until they die a natural economic life. As an interesting side note, the same FCC has poured billions into extending the life of copper networks through the CAF II program.

Discussions about upgrading connections between carriers to IP fizzled out. The original FCC vision was to take a few years to study the best path to an all-IP interconnection network and then require telcos to move from the old TDM networks.

I recently had a client who wanted to establish an IP connection with one of the big legacy telcos. I know of some places where this is being done. The telco told my client that they still require interface using TDM, something that surprised my client. This particular big telco was not yet ready to accept IP trunking connections.

I’ve also noticed that the costs for my clients to buy connections into the SS7 network have climbed over the past few years. That’s really odd when you consider that these are old networks and the core technology is decades old. These networks have been fully depreciated for many years and the idea that the cost to use SS7 is climbing is absurd. This harkens back to paying $700 per month for a T1, something that sadly still exists in a few markets.

When the FCC first mentioned the IP transition I would have fully expected that TDM between carriers would have been long gone by now. And with that would have gone SS7. SS7 will still be around in the last-mile network and at the enterprise level since it’s built into the features used by telcos and in the older telephone systems owned by many businesses. The expectation from those articles a decade ago was that SS7 and other TDM-based technologies would slowly fizzle as older products were removed from the market. An IP-based telecom network is far more efficient and cost effective and eventually all telecom will be IP-based.

So I am a bit puzzled about what happened to the IP transition. I’m sure it’s still being talked about by policy-makers at the FCC, but the topic has publicly disappeared. Is this ever going to happen or will the FCC be happy to let the current interconnection network limp along in an IP world?

Verizon Phasing Out Copper Services

fios vanVerizon has asked the FCC for permission to discontinue a number of legacy copper-based products in the Northeast. This signals the company’s ongoing plan to pull down copper wires and go to an all-digital network.

Specifically Verizon wants to eliminate their Voice Grade, WATS Access Line, Bonded Digital Link, Digital Data, and DIGIPATH Digital Service II. These are somewhat obscure services, mostly used by businesses, and which for the most part have been supplanted by better products over the years.

What this filing doesn’t specifically say is that Verizon will eventually accompany this tariff change by a request to remove their copper network. That’s what they did earlier this year in New Jersey.

I find it hard to criticize the company for wanting to move customers from copper to fiber. I have a lot of small telco clients who have done the same thing over the last decade. There are a few customers that worry about such a transition because they have some legacy function like fax machines, health monitors, burglar alarms or T1s that they are afraid won’t work with the updated technology. For the most part there are not very many such applications around that can’t be made to work on fiber. Fiber technology can emulate almost every TDM copper-based function.

There comes a point where it doesn’t make economic sense to maintain an old copper network for a tiny handful of customers using old applications. I have a hard time thinking that customers have a right to stay on copper when there is something better available.

But I also think the public unease over these transitions is because the public doesn’t trust Verizon. Verizon got a lot of bad press after hurricane Sandy hit Fire Island and the company wanted to replace the destroyed copper with cellular service.

The problem is that Verizon doesn’t have fiber everywhere –not even close to everywhere. What happens where there is no fiber availability? When Verizon built FiOS they only built fiber where the costs to do so were low, and this resulted in a patchwork fiber network – where one street or one subdivision has fiber, but the next doesn’t. The company also largely built fiber in the suburbs of the major cities and they largely ignored downtown urban neighborhoods as well as smaller towns and all rural areas.

AT&T is being open about its plans to move homes to a fixed cellular connection. But as Verizon starts pulling down more copper they are either going to have to build new fiber to people or offer the same kind of cellular connections as AT&T. And it doesn’t seem likely that Verizon is going to extend FiOS fiber networks today to neighborhoods they judged too expensive to build fifteen years ago.

Verizon’s union members have been complaining for years that the company has been neglecting the copper plant – and these technicians are right. It’s a behavior we have seen from all of the large telcos for decades. Twenty years ago Verizon started trying to find a buyer for their network in West Virginia. It took them more than a decade to finally sell it to Frontier, and during the interim they cut maintenance to the bone. But this is not a singular example and huge parts of the Verizon, AT&T and CenturyLink networks are in bad shape due to many years of neglect.

The shift away from copper is inevitable. A lot of these networks were built in the post WWII decades and they have lasted longer than intended. It’s a testament to the high standards of the old Ma Bell system that these networks are still working today. Critics of Verizon want the company to maintain the copper networks for a few more decades – but that is unrealistic, and in many cases becoming technically impossible, and it’s time to shift focus to make sure Verizon doesn’t start quietly dropping homes and leaving them stranded with no communications options.

Special Access Rate Investigation

FCC_New_LogoThere is an investigation going on at the FCC that is probably long overdue involving looking at special access rates. Special access rates are rates used by telephone companies to charge for TDM data circuits such as T1s.

You might think that T1s and TDM technology would be fading away, but the large telcos are still making a fortune by requiring other carriers and large businesses to interface with them using TDM circuits and then charging a lot of money for the connections. As an example, the connections between a large telco like AT&T and CLECs or long distance carriers are still likely to be comprised of DS-3s (28 T1s).

There are also still a lot of businesses that use T1s. There are still a lot of older phone systems sitting at small businesses that need a T1 interface to connect back to the phone company. And in very rural markets where there is no last mile fiber the telcos are still selling T1 data connections to businesses and delivering the paltry 1.544 Mbps of data that it can deliver.

The main thrust of the investigation are the prices being charged. In places with no competition the telcos might still charge between $400 and $700 per month for a T1 connection. And it’s not unusual for carriers to have to pay thousands of dollars per month to interface with the large carriers at a regional tandem switch.

There was a time when the prices charged for TDM circuits were somewhat cost-based, even though as somebody who did some of the cost studies behind the rates, I can tell you that every trick in the books was used to justify the rates at the highest possible cost. But in a 100% copper network there was some logic behind the charges. For example, if a business bought a T1 the phone company had to dedicate two copper pairs throughout the network for that service plus provide fairly costly electronics to supply the T1. I remember when T1s first hit the market and they were a big step forward in telco technology.

But technology has obviously moved forward and we now live in an Ethernet world. The FCC has been working on a transition of the public switched telephone network from TDM to all-IP and this investigation is part of that process.

The prices for TDM special access are all included in tariffs, and for the most part the rates have not changed in years, or even in decades. Where it used to be a big deal, for example, for a telco to send a DS3 between its offices, the bandwidth on a DS3, at 45 Mbps, is barely a blip today inside of the huge Ethernet data pipes the phone companies use to connect their locations. Even if the cost for a DS3 was justified at some point in time, when those same circuits are carried today inside much large data pipes the cost of transporting the data has dropped immensely.

It’s good that the FCC is investigating this, but to a large degree it’s their fault that the rates are so high. It’s been decades now since either the FCC or the state regulatory commissions required cost studies for TDM circuits. And without the prodding by the regulatory agencies the telcos have all let the old rates stand in place and have been happily billing them year after year. This investigation should have been done sometime soon after the Telecommunications Act of 1996, because the rise of competitive telecom companies created a boom in special access sales, all at inflated prices.

Special access rates matter a lot to small carriers. For example,special access is one of the largest expenses for any company that wants to provide voice services. It’s not unusual for a company to spend $100,000 or more per year buying special access services even if they deliver only a tiny volume of voice traffic to the world. As would be expected, the high costs adversely affect small carriers to a much greater extent than large carriers who can better fill up the pipe between them and the large telco.

For years the telcos have hidden behind the fact that these rates are in a tariff, meaning that they are not negotiable for other carriers. But at the same time, the telcos routinely drop rates significantly when selling special access circuits in a competitive market. The high special access rates apply only to those small carriers and businesses who are too small to negotiate or who do not operate in a competitive part of the network. It’s definitely time for these rates to be brought in line with today’s costs, which are a small fraction of what the telcos are charging. It would not be shocking for the FCC to determine that special access rates are 70% to 90% too high, particularly when you consider that most of the network and electronics that support them have been fully depreciated for years.

AT&T’s IP Transition

telephone cablesA few weeks ago I talked about how Verizon and AT&T are using the FCC’s IP Transition to try to get out of serving regulated services on copper landlines. Today I want to talk more about AT&T. Earlier this month they met with the FCC staff to talk about their ideas on the IP transition and they followed up that meeting with this memo.

The FCC’s IP Transition looks at replacing the PSTN (Public Switched Telephone Network). This is the complex network that has been used to carry voice traffic and that assures that every call attempted ought to be completed. It’s now an old network, is separate from the Internet and it still mostly uses time division multiplexing technology (TDM) that is based upon using circuits that are some multiple of T1s.

The PSTN has served the country well, but pure IP technology is a lot more efficient and the FCC is working towards replacing the PSTN with something new and IP-based. The PSTN is a series of connections, called trunks that connect the central offices switches of all of the carriers in the country along with the electronics that control the network. It’s important to note that the PSTN is only the network between carriers and does not involve any connections to customers. The PSTN has always been technology agnostic in terms of supporting any kind of network such as copper, coaxial or wireless and in allowing any kind of phone or customer device as long as a carrier can locally support it.

But the PSTN network is more than just the wires connecting carriers because it includes things like the SS7 network that is used along with each call to transmit the calling number and other information. And the PSTN comes with a number of specific regulatory requirements that define the ways that carriers of different types can interconnect with each other. The FCC’s major role in this process is to rework all of the rules that define how carriers interact in the new IP world. So the FCC is being careful in dismantling the PSTN because if it’s done incorrectly there could be chaos between carriers and even problems in completing calls. Before they order a mass migration from the PSTN the FCC has authorized a number of trials to convert parts of the PSTN to IP to look in detail to make sure that everything works as hoped.

But AT&T and Verizon have hijacked the IP Transition and persist on using it as an excuse for replacing copper connections with something else. In the case of AT&T they talk about wanting to replace millions of home phones on copper with cellular. The FCC’s IP Transition never intended to require or be associated with changing technologies used by customers. The FCC must be getting very frustrated to see AT&T and Verizon continuously blame them for the changes they are trying to foist on customers. I’m actually somewhat surprised that the FCC hasn’t told them to knock it off. If I was the FCC I would be telling customers that it is not my intention to kick people off of copper. There is no technical or regulatory reason that copper networks can’t work with an IP version of the PSTN.

As you can see by this memo, AT&T intends to kick people off copper in several communities as part of what they call Technology Trials. But they don’t want to say ahead of time where those communities are because they know full-well they will be met with a lot of resistance. It’s funny that AT&T says they don’t want to divulge where their trials will be done due to fear of how competitors will act. The only competitors that benefits from AT&T’s plan will be the cable company in each town which will pick up most of the abandoned customers along with AT&T’s own cellular business. It’s incredibly unlikely that the cable companies are going to find any problem with AT&T’s plans or that releasing the information early would somehow give the cable companies some kind of edge.

AT&T has some really great writers and their memos always sound very logical and well thought out. This memo certainly seems reasonable if one didn’t understand what they are actually talking about doing. They want to knock people off copper, wait until the last possible minute to announce who that will be, and then blame it on an FCC as part of the IP Transition. AT&T will largely be forcing customers to the cable companies if they want landline voice or data.

One might not think this is all that bad of a thing. After all, the copper is getting old and perhaps it is time for it to go. But a large part of the reason rural copper is so bad is from years of neglect. One might not feel so bad about people living in small towns who end up having to go to the cable company if AT&T bails on a town. But you have to realize that small-town cable networks are sometimes in worse condition than the copper. And in most places, if AT&T shuts down the copper then the cable company becomes the only game in town.  One really has to feel bad for the people who live outside rural towns, outside the reach of the cable companies. They are going to lose the only wire to their homes.