Peter Mueller's Blog

Here's what's lurking inside of Amazon Fire

2014-04-08T12:47:00.002-04:00

I'm going to go out on a limb here and say that, despite all the press, the Amazon Fire has little to do with your TV or Gaming, and has more to do with building a Peer-to-Peer distribution center inside your firewall. Imagine if Amazon were to use your garage as a remote distribution center for physical goods, and you start to see what they're up to with your broadband connection.

The hardware specs are impressive, and are needed - I suppose - for this thing to become a gaming center. But the specs are overkill for simple video streaming, so there must be something else in play.

Without getting into the Net Neutrality debate, you already know that Netflix and Comcast have differing views as to who should pay for incredible bandwidth demands of millions of viewers all streaming House of Cards at once. Comcast thinks Netflix should pay its share for 'over the top' use of their network. Netflix thinks Comcast should just be quiet (and grateful) that all their customers pay them fees each month access the Internet.

What tends to go missing here is the impact of the so-called "Busy Hour" problem. Basically, it's not that Netflix users need to download so many Megabytes that's the problem, it's that they all want to do it during the peak viewing hours of 6-11PM where, by some measures, Netflix accounts for >50% of all Internet activity.

Which, brings us back to Peer-to-Peer. If Amazon turned every Fire device into a peer-to-peer client and server, then all those bytes wouldn't have to be served directly off of S3, and 'peered' into Comcast. Fundamentally, I think Comcast could get behind this idea, because they don't really want all that excess burden on their networks either. Why pull the next episode down from the Internet (and cross multiple routers) when your neighbor already has it on their own Fire? If all that traffic stays on-net (from the vantage point of a Comcast neighborhood router), then everybody wins.

Furthermore, Netflix already has all the algorithms it needs to predict and prepare for what you are about to watch. If you've gone to bed at 11PM having just finished episode 3 of House of Cards, are they really going out on a limb to think you might return the following night to watch episode 4? And, if that's the case, why not download that episode ahead of time, in the middle of the night, when Comcast doesn't really care what you're up to anyway? Remember, what Comcast worries about is having their customer complain, during prime time, that ISP speeds are slow. If Amazon and Netflix could timeshift their downloads and/or seamlessly do neighborhood peer-to-peering, most of the download/Capex hassles vanish for Comcast. They get less network traffic, Netflix avoids peering charges and you experience 'faster' speeds because some content is already cached and waiting for you. Everybody wins.

(Also, remember that when you're watching Netflix, you're actually downloading a file stored on Amazon Web Services. Netflix has always been a huge customer for AWS, and AWS has always done a good job of serving them. Werner Vogels and Jeff Bezos have never seen it as a conflict of interest that some of their employees need to provide the best possible Infrastructure to Netflix, and others have to develop streaming video services to compete against them. Solving the problem of peering and busy hours would be a huge benefit to both of them. Interestingly, it's a problem already solved by all the illegal filesharing sites, so adapting it to AMZN and Netflix is very doable.)

Let's see how this plays out, but I think your broadband connection and that little-bit-of-shelf-space beneath your TV just became part of Amazon's worldwide distribution network. You can choose to connect the HDMI cable if you want to, but I don't think Amazon cares. Just remember to hook up the Ethernet, will you?

$578: AT&T's per-sub-valuation of T-Mobile

2011-03-25T11:21:00.001-04:00

Interesting behind-the-scenes look at the AT&T / T-Mobile deal, how private-vs-public sales deals work, and what the main assets are. (Namely, subscribers and spectrum in population-dense areas.)

Ar(T) history...

2010-09-24T16:13:00.001-04:00

CenturyTel and Qwest in $10.6 Billion Deal

2010-04-22T09:47:00.002-04:00

Early news this morning about a *major* consolidation in the Wireline space. It's an all-stock deal.

http://dealbook.blogs.nytimes.com/2010/04/22/centurytel-and-qwest-said-close-to-a-merger/?hp

What 'multitasking' means for Phone Operators...

2010-04-09T11:28:00.001-04:00

In short... not good.

When smartphone companies like Apple announce they'll support multi-tasking as a 'new' feature, it sounds a bit like a TV company saying they'll support color. Hasn't multi-tasking been around for a while already? Yes, but Apple (and others) have essentially crippled it, so that all of the phone's resources were available to one app at a time. (Some privileged apps, usually Apple's, were allowed to run in the background.)

Until now, this 'running in the background' has kept a lot of voice competitors out of reach. After all, how useful is Skype - for receiving calls - if you're considered 'offline' whenever your phone is off, or you're doing something else, like listening to music. If Skype could run in the background, then the whole dynamic changes. In fact, that's just what Apple's recently announced.

Expect a huge influx of competing apps that will make use of this, and allow routing calls over VOIP, thereby bypassing the wireless networks in homes and offices. Also, assuming Verizon and AT&T can't block it, you can expect a lot more 'data' to flow over the carriers and - eventually - less 'voice'. You still need the 3G connection (at least, when you're away from WiFi) but you won't need nearly as many minutes. Presumably, AT&T knows this, and will bolster revenue from the 3G-side, as well as work their Interconnection magic to recoup some costs from voice termination. This will 'hide' the costs from the consumer, and spread it around a bit to the network players. But, in all things consumer, we are definitely headed towards a 'per Megabyte' pricing model rather than a 'per minute' one. And, regardless of costs, we're about to see a lot more players in the 'telco' business, if you broaden 'telco' to mean: An application that can originate or terminate a voice call to your mobile device.

FCC loses key case

2010-04-06T15:57:00.002-04:00

An appeals court ruled in favor of Comcast on the matter of Net Neutrality, declaring the FCC didn't have jurisdiction. Offhand, it's interesting this came out the same week as the IPad - undoubtedly, Steve Jobs was planning on passing much 'content' over the neutral net. In any case, I agree with the WSJ's opinion that the immediate winners are the Cable companies and the Wireless carriers. No doubt the big "content pushers" (i.e. Google, Apple, Amazon, etc...) are licking their chops and readying an appeal.

More here.

New life for copper?

2009-03-26T16:39:00.002-04:00

Mark Twain once famously said "The reports of my death are greatly exaggerated". Maybe the Telcos are thinking the same thing. A new report out by equipment maker Ericsson indicates that, under the right circumstances, they can get up to 500 Mb/sec with multiple twisted pairs "bonded" together over a relatively short distance of 500 meters. Still, if this last "quarter mile" can be made more economical with existing infrastructure, then we'll be seeing competition heat up even more for Broadband.

See the whole article here...

Verizon to Charter: Not so fast!

2009-01-30T17:35:00.003-05:00

In the escalating speed war of download speeds, Charter seemed to up the ante this week with a (limited) 60Meg/sec DOCSIS 3.0 alternative.

Verizon responded immediately by pointing out that the speeds were only achievable if certain (upstream) pipes were widened -- something they might already have done.

Interesting concept embedded in the back-and-forth is the concept of guaranteed quality-of-service. You can't offer a triple-play without partitioning bandwidth. For example, if you put all your eggs in winning the download competition, your users may not have enough bandwidth allocated to make a simple phone call. Similarly, triple play users can reasonably expect to watch TV and make a phone call at the same time.

This is going to get interesting... More here:

http://www.cedmagazine.com/Verizon-DOCSIS-3-0-not-answer-013009.aspx

Broadband usage caps con't

2009-01-07T11:03:00.004-05:00

We've looked before at how long "all you can eat" broadband subscriptions would survive. And, we've seen that Comcast (and others) began experimenting in 2008 with charging by the Gigabyte. Now, a new white paper has emerged that challenges the tenet that phone & cable companies should sequester their top 5% of users by charging them more. The reason? These "data hogs" might actually be performing a useful function by effectively serving as a "node" on the network.

For more info, check out this link.

Will 2009 see the continued shift of Broadband to usage-based-pricing at the same time as Voice continues to slide to flat-rate?

Time Warner Cable to meter Internet usage

2008-01-21T16:01:00.000-05:00

Is the "All you can eat" Internet buffet winding down?

Time Warner Cable announced last week that, it an effort to curtail the heavy costs imposed by 5% of its users, it would try to meter their usage and charge accordingly. It's probably no surprise that this announcement came in the same week as Steve Jobs pitching Apple TV version 2.0. (That app makes turns ITunes into a superstore of Video as well as Music --- Video downloads that must ride the backbone of ISPs down to the user's hard drive.)

In the 25-year old "Bell Head" vs "Packet Head" battle-royale, you can chalk one up for the Bell Heads this week. Call Detail Records (CDRs) will probably live to see another day, this time reincarnated as something like "Byte-Detail-Records" --- and, if successfully, they'll get people to stop drawing Internet clouds (whose usage is always perceived to be free).

http://www.reuters.com/article/industryNews/idUSN1639580720080118

Non-neutral-net and the China Syndrome

2007-10-19T11:31:00.000-04:00

Readers of this column have heard me pontificate before that the 'all you can eat buffet model' (aka The Internet "Cloud") is unsustainable. Flat-pricing models (the ones we're all used to when it comes to bandwidth) are unsustainable over the long run, because 'cost' can't be distributed efficiently enough.

Unlike other flat, per monthly fees (say a bus pass or a gym membership), bandwidth outliers can use thousands of times more capacity than the model, casual user. So, what's a carrier to do?

1- Do they charge each other - effectively bringing intercarrier compensation to IP packets?

2- Do they implement 'tiered' services for the end-user?

… or, hot off the press this morning …

3- Do they selectively block high-bandwidth services such as BitTorrent and other file 'sharing' applications?

http://news.yahoo.com/s/ap/20071019/ap_on_hi_te/comcast_data_discrimination_2

Some new evidence that ISPs, in this case, Comcast are opting for number 3. By blocking access to the bandwidth 'hogs', they ensure better service (and less CAPEX) to maintain the all-you-can-eat-flat-pricing.

But that hardly seems a lasting solution, especially given the bad press that will come of this.

Fiber as a different kind of weapon?

2007-05-11T16:08:00.000-04:00

Interesting analysis by Brian Santo at CED today (http://www.cedmagazine.com/article.aspx?id=147702) about how ILECs decommission their copper lines and what the impact might be on competition. You see, when an ILEC gets approval from the FCC to replace Copper (which must be resold -unbundled- to the CLEC) with Fiber (which has no such imperative) ... guess what: They do just that.
Now the CLECs are worried that, were this tactic used aggressively, it might diminish their capacity (via leased copper) in specific markets.

I've long been a sceptic about Fiber-to-the-home... not because it isn't great service (I hear it is!), but because the price seems so high. (Wall Street shares my view... Every conference call is littered with questions akin to "how are you going to make back the thousands of dollars of Capex for each home passed by fiber? Is there really that much demand for pay-per-view? Aren't you inviting a commodity-price-war with the Cablecos?)

But, alas, this might explain it: What if the real goal is to decommission copper? Then "Capex per home passed" is really an investment in "home-not-passed-by-CLEC".

Pretty, pretty smart.

The coming IPTV Interconnection Debacle

2007-02-06T11:34:00.001-05:00

We've talked here before about how the 'old/POTS/Voice/Interconnection' model might offer a fit to the upcoming IP mess. Now, there's extra reason to believe it.

In this week's Cringely column (http://www.pbs.org/cringely/pulpit/2007/pulpit_20070202_001566.html), we learn that - like voice trunks - bandwidth is 'engineered' on the assumption that not everyone is on-line at the same time. (And even if they are, they're staring at a screen a good portion of the time --- not actively downloading something.) That's the same premise as voice, of course --- no telephone switch in the world is equipped to let every user make a call at the same time.

His quote:

"While you may think your 1.5-megabit-per-second DSL service or your 3-megabit-per-second cable modem service is actually backed by 1.5 megabits or 3 megabits of Internet bandwidth, they really aren't. ISPs provision backbone access based on the expectation that people usually aren't on the Internet, and even when they are on the Internet most of their time is spent reading the screen, not actively sending or receiving packets. As such, ISPs have been able to get away with buying 20-30 KILOBITS per second of Internet backbone capacity for every MEGABIT per second of Internet service they are selling at retail. This 20-to-1 provisioning ratio of what's sold to what is promised (and believe me, 20-to-1 is me being generous to the ISPs since it is probably much higher than that) is what creates the burgeoning Internet video problem."

All of this means that newer services, such as IPTV, can only be implemented by improving on the 20-to-1 ratio. And that costs money --- big money coming out of the ISP's pockets. When you think about it, this is a pretty good reason for the ISP's wanting to bust up the net neutrality argument.

So, we stick by our long-standing prediction that a new metric will emerge (similar to originating and terminating minutes in the POTS world) that will adequately spread the cost of provisioning these facilities on an intercarrier basis. It won't be long before we see a "Missoula Plan" for Video over IP. (http://policycouncil.nationaljournal.com/NR/rdonlyres/91E0BD63-F543-4F0D-AB00-7BBDD3E1BB7D/36101/060719251MissoulaPlan1.pdf)

AT&T Merger explained...

2007-01-29T14:15:00.000-05:00

... by 'telecom analyst' Stephen Colbert...

http://www.youtube.com/watch?v=Bj1Mtv9cD0I

Google's next step in world domination

2006-08-02T15:32:00.000-04:00

Looks like Google is at it again… This time, they’re going to target-place ads inside a radio stream on XM channels.

http://online.wsj.com/article/SB115452648259524557.html?mod=hps_us_at_glance_mm

Can TV be far behind?

Is VOIP quality getting worse?

2006-08-01T10:30:00.000-04:00

My colleague Paul Gilbert has, for some time, been warning about the perils of 'real time' traffic being placed on IP Networks. He's got some backing from Om Malik - who questions whether VOIP quality is getting worse.

According to Brixnet, which monitors such things, the rapid growth of peer-to-peer applications (such as Bittorent - used to download videos, music and everything else - now accounts for perhaps more than 25% of all public IP traffic) is causing your VOIP packets to have a lot more competition at each and every "hop". (A useful reminder of how routers work is found here, at Verizon's Learning Center.)

In the circuit world, it's all or none. You either get a circuit, in which case you can count on hearing a pin drop from thousands of miles away. In VOIP, you don't get a fast-busy, but you can get degradation of service due to underlying infrastructure issues.

No wonder we're seeing the big carriers invest so much in Fiber to the Premises ... and no wonder we're seeing battles rage over "Net Neutrality"...

LERG, meet DNS...

2006-07-01T10:37:00.000-04:00

"Bit heads" and "Bell Heads" have, by nature, trouble seeing eye-to-eye. The whole idea of Voice over IP is anathema to some who think in terms of physical trunks, number plans and so forth.

Now, ENUM (the worldwide numbering authority) is pursuing a new project that promises to bridge the gap ... and I mention it here because it makes an important point: That 'location' can be determined using decades-old web technologies, and propell the concept of Local Number Portability far into the future.

First, a little background. When you pointed your browser to this page, your browser did a little sleight-of-hand. It turned a web address (say, www.atso.com) into an IP address (64.243.114.17) -- keeping you from having to remember an awkard set of numbers, and giving us the flexibility to move the web site to another server without having to inform everybody. In fact, all we do is tell certain servers (called DNS, for Domain Name Server) of the change, and that's it... The results seamlessly propagate around the globe to anyone who wants to know where "www.atso.com" resides. The same hold for other servers (or sub-domains) in our domain, say simcall.technologies.products.wayne.nj.atso.com... don't bother looking that one up!

So, what ENUM is proposing is to do the same thing for telephone numbers. Think of a phone number - 973-696-0990, being turned into something like 1.9.7.3.6.9.6.0.9.9.0.arpa.net - and you see what's up. Suddenly a phone number looks suspiciously like a web address, and the DNS-world can take over, telling the world where you are, which services you have, and who your telecom carrier is. Given a front-end to this system (which is, by nature, dispersed), you, the end-user could update your own records, allowing you to have one-number, forwarded to any of your 'services' you desire: e-mail, voice mail, VOIP, cell phone, etc...

It's a trend worth watching, because if it takes off, we'll see the further degradation of area codes and prefixes meaning anything remotely geographic. It will also mean that new data sources will emerge to be the authoritative source of which company "owns" which customer, and that will have far-reaching repercussions in the areas of Interconnection compensation and routing. Today's Local Exchange Routing Guide (or LERG) will no longer help much in routing calls, or classifying traffic as belonging to a particular carrier.

For more information, check out:

An overview of the ENUM committee
http://www.enumf.org/

An overview of the "TN numbers homed in a DNS server" project:
http://www.enumf.org/documents/6000_1_0.pdf

One Step closer to "Net Neutrality"

2006-05-26T16:29:00.000-04:00

Today, a Congressional House Panel voted to allow "Net Neutrality", which is great news for content providers (think: Google Videos and Apple Itunes) but bad news for broadband providers (think: AT&T, Qwest, Cablecos, etc...)

I'm guessing a few lobbyists are going to get fired for this one. After all, much of the Bells' investment strategy relied on getting both you and the content-provider to pay at the same time. Now, they have to let Google pile on all the high-broadband content they want, and invest in the infrastructure to carry it.

Of course, Google would argue, they're doing the Bells a favor: After all, DSL is plenty fast for e-mail and other now-low-tech-apps. If someone, like Google, doesn't dream up new high-bandwidth IP-based applications, then no one will care how much FTTH (Fiber to the Home) Verizon invests in.

Verizon may think: "Build it and they will come", but Google thinks it can help with a few high-bandwidth digital apps to serve as a carrot.

In the end, they need each other, I guess.

A new BPL player gets ready

2006-05-04T09:26:00.000-04:00

We've made the case before here that even though the cost of "retail" voice minutes is plummeting, the "intercarrier" complexity remains. For as far as the eye can see, who gets to carry traffic along portions of a call's route will remain at least as important as what the customer pays for the call. And, as long as these turf wars remain, there will be handoff issues as well --- all of which means that intercarrier compensation schemes will only get more - not less - complex.

This morning's Wall Street Journal reports on the progress of Current Communications, and their foray into Broadband over Power Lines (BPL).

A new choice for the consumer. A new method of providing Broadband service. And a new connected party to the PSTN.

__________________________________________________________________

Current Communications Gets New Funding
By AMOL SHARMA

Current Communications Group LLC, which is offering Internet access over power lines, plans to announce $130 million in new equity funding, including backing from General Electric Co. and EarthLink Inc.

The start-up, whose other investors include Google Inc. and Goldman Sachs Group Inc., has developed a technology -- available in Cincinnati and soon in Dallas -- that offers consumers high-speed Internet access using small adapters that plug into an electrical outlet. Utilities also can use the network to monitor the security and efficiency of their grids.

Broadband-over-power-line technology, known as BPL, has been identified as a potential threat to the cable and phone companies that dominate the market for high-speed Internet access, but there have been few examples to point to thus far.

The infusion of money from well-established companies shows that Current's efforts are gaining traction. The recent funding also came from Texas utility TXU Corp. and Sensus Metering Systems Ltd.

Current is controlled by a partnership of chairman and co-founder William Berkman and Liberty Media Corp.

Prediction: Reciprocal Comp and CABS Coming to Broadband

2006-01-15T10:01:00.000-05:00

It's a funny thing: Though it's one of our core businesses, we often hear from our customers that they're not sure how much longer intercarrier compensation will last. It's true that rates have plummeted, and it's true we've been watching the all-you-can-eat-voice-offering take hold for some time. But, at the end of the day, no carrier wants to presume that traffic is completely balanced over their networks... so they measure it, and charge for it.

An article in this morning's New York Times caught my eye.

With Yahoo, Google, Apple and others racing to deliver video over broadband connections, what impact will that have on 'last-mile' providers such as the Baby Bells and the Cable companies? Won't the bandwidth consumption needed to download that latest episode of "Desperate Housewives" require a kind of dedicated high-speed lane, provided by the carrier?

The short answer is, of course, yes... But it remains to be seen who will pay for it. It's conceivable that the broadband providers (and by that I primarily mean DSL and CableModem providers) will create new tiers of 'usage' for their heaviest users. But, with competition already tight (not to mention the growing use of terms like "parasitic" to describe the likes of Vonage because they benefit completely from others' hardware investments), it's much more likely they'll pursue an 'intercarrier doctrine' with the content providers.

Besides, it's ingrained in the Bell tradition (brought on by divestiture itself) and there are a lot of 'Bell heads' looking for a place to hang their hat in the new world. Devising a new Carrier Access Billing System standard (with the new C perhaps standing for 'Content') for the interplay of content providers and network operators is something that should get done and - my prediction here - will start in earnest in 2006.

PS A follow up to my post. Some readers wrote in deriding this idea as a new revenue source for the carriers. But, there's no reason to assume the carriers will be 'net beneficiaries' any more than the Bells thought they'd be 'net beneficiaries' of recip comp agreements with the CLECs. After all, we know that most Bells are working hard to get into the TV business (Notably, AT&T and Verizon). To do so, they'll have to sign the same kind of content deals that have been the staple of Cable for so long. So it might be that a new CABS arrangement would offset content rights (a revenue stream to the content providers) with network-usage (a revenue stream to the Bells and others). Steve Jobs did this with the record-labels. His revenue sharing model works like this: "Let me sell your song for 99cents, and I'll give you - roughly - half in royalties. I'll make the rest of my money selling the IPOD hardware. How the song's bits and bytes gets to the IPOD will be some network provider's problem. We'll leave them out of the equation for now. Call it CABS without the middleman, and let's hope it lasts for a while."

No wonder the Bells want a new arrangment.

What the Skype deal means to the rest of us…

2005-09-28T11:35:00.000-04:00

Since so much of this column’s print has been devoted to VOIP this year, I’d be remiss if I didn’t try to make sense of our industry’s latest big news: Ebay buying Skype for 2.5 billion. (For some perspective, that’s one third of what Verizon will pay for MCI.) Skype, as you probably already know, is a ‘peer-to-peer’ software service that lets anyone BYOMSB (Bring Your Own Microphone, Speaker and Broadband ; my acronym – sorry) and make voice calls to each other for free.

Whenever I hear “for free”, I cringe. After all, why pay around $150 for each of the estimated 20 million active subscribers if they don’t generate any revenue? At first glance, this looks like déjà-vu-all-over-again. Didn’t Microsoft buy Hotmail (another free service) for $400 million a few years back? Is this the same kind of ‘market-share-is-everything’ gamble? The answer is no. Here’s why:

1. First of all, other “free” services (like Hotmail) didn’t threaten any entrenched monopolies. Yes, a few people (like AOL) were making some money charging for e-mail at the time, but this was pre-Broadband and most people were happy to fork over twenty bucks a month just to get on-line via dial-up.
2. If $150 dollars per subscriber sounds like a lot, remember that large ILECs are typically valued at something approaching ten times that number. A traditional POTS customer pays a monthly annuity, so the valuation is a lot higher.
3. Let’s not forget about the cost to acquire a new customer. Here, ILECs, Cable and Wireless companies often peg this number somewhere in the $200-$500 range – a fee that includes marketing and hook-up charges. Skype isn’t encumbered by these costs since – as I mentioned earlier – each new customer brings their own equipment and bandwidth to the table. One point of view on this Judo tactic is that it presumably shifts the ‘cost’ of the service to those last-mile providers of POTS and ISP services. It’s not clear that those services have managed to effectively convey their true cost structures to regulatory boards, without sounding like whining kids who’ve had their toys taken away.
4. Skype has expanded into the “paid” arena. “Skype-IN” and “Skype-OUT” are paid services that interconnect to the POTS network. Clearly, Skype users occasionally still talk to people on the POTS network, and that entails a myriad set of interconnection charges at some point. (Check out our archives for a discussion of what happens when this goes awry, and an unsuspecting carrier is caught trying to figure out where all this phantom traffic is coming from!) So, if Skype could turn 5% of users their into paying customers dishing out $5 per month for interconnection fees (and I guarantee you Ebay will find a way), then – suddenly – there’s a newfound $60M revenue stream, which is about Skype’s forecasted revenue this year.

So who wins, and who loses, in this new deal?

1. I’m reluctant to jump on the “end is nigh” bandwagon for the ILECs, since the ‘last mile’ is still a key piece of the puzzle. The real issue is that ‘one- trick-voice’ ponies are going to feel a lot of pain as voice becomes more and more of a commodity. Interestingly, this may not be the ILECs who – despite their sluggishness – are deploying different services, and getting a bit more of a revenue mix. (Fully 1/3rd of SBC’s revenues are not derived from voice.)

2. More likely, Mobile operators (who typically derive more than 80% of their revenues from voice) are going to feel the pinch. In their (often) misguided rush to build 3G networks (data networks that will easily handle VOIP traffic as well), they’ll probably hand over the keys to Vonage, Skype and others whose customers will use these new networks to float their voice traffic – significantly undercutting the Mobile operators’ voice revenues.

3. Incumbents who try to fight this trend will lose. Madison River, an ILEC, was fined recently for blocking access to the Skype web site. Cable companies’ attempts to distribute cable-modems that block competitive VOIP IP addresses (such as Vonage) will also take it on the chin.

4. Intermediary players that provide Tandem and Media Gateway services (e.g. Packet8, Level3) will gain, since VOIP/PSTN traffic-ferrying is now an important piece of the puzzle.

5. Software providers (like us) will do well by providing greater interconnection validation and clearinghouse functions. The nature of this new traffic is too complex to rely on old-style CABS billing systems. Other Revenue Assurance tools, including network testing tools, will need to be rethought in the VOIP world. After all, conformance to state tariffs – a driver for traditional RA activities – may not apply in the new order. But, the correct validation of routing, provisioning, call types and intercarrier billing still play a role. In other words, there are still real network elements under the Marketing department’s fanciful portrayal of the IP network as a mere cloud.

6. Carriers who jealously safeguard their last-mile monopoly will lose, if they don’t separate their ‘physical service’ (outside plant) from the ‘application services’ (Voice, ISP, TV) that ride on top of them. More and more, these application services will simply flow over the cheapest bandwidth to the consumer. Better accounting of their ‘physical’ costs may even give them some regulatory relief if the FCC realizes that the costs of handling IP/PSTN calls are unfairly being shouldered by the ILEC. This will probably play out first in the Tandems, where much of the cost currently sits.

7. Finally, I’m tempted to say the VOIP providers lose unless they fix their Quality of Service problems. I’ve uncovered that most people who proselytize about VOIP have never actually used it. It sure doesn’t have the quality of POTS yet, but it’s probably on the way to getting better. Besides, the Wireless industry proved that quality isn’t everything… how many dropped calls did you experience last week?

Will Better Routing Save Telecom?

2005-05-09T17:58:00.000-04:00

I read a fascinating article today about a guy who wants to start an airline. What's revolutionary is this: He wants to to carry you, and up to 5 other passengers, anywhere you want to go for a maxium price of about $3 per person, per mile. Figure out where you want to go, and he'll take you there, at a price that's higher than commercial aircraft, but far less than Executive Jets.

If you were him, what would most concern you about your business plan? The cost of fuel? The cost of planes? The insurance? Nope. Just minutiae.

What really worries this guy is effective routing. Before he invests in 700 planes each capable of landing at (roughly) 700 airports, he wants to ensure that he can dispatch the right plane to the right place, using the best route, for the lowest cost. As soon as one plane lands, or a new 'ticket' is purchased, he needs to recompute it all over again. So, what's a guy to do? Hire a bunch of programmers to develop this 'traffic simulator' and prove to him it can all be done in under 5 minutes. (By contrast, the commercial airlines do this once-a-day, max, and have far less permutations.)

Once it's done, he starts building the airline. Check 'em out. When was the last time you heard of any major investment hinging on the outcome of a software simulation? Where else would this be useful?

OK, folks. What's the parallel in Telecom? It's not too late to save millions of better-optimized circuit switching. It's certainly not too late to come up with a - any - optimal voice routing in VOIP.

Cisco, are you listening? Lucent? Anyone?

FCC delays important VOIP decision

2005-03-22T13:12:00.000-05:00

A long-awaited decision will have to wait a little longer. Because of the recent change of leadership at the FCC, that court could not rule on whether or not access charges needed to be paid for "VOIP in the middle" - calls that originate on POTS, traverse the IP cloud, then terminate again on POTS. Stay tuned.

http://news.zdnet.com/2100-1035_22-5629045.html

Spam is to e-mail as ???? is to VOIP:

http://voip.weblogsinc.com/entry/1234000470037174/

And then there were four.

2005-02-06T18:50:00.000-05:00

A nice article in the NYT showing how 7 "children" ultimately brought (or is that "bought"?) their own Mother back into the fold.

What's Apple up to?

2005-01-15T13:34:00.000-05:00

OK. By now, we all recognize the tremendous marketing clout of Apple. Through the IPOD, they've turned a small hard disk into the Sony-Walkman reincarnate of this century. But now, we also see that they're going after the PC space as well. Steve Jobs recently joked that he wished he'd had a nickel every time someone asked him why he didn't pursue a 'downmarket, outsourcing' strategy similar to Mr. Gates' "let everyone use DOS, then Windows" strategy. Well, he says, now he does. Enjoy the Mac Mini.
But, those of us in the telecom space need to take note of what's really going on here. What he's distributing is actually a mini-Unix server (take note Software Providers touting state-of-the-art ASPs: How does this revelation change things? Well, it could mean that distributed computing will get in vogue real fast; faster than you can say "My customers want my app more than they want to shuttle date to and from me." More on that in another post.)
It also means that Apple's little-talked about communication device and software, ICHAT is going to start taking aim at Cisco's expensive VOIP phones and, maybe, even a slice of ol' Ma Bell. With the FCC set to rule on VOIP's implication rulings (likely to remain unregulated, except for when it terminates on the POTS network), there's going to be increasing reasons for VOIP players to find a way to avoid the POTS network. Getting business consumers hooked on ICHAT & MAC makes Apple a telecom player, not just a computer manufacturer. Watch out, Ma Bell ... Ma Steve has plans.