VOIP services

VOIP Peering: Incumbent Killer?

NEW YORK -- Stealth Communications Inc. is aptly named. The tiny company is 10 years old and has only eight employees. It isn’t widely known outside a small circle of service providers that use its technology.

All the same, it could turn out to be an incumbent killer.

Stealth for years has been developing its Voice Peering Fabric (VPF), which many experts say is now one of the largest VOIP peering points in the world. VPF and other VOIP peering points offer a way for providers of VOIP services to link networks and exchange traffic, routing it around the more expensive PSTN (public switched telephone network) and avoiding the less secure public Internet.

What's interesting is these services are becoming more sophisticated. Yesterday Stealth announced it is adding more features to help VOIP upstarts use VOIP peering points to circumvent the PSTN and the public Internet (see Stealth Intros VOIP Peering).

At telx’s Customer Business Exchange (cbX) conference, Stealth launched its ASP Market, which will add a range of important telecom signaling and directory services, including 411, 8xx services, caller ID, local number portability (LNP), and SS7 (Signaling System 7). Initial partners in the project include SNET Diversified Group, Syniverse Technologies Inc., and VeriSign Inc. (Nasdaq: VRSN).

With more PSTN-like features like caller ID and LNP being built into VOIP peering points, it will become less important for services providers to access the PSTN -- and this threatens the business of many incumbents.

”Established wholesale providers and incumbents such as AT&T and BT really don’t like this; they can become completely disintermediated,” says Caroline Chappell, an independent analyst and author of a recent Light Reading Insider report on service delivery platforms (SDPs). “This could change the whole structure of how traffic is passed around.”

Stealth CEO Shrihari Pandit says the company’s VPF processed 2.5 billion minutes of VOIP in 2004, and he expects it to do 9 billion minutes this year.

”As VOIP starts growing, you want to redeem the risk of going over the public Internet,” says Pandit. “VOIP networks will eventually replace the PSTN, but they need more interconnecting.”

Stealth also operates one of the largest ENUM databases, offered as a free service to its customers (see ENUM Heads for Primetime and Carrier ENUM Gains Ground). The database now holds 6.5 million numbers, says Pandit. He adds that Stealth is profitable and does $10 million to $15 million in revenues per year, charging service providers a flat per-port monthly fee to connect into the peering point.

Skeptics point to Stealth’s tiny size and the fact that VOIP peering really doesn’t make a lot of money. They also point out that VOIP peering points rarely consist of more than a few Ethernet switches cobbled together with some database software, making the barrier to entry low.

But that's precisely what makes the trend scary to the big telcos, and it’s clear the approach is growing -- like a weed. Stealth isn’t the only one working on VOIP peering fabrics, even though it’s one of the biggest. There are many different forms of VOIP peering. Other peering services from InfiniRoute Networks Inc. and Interoute Telecommunications Ltd. cater to different customer bases. For example, Interoute focuses on connecting incumbent TDM networks to VOIP networks, says Chappell.

Yet another threat to incumbent voice providers, notes Chappell, is if large enterprises start using the technology to create their own large VOIP networks.

There are even more VOIP peering services from providers such as Syniverse and VeriSign that involve specialized database, security, and signaling functions. Stealth caters primarly as a traffic exchange for VOIP pure-play operations such as Packet8 and Net2Phone Inc. (Nasdaq: NTOP).

Telecom experts expect these VOIP connection services to proliferate, offering ever more options and ways to exchange traffic and data.

”VOIP peering is a multi-layer proposition,” says Hunter Newby, the chief strategy officer of telx, which hosts connection points for major services providers. “In order to derive the maximum benefit from it, you first need to understand what you want to do with it. With VOIP peering, like many other things, if it makes sense, solves problems, and saves network operators money, then it will succeed."

— R. Scott Raynovich, US Editor, Light Reading

Page 1 / 4   >   >>
OldPOTS 12/5/2012 | 3:10:02 AM
re: VOIP Peering: Incumbent Killer? Scott - It was an accurate chide G

How can you tell if they are making a profit or not?
You must be listening to the Telcos who have to deal with regulators and who can allocate haul profits or costs to their various business divisions/groups. If an expensive Class 5 is part of the local phone business, then are those trunk ports and lines to the Class 4 part of the local phone business? Or are those ports and lines part of the haul network? Heads I win, Tails you loose.

They spent a lot of money to upgrade from unreliable T carrier to fiber DWDM. But to make the network reliable they were spending big bucks on alternate route switching, so upgrading to fiber really timed out right. The fiber was a big cost that was timed to compete and show regulators low profit, but that fiber now monopolizes and keeps traffic growing and the money rolling in. Think long term. Why would VZ spend a $1B or two for ROADM? More capacity and flexible connectivity for that FIOS traffic.

BTW Wiltel was a competitor that figured out how to quickly and cheaply provide long haul fiber. They used GpigsG to pull fiber through their extensive and existing pipeline network. But they also got in other, less profitable businesses and had problems for a while until they moved back to hauling. And now they are friendly telco haul competitors and suppliers.

The Telcos distribute many of the 'long' hauling costs and profits between their various divisions/groups and no one, but they, know what they make at this business (proprietary). I've priced both sides as a buyer and as a supplier and have concluded there is a good profit, but profitability depends how quick and where one depreciates the costs. They are masters at those calculations, trained by the regulators. But I can assure you they track the profit/traffic regularly to measure how much traffic they and their competitors are generating (since they have a near haul monopoly). Then if any new business segment represents a threat they will take action.

Actions explained-
They do Marketing from a high, long term level and try to control the pace of a new technology business development, especially of their competitors. The regulators taught them well. Their first competitive action is Legislative/regulatory to slow competitors and their suppliers progress. This is important so the old investments, with the long ROI (competitive entry barrier), keeps generating cash. Meanwhile they invest in that potential disruptive technology through private startups (used to be BellCore) to monitor technology development and for later use to manage their 'carrier class' suppliers development features and pricing. They generate (see VZ ROADM) RFPs to guide this development and monitor supplier progress. Never buy before at least three RFPs.

By watching the competitor's traffic progress they can decide when it is time to become more actively involved. At a point when it appears that it is possible that the competitors with new technology are starting to generate some traffic and may have a possible working business model. They usually quietly start small trials in the lab (check out their startup investments and suppliers) and then test with small employee groups (FTTU for ~5 years) to primarily understand the business/customer case and secondarily how to adapt the technology to operate in their large networks.

They only announce any activity if they need to lend credibility to a new technology (IPTV), to generate more hauling traffic, steer competitors or satisfy legislators/regulators. At this point they will slow roll a small test business for checking technology credibility, adaptability and to further verify that there is a good business case.

Note they have time on their hands as they always share in the revenue by hauling any new technology traffic. So they roll slow while making money on the old technology. If they really see the business will be profitable, by their standards, they finally start pushing the roller. But they usually don't announce this. In fact it is usually preceded with more legislative/regulatory activity. The attempted Texas legislation and now moving to congress is why I see they perceive triple play as a threat. But they may not do it all with IP. They will slow roll the Voice & TV IP adaptation and let others (competitors and suppliers) make the technology work until there is a business case. Even FIOS doesnGt yet offer IP Voice or any TV. Why, even if it is easy with FTTU.

The cell phone business is an example. This business now contributes significant haul traffic over those leased lines between the (~ Now Telco) cell phone companies. The PSTN and VoIP traffic together will probably generate an equal amount of traffic to cell phone business. Local phone (TDM, DSL or IP) business isn't that much of a profitable business by itself as the existing plant and switches (Class 5) are expensive to maintain. Importantly the Long distance traffic captured by the phones provides the profitable business. Remember cell phone long distance traffic must also be long hauled. They can afford to lay off many local phone service people and reduce this business segment, but would they admit it to regulators and the public? And then there is WiMax on the access horizon.

But the threat of Triple play is significant as most threats are only against one of their business segments, but 3P is against all segments. Actually 4 counting cell phone convergence. But alternatively a slow roll makes sense for Telcos, as the triple play offers a way to upgrade the expensive access plant (gradually to fiber) and share the expense between multiple services/groups. Allocating the cost to whichever market they see fit, including the long haul network. Competitors beware.

But MSOs have already invested much of the cost of entry into this 3P market. They also can share the cost of access between the services/businesses, making them a formable competitor. That is why GǣThere's this interesting little battle going, call it, "If you can deliver phone via cable, then I'm going to deliver TV over the phone line and let's see who cries, "uncle," first."Gǥ Or better yet; I donGt really want to do TV and you donGt really want to invest in all that complicated phone switching equipment (servers). Besides in the last few months we are seeing that customers expect any new services for the same or slightly more than what they already pay for access or they will churn. So where is the profit in this? LetGs call a truce and let others develop those killer apps for our networks and weGll avoid the churn by pursuing our prime businesses.

Enough educational rambling
Do my retirement finances and then
Gone Sailin, the winds are great.
And HeavyDuty - they still won't get it
Keep your day job!
rjmcmahon 12/5/2012 | 3:10:08 AM
re: VOIP Peering: Incumbent Killer? Hauling transport has been a near-disastrous business to be in.

Have there been any exceptions to this? How has Wiltel's hauling business been going?
spelurker 12/5/2012 | 3:10:09 AM
re: VOIP Peering: Incumbent Killer? > HeavyDuty asks:
> I'd be real interested in where you found your
> statistical data.

> alchemy mentions:
> http://www.cabledigitalnews.co...

That is one of my sources, another is the USTA web site. (Though I did make the statement from memory) Note that USTA is a lobbying group, paid to promote the telecom folks, and CDN is a webzine which is by nature a cheerleader for the cable MSOs.
Scott Raynovich 12/5/2012 | 3:10:10 AM
re: VOIP Peering: Incumbent Killer? OldPOTS,

You said:

"And the money just keeps rolling in. Don't kill that cow, hauling network."

I can't tell whether this post is facetious or delusional. Hauling transport has been a near-disastrous business to be in.
HeavyDuty 12/5/2012 | 3:10:10 AM
re: VOIP Peering: Incumbent Killer? "Check your facts. You are confusing many things..."

Re-read the paragraph in question, it's not as confusing as you think.

"the Internet is not run on leased lines"

The Internet is run on leased lines. Ask your ISP what their connection to the peering points are and they'll probably list DS-3s or OC-3s; leased TDM circuits.

"(facility-based) ISPs have their own DSLAMs, and only rent copper and possibly some backbone fiber."

ISPs colocate their DSLAMs in an RBOC's Central Office, then lease cable pairs, from the RBOC, to the customer premesis. Once again, ask your DSL provider.

"So, if the end customer switches from just PSTN to DSL and VoIP, the ILEC loses 50$ voice and gains 10$ for copper rent. If the VoIP is over cable, the ILEC loses everything."

When I say, "old-timers like me have a great future, 'cause the kids don't understand OSI layer 1& 2 connectivity," I'm talking to, and about, you and Stephen...

Thanks for the job security!
jepovic 12/5/2012 | 3:10:13 AM
re: VOIP Peering: Incumbent Killer? "What is lost on most people is that the Internet is run on leased lines that are a subset of the PSTN. If you get DSL (even if you don't get it from the LEC), the service is almost always provided via the LEC's copper pairs."

Check your facts. You are confusing many things here. Leased line is not the same thing as a "copper pair". And the Internet is not run on leased lines except for some business customers, it never have been, but on DSL or cable. Dial-up Internet uses PSTN, but not broadband Internet.

(facility-based) ISPs have their own DSLAMs, and only rent copper and possibly some backbone fiber.

Anyway, a VoIP provider does not pay anything for the undelying connectivity, except for the one port for the SIP server. The end customer uses the same broadband connection.

The broadband provider buys an unbundled copper pair, actually paying for the right to use the non-voice analogue spectrum on the copper. This costs maybe 5-15 dollars per month, much less than the end customer pays for PSTN.

So, if the end customer switches from just PSTN to DSL and VoIP, the ILEC loses 50$ voice and gains 10$ for copper rent. If the VoIP is over cable, the ILEC loses everything.
stephencooke 12/5/2012 | 3:10:21 AM
re: VOIP Peering: Incumbent Killer? Once again, I'm confused... So, which is true...?

"Why is SBC, Verizon, etc... investing in same? Because if some folks insist on getting the latest thang, wouldn't it be best if they could get it without changing vendor; best for the vendor!"


"They're trying to outlast, not encourage, VoIP use."


In the first instance you seem to be making the case that carriers are investing $Billions to give their customers the latest 'thang' yet in the second instance you are saying that they aren't interested in going there or at least nothing permanent so that they can 'outlast' VoIP (my interpretation of your comment is that carriers feel VoIP will go away eventually)?

You say:
"There's this interesting little battle going, call it, "If you can deliver phone via cable, then I'm going to deliver tv over the phone line and let's see who cries, "uncle," first.""


"The local phone company may not get a check from the consumer to it's voice group when a customer choses VoIP, but it gets a check from the VoIP service provider to the leased line/statistical multiplexed services group, and it is sufficiently large to cover the difference"

As mentioned before the carriers have tried video in the past and have failed miserably (ie: they have tried to fight this battle before and lost, badly). So why are they spending $Billions (yes, that's with a 'B') to potentially lose this battle again now? Carriers don't spend a dime unless they are forced to or they can see a really big reward due to the investment. If, as you seem to be suggesting, the revenue equation for losing telephone customers to VoIP is a wash ("What I'm saying is that the revenue stream may be redirected, but overall the LEC lose is minimal, and in some cases (for instance, when the VoIP provider is the LEC) it's possible that there won't be any lose in revenue.), why would they go into the TV business at all, as their revenue should just get better due to cable-based VoIP? Displacing TV customers from the cable companies will not be a walk in the park by any means, there will be huge price erosion on both sides and customers will take a while to warm up to IPTV. With this kind of competition on the video side, as the RBOCs have to re-coop their $Billions invested in the network, this is no slam dunk for the telcos.

I used to talk to the management of many carriers, some on a daily basis. They always used to say that the money was in the services, the 'value-add', not the carrying bits. This was why MCI & AT&T tried to get into the local phone service business. Perhaps you are telling me that these guys didn't know their business or that VoIP has changed the transport economics so that VoIP carriers now have to pay charges that are "sufficiently large to cover the difference"?

"But I was a witness to the CLECs' self-inflicted debacle, and see no reason to belive that VoIP will be different."

Most of us on these boards were witnesses to this. What I saw was enormous price erosion in the long distance and corporate business units of carriers that included significant layoffs and new legislation (eg: co-lo's, Telecom Act, etc.). CLECs weren't funded well enough to install their own complete networks so could never have killed RBOCs but it was none-the-less a very bloody battle. MSO's have their own networks, complete with last mile access, independent of the twisted pair, and already have huge revenue streams from their TV businesses. This battle looks far less certain, and far more bloody, than anything to do with the CLECs.

Incidentally, at a recent telecom summit in Banff, the CTO of Alliant (incumbent in Canada's Maritimes, part of Bell Canada) got up and said that their local cable company had been offering VoIP-type service since 2000. Alliant has lost 8% of their telephone revenue due to this, he said. He also said that the threat from VoIP is real and will only get worse.
alchemy 12/5/2012 | 3:10:23 AM
re: VOIP Peering: Incumbent Killer? HeavyDuty asks:
"There are actually more HFC endpoints providing high-speed IP to subscriber homes from CableTV MSOs, than all PSTN leased line & DSL services combined."

I'd be real interested in where you found your statistical data.

In the US, this statement is true though DSL is starting to catch up. PSTN leased line, a business service, is tiny compared to residential DSL or cable modem service. Market share varies depending in where you are in the world and this is driven more by regulatory and business issues than anything else. In Europe and Asia, DSL dominates.

OldPOTS 12/5/2012 | 3:10:24 AM
re: VOIP Peering: Incumbent Killer?

LR - Verizon has released a WDM request for proposal. See;


"The RFP targets Sonet/WDM equipment for backhauling video traffic behind the RBOCGs Fiber-to-the-premise [sic] (FTTP) initiative."

"Notter saying a Verizon deal for optical gear, including reconfigurable optical add/drop multiplexers (ROADMs), could be worth $1 billion."
"included optical systems spending over two to three years."

"The bottom line is that Verizon is working on a network transformation at some level -- and ROADMs will be a big player (see ROADMs: Almost Famous). "This is very positive news for the optical market," Clavenna says. "It does validate that triple play can't be seen as just having impact on the access network."

Yes it also impacts the services hauling network.
And the hauling network can cost a $1B. (That's a Billion with a "B")

OldPOTS 12/5/2012 | 3:10:25 AM
re: VOIP Peering: Incumbent Killer? HeavyDuty it sounds as though you will have a long employment as no one seems to get that there is major money in hauling those services.

It is clear that only steven gets that any nation wide (end to end) service must be transported over a very very large nationwide mesh network between all those service aggregation points. If you transport a large share (near monopoly) of these VoIP services, especially your own, wireless and the MOS's, there is plenty of traffic for a huge profitable business. And then IPTV raises this by at least an order of magnitude.

And as rj pointed out MOS and RBOCs will divide up providing service and then will allow those other "competitive services" a very small share of the revenue. And RBOCs and MOS will let those "competitive SPs" exist so they can get all the "competetive" legislation they need to significantly raise their revenues.

I costed hundreds of very large enterprise, tier 1 and tier 2 networks before I retired. Steven the margin on hauling services is so great that RBOCs have to use all kinds of methods to distribute/hide the margins and profits among their various organizations/services. That is why I referenced WilTel contract initially.

I haven't checked lately, but I believe WilTel has let their customer leased line business atrophy as they make much more money (margin and profit) hauling only portions of (adjunct) RBOC's services. The cost of entry into this business is as big as entry into providing those services keeping entry low. And then the other guys (service providers including wireless) sell your hauling services for you. Thus providing furthering higher profits from that already high margin.

In my costing experience the cost of ports is rather insignificant, especially IP/Eth switches. They already own the existing expensive class 5/4 switches, so if traffic increases they won't miss the small margin they loose to other haulers. But they can provide VoIP peering points in their hauling network with those cheap switches (MPLS/IP/Eth) they already own if needed.

Between RBOCs and WilTel and the old Metro Fiber, they own most (~95%) of the fiber that will haul all those services. Listen to HeavyDuty. Steven pointed out that most of the VoIP services are local. So are POTS phone calls. But there are still more than plenty of VoIP CCSs/erlangs that must be hauled over a true nationwide mesh network. (Very much more than a few 10GigE/OC128 lines - think DWDM) Note that I don't call this a core network as most cable and eye pee'ers only think of a core net in local or maybee regional terms and not the scale of large RBOC mesh networks. Even a WilTel list of large circuits is so huge that the spread sheet must be split up.

Also steven I have worked with the RBOCs testing eye pee BW requirements and it is about ((72-64)/64) = 12.5% more. But that does not included those additional very large IP/Eth headers that they know how to avoid. With the headers a minimal increase in traffic can be up more than ~50%. And if anyone does come up with the "killer app", the RBOCs get an even more significant amount of traffic to haul, even when they don't offer any new services.

And the money just keeps rolling in. Don't kill that cow, hauling network.

Page 1 / 4   >   >>
Sign In