The Future of Interconnection

A Verizon payphone with the Bell logo.

A Verizon payphone with the Bell logo. (Photo credit: Wikipedia)

AT&T and Public Knowledge both testified yesterday at a House Communications Subcommittee hearing about the transition of today’s PSTN to an all-IP network.

Both parties agreed that there were five areas that must be addressed to maintain a functional telephone network:

  • Service for everybody
  • Interconnection and competition
  • Consumer protection
  • Reliability
  • Public Safety

I want to look a little more at the issue of interconnection and competition. Today a large percentage of my clients have interconnection agreements with the incumbent telephone companies. Most of my clients are CLECs but a few are wireless carriers, and each negotiates interconnection under a different set of FCC rules.

Interconnection is vital to maintain competition. Interconnection basically covers the rules that define how voice traffic gets from one network to another. The agreements are very specific and each agreement defines precisely how the carriers will interconnect their networks and who will pay for each part of the network.

For the most part, the rules of Interconnection adopted as part of the Telecommunications Act of 1996 work well and there are probably over 2,000 companies using these agreements to interconnect with each other.

There is a lot of danger that changing the interconnection rules could harm and force competitive companies out of the market. Let me just revisit a little bit of history to talk about what I mean. A long time ago the FCC decided that interconnection for local calls between incumbents should be free, and so incumbent telephone companies don’t charge each other to exchange local minutes. However, I can think of at least five times during my career when the RBOCs like AT&T tried to put in reciprocal charges for this traffic. That means that both parties would pay each other the same amount for terminating local calls from the other. Sounds okay until you recall that AT&T basically serves all of the metro areas in the country while smaller telcos serve the rural areas. Still today there is a lot more calling made from rural areas into metros than in the other direction, and if such a change was made the rural companies would be sending big checks to the RBOCs for ‘free’ calls

And the RBOCs have tried to do similar things to competitive carriers with interconnection. The FCC’s interconnection rules say that a competitive carrier can choose to interconnect with a larger company at ‘any technically feasible point’, and yet every few years the RBOCs try to change interconnection agreements to force carriers to carry the traffic to the RBOC hubs. Again, this is a matter of money and the RBOCs want the competitive carriers to pay for everything.

Changing to an all-IP network is likely to open up the same battles. Rather than maintain a system today of many tandem offices in a state, it is not impossible that the RBOCs will have only one hub in each state, or even only one hub in each region of many states. And if they make that kind of change you can expect that they will then expect competitive carriers to pay to carry all if their traffic to and from such hubs. I can tell you that such a change would devastate the business plan of many competitive carriers and would greatly reduce competition in the country.

The FCC has to be diligent in making the changes to IP. Everybody agrees that the technological change needs to be made. It’s more efficient. But we can’t let a technology change be grounds for a land-grab by AT&T and Verizon in an attempt to quash competition. They will, of course, claim that they are not trying to do that, but during my 35-year career I have seen them try exactly that kind of change a whole lot of times. And there is no doubt in my mind they will try to do it again.

Switching in an IP Environment

FCC HQ

FCC HQ (Photo credit: Wikipedia)

In this industry there are always interesting fights going on behind the scenes. In fact, it seems like a lot of the policies made by the FCC are in response to battles being waged between carriers. As the FCC intervenes in these fights they end up creating policy as they help solve issues.

This Letter is a correspondence with the FCC about a current dispute that is going on with Verizon and AT&T disputing the way they are being billed by Bandwidth.com. and Level3. This fight is an interesting one because it asks the FCC to affirm that is supports a migration to an all-IP network.

The dispute is over what is called OTT (Over-the-top) VoIP. OTT in this case means that there are voice calls being made from a service provider’s network for which the service provider is not providing the switching. Instead the service provider is buying switching from a CLEC like Level3. And all of the calls involved are VoIP calls, meaning that they are being delivered from the customers to the switching CLEC using the IP network rather than the public switched telephone network.

Here is how this might happen, although there are other configurations as well. The network in question is clearly an IP network to the customer in order for this to be considered as VoIP. That means it is either a fiber-to-the-home network, DSL over a copper network or a cable system that has been upgraded to send the voice over the data path. In a traditional TDM network the calls from customers are routed directly to a voice switch and that switch will decide what to do with the call based upon the numbers that were dialed. But in this scenario there is not a switch in the subscriber’s network. Instead, when a customer makes a call, a signal is sent to wherever the switch is located telling it where the customer wants to call. That remote voice switch then tells the network owner where to send the call. It is no longer necessary in a smartswitch environment for the call to actually touch the switch, but the switch is still the device that decides how to route the call.

The parties are fighting about whether access charges ought to be charged for an OTT VoIP call. Access charges are fees that long distance carriers pay at both the originating and terminating end of a call to compensate the network owner at each end for processing the call. Verizon and AT&T don’t want to pay the switching component of the access charges for these calls. They are arguing that since there is not a physical switch in the originating network that such charges aren’t warranted.

Broadband.com and Level3 are arguing that the switching is being performed regardless of the location of that switch. They point out that for the FCC to rule otherwise would be counter to the FCC’s desire for the telephony world to migrate to an all-IP environment.

If the FCC rules that AT&T and Verizon are right, they will be saying that a carrier performing a switching function on legacy TDM technology can bill for performing that function but that somebody doing it more efficiently in an IP environment cannot. I just published a blog yesterday talking about ways to share a softswitch and that is exactly what is happening in this case. In an all-IP environment the network can be more efficient and not every carrier needs to buy and operate a switch. They can instead contract with somebody else to switch calls for them which is easy to make happen in an IP environment. Access charges are designed to compensate local carriers for the cost of performing certain functions and one has to think that the network owner in this case is still having to pay for the switching function and should get to recover some of that cost.

In fact, there has been switch sharing for years even in the TDM world. I know several rural LECS who lease switching from their neighbors and who have not owned a switch for decades, and they have always billed the switching access charge element. That element reimburses you for the cost of switching and it really shouldn’t matter if that cost is made up of the depreciation on a box you paid for or else a fee you pay to use somebody else’s box. Cost is cost and the key fact is that calls can’t be made or received from an area if somebody isn’t doing the switching.

I always find arguments by the large RBOCs to be interesting because they wear many hats. AT&T and Verizon are wireless carriers, LECs and long distance companies, and often when one part of the large companies make regulatory arguments it will be contrary to the interest of one of the other branches of the company. In this case the long distance branches of the RBOCs are looking for a way to avoid paying access charges. But the LEC side of both Verizon and AT&T share switching and they do not have a switch any more for every historic exchange area. So to some degree these companies are arguing against something that another branch of their company is doing. And this is often the case in many regulatory arguments since these companies do so many things.

Hopefully the FCC will agree with Broadband.com and Level3. If they rule otherwise they will be telling carriers that it is not a good idea to establish switch-sharing arrangements that are more efficient than having every carrier buying the same expensive boxes. If the FCC really wants the telco world to move to IP they need to get rid of any regulatory impediments that would make an IP network less desirable than a legacy network. Hopefully the FCC sides with efficiency.

Is Wireless a Substitute for Wireline?

English: A cell phone tower in Palatine, Illin...

Last week in GN Docket 13-5 the FCC issued an update that asked additional questions about its planned transition of the historic TDM telephone network to all-IP network. This docket asked for comments on several topics like having a trial for transitioning the TDM telephone network to all-IP, for having a trial to go to enhanced 911 and for making sure that a switch to IP would not adversely affect the nationwide telephone databases.

But the docket also asks for comments on whether the FCC should grant telephone companies the right to substitute wireless phones for wireline phones and abandon their copper network. The docket mentioned two companies that wanted to do this. For example, Verizon said they intend to put wireless on Fire Island off New York City as they rebuild it from the devastation of hurricane Sandy. But AT&T has told the FCC that they are going to request permission to replace “millions of current wireline customers, mostly in rural areas, with a wireless-only product”.

Let me explain what this means. There are now traditional-looking telephone sets that include a cellular receiver. To replace a wireline phone, the telephone company would cut the copper wires, and in place of your existing phones they would put one of these cellular handsets. They would not be making every family member get a cell phone and there would still be a telephone in the house that works on the cellular network.

This make good sense to me for Fire Island. It is mostly a summer resort and there are not many residents there in the winter. It’s a relatively small place and with one or two cell phone towers the whole island could have very good coverage. And if the cell phone tower is upgraded to 4G there would be pretty decent Internet speeds available, certainly much faster than DSL. One would have to also believe that the vast majority of visitors to the island bring along a cell phone when they visit and that there is not a giant demand for fixed phones any longer.

It is AT&T’s intentions, though, that bother me a lot. AT&T wants to go into the rural areas it serves and cut the copper and instead put in these same cellular-based phones. This is an entirely different situation than Fire Island.

Anybody who has spent time in rural areas like I do knows the frustration of walking around trying to find one bar of cellular service to make or receive a call. Cell phone coverage is so good today in urban areas that one forgets that this is not true in many places. I have a client, a consortium of towns and the rural areas of Sibley and Renville Counties in Minnesota. Let me talk about my experience in working with them as an example of why this is a bad idea.

My primary contact works in the small town of Winthrop. I have AT&T cellular service and when I visit him my cellphone basically will not work. I sometimes can move around and find one bar and get a call through, but I can’t coax the phone to get a data connection so that I can check email. And if you go west from Winthrop the coverage gets even worse. AT&T’s coverage maps show that they serve this area, but they really don’t. There are places in the east end of Sibley County that have decent coverage. But there are also plenty of farms where you can get coverage outdoors, but you can’t get coverage in the house.

The traditional cellular network was not built to serve people, but rather cars. Cell phone coverage is so ubiquitous now that we already forget that cellular minutes used to be very expensive, particularly when you roamed away from your home area. The cell phone network was mostly built along roads to take advantage of that roaming revenue stream. If you happen to live near to a tower you have pretty decent coverage. But you only need to go a few miles off the main highway to find zero bars.

And I use the Renville / Sibley County client as an example for a second reason. The people there want fiber – badly. They have been working on a plan for several years to get fiber to everybody in the area. The area is a typical farming community with small hub towns surrounded by farms. The towns have older cable systems and DSL and get broadband, although much slower than is available in the Twin Cities an hour to the east. But you don’t have to go very far outside of a town to get to where there is no broadband. Many people have tried satellite and found it too expensive and too slow. There are any homes still using dial-up, and this is not nearly as good as the dial-up most of you probably remember. This is dial-up delivered to farms on old long copper pairs. And it is to get access to an Internet that has migrated to video and heavy graphics. Dial-up is practically useless for anything other than reading email, as long as you don’t send or receive attachments.

Over 60% of the people in the rural areas in Renville and Sibley Counties have signed pledge cards to say that they would take service if fiber can be built to them. One would expect this would translate to at least a 70% penetration if fiber is built. They refer to the project locally as fiber-to-the farm. There has been a cooperative formed to look at ways to get fiber financed. And any financing is going to require local equity, meaning the people in the County are going to have to invest millions of their own dollars in the project – and they are certain they can raise that money. That is how much they want the fiber. And this same thing is true in rural areas all over the country. Most of rural America has been left behind and does not have the same access to the Internet that the rest of us take for granted.

AT&T’s idea is only going to work if they make a big investment in new rural cell towers. The current cell phone network in rural areas is not designed to do what they are proposing, even for delivering voice. And even if the existing rural cell phone towers are upgraded to 3G or 4G data (which almost none have been), most people live too far from the existing towers to get any practical use from cellular data. Cellular data speeds are a function of how close one is to the tower and, just like with DSL, the speeds drop off quickly as you get away from the hub.

I hope rural America notices this action at the FCC and files comments. Because as crappy as the rural copper wires are today, when the wireline network disappears many rural households are going to find themselves without telephone service. And forget about fast rural data. The AT&T plan is really just a plan for them to abandon and stop investing in rural communities.