& cplSiteName &

WiMax Guide

Light Reading
1/5/2005
50%
50%

2005 will probably be the year in which WiMax sinks or swims as the next big thing in telecom technology. For that reason, plenty of people are going to need a quick and easy way of getting up to speed on what WiMax is and why it's stirring up so much interest.



This report aims to provide that quick and easy guide, by answering questions that are most frequently asked about WiMax. In the report's initial form, answers to eight basic questions are given, one per page. But the idea is that readers can ask further questions on the message board attached to this article. If you want to send a private message, please email [email protected] and include "WiMax Guide" in the subject field. Frequently posed questions will be answered by adding pages to this report.

For those that just want a high-level understanding of WiMax, here are the key points:

  • WiMax is a broadband access technology that delivers high-speed, carrier-grade Ethernet data communications by wireless over city-sized distances.
  • Its main use initially will be to deliver broadband Internet connections to buildings by wireless and to provide links between WLAN hotspots and carriers’ core IP networks (backhaul); later, individual user devices such as PCs will be connected direct, and, eventually, mobile devices.
  • Its big attractions are expected to be the usual Ethernet and wireless virtues of low costs, flexibility, ease of rollout and use, and interoperability.
  • However, WiMax is not expected to be fully commercialized until about 2006.


Industry opinions are divided over WiMax's prospects, judging by early results of a poll being conducted on Unstrung, Light Reading's sister Website covering wireless technologies. So far, 48 percent of respondents say WiMax will "emerge as a credible wireless replacement for DSL/cable" and exactly the same percentage say it will "splutter onto the market with minimal impact."

To take the poll yourself and see the full, up-to-date results, please click on this link.

Here's a hyperlinked list of the questions and answers in this report:



Webinar Archives on WiMax



Paid Research Reports on WiMax

— Tim Hills, Freelance Telecommunications Writer and Journalist


Need to know more about the latest developments in WiMax? check out the coming Light Reading Live! conference:

WiMax – Why Now?
at the W Hotel, Union Square, New York City, on Tuesday, January 20, 2005


This one-day event, hosted by Rick Thompson, Heavy Reading Analyst at Large, will provide qualified attendees from Light Reading's global audience an education in how WiMax will fit into their networking development plans for 2005 and beyond.

  • For more information, click here
  • To register, click here
  • Sponsorship opportunities are still available. Direct all inquiries to [email protected].


(69)  | 
Comment  | 
Print  | 
Threaded  |  Newest First  |  Oldest First        ADD A COMMENT
Peter Heywood
50%
50%
Peter Heywood,
User Rank: Light Beer
12/5/2012 | 3:15:20 AM
re: WiMax Guide
http://www.answers.com/topic/w...
keelhaul42
50%
50%
keelhaul42,
User Rank: Light Beer
12/5/2012 | 3:28:26 AM
re: WiMax Guide
RJ:
Solving this is no small task. One wonders if Iraq will become a democratic state before the US has a modern communications infrastructure. Let's hope for our children's sake both can happen within a generation.
>>>>>>>>>>>>>
RJ, you sure can turn a phrase. I thought I had it nailed by saying we were set to fall behind N. Korea but I can see you've done me one better.

-kh
rjmcmahon
50%
50%
rjmcmahon,
User Rank: Light Beer
12/5/2012 | 3:28:27 AM
re: WiMax Guide
Structurally, the only way it can happen is if the government "buys out" the RBOCs.

The following paper from the Electric utility industry gives some guidance on how this could be done.

Valuation and Compensation Issues in Establishing a Public Power Utility

http://appanet.files.cms-plus....

It suggests to me that the existing local loops will need to be condemned and brought under local ownwership.

The idea of using US tax payers dollars to "buy out" RBOC shareholders doesn't solve the problem. That's a financial game where nothing of real value gets created. Value creation will occur when people start investing into the upgrades of the outside plant and into the establishment of modern, Internet Protocol capable, operations and management staff. Motivating value creation is what needs to be done.

Solving this is no small task. One wonders if Iraq will become a democratic state before the US has a modern communications infrastructure. Let's hope for our children's sake both can happen within a generation.
keelhaul42
50%
50%
keelhaul42,
User Rank: Light Beer
12/5/2012 | 3:28:29 AM
re: WiMax Guide
>>>>>>>>>>>>>>
That's fine, but the truth is that the RBOCs are doing bo diddley against those spaces anyway. Cablecos don't need the ILECs, except to interconnect, and have the clout to retain it. So that's kind of like asking to not let elephants into the dining room.
>>>>>>>>>>>>
Well, no. If the RBOCs start feeling some heat from the Cablecos they will respond in the way they know best: get the FCC to administer a dose of regulatory poison. (Instead of competing GǪ)
They surely have a regiment of lawyers & lobbyists in training for this should they need to use them.


>>>>>>>>>>>>>>
First off, the RBOCs are no longer required to offer line sharing anyway -- where have you been? They sometimes offer it commercially (not as a UNE) because it's incremental revenue and takes away from cable.
>>>>>>>>>>>>
Sorry, poor phrasing/choice of words. The RBOCs wailed bitterly about line sharing, said they can't buildout tomorrow's access networks if forced to share at below market rates etc, etc, ...
Enter the carrot: get something done that we want (high speed buildout) and we'll spare you the stick (forced line sharing)
I think the FCC made a poor deal on this one, could and should have squeezed the RBOCs much harder.


>>>>>>>>>>>>
In any case, you're right back in the Industrial Policy area that I referred to earlier.
>>>>>>>>>>>>
Remember I said this was partially tongue in cheek? It's the path of least resistance: requires very little other than holding the line on new regs and extracting just a little more from the RBOCs in exchange for releasing them from line sharing. Besides, I'm not an ideologue either, moderation, all that jazz :-)


>>>>>>>>>>>>>
But getting ADSL2+ to 75% of homes within a decade is trivial.
>>>>>>>>>
Huh? Billions of $$, digging, cutting, splicing, etc? What are you telling me and when can you start?


>>>>>>>>>>
You'll end up with a simple duopoly, one CATV and one ILEC, each with its own ISP, and no other ISPs. Then those ISPs will offer filtered, walled-garden services (think "Great Firewall of China", only for commercial reasons as well as to curry political favor). Much worse than we have now. Who gives a goat's bzadeh if the ILEC can stream Fox News at you in HDTV?
>>>>>>>>>>>>>>>
Two is better than one, especially if they both have to answer to shareholders. You forgot wireless (three). There's also satellite for those who just want Fox News in HDTV (3 -1/2).

>>>>>>>>>>>>>
Now, WRT pubic ownership, you're reciting the precise ideological argument that I was expecting. The government is actually pretty good at some things. Who else provides roads, bridges, sewers, flood control structures, and similar infrastructure? And for that matter, Social Security has an extremely low overhead, far better than the mutual funds. So I don't think letting them own a LoopCo would result in a bad job being done.
>>>>>>>>>>>>>>
No, I gave a factual argument. Granted, some infrastructure seems to have been handled reasonably well by govt agencies. But very often it can be - and has been - privatized with excellent results. So the jury is still out on that. Let's give Fredric Bastiaat some credit and pay attention to what is seen and what is not seen.
Re: social security, please educate yourself on this. If you want that can be a separate discussion that easily rivals this one. But please don't mention it in an attempt to bolster a case for govt programs -- that's bad salesmanship.


>>>>>>>>>>>>
Nor am I calling for it to be a monopoly. It is by its nature a natural monopoly, which means that there won't be a lot of competitors even if it's allowed.
>>>>>>>>>>>>>>>>>
I thought a monopoly meant ONE entity. Anyway, we have the likes of Comcast et al spending capital to bring cable service to residences, wireless is getting a hard look, ...
Two or three choices are better than one and one is all we'll get if we fall into this natural monopoly mode of thinking.


>>>>>>>>>>>>>
Now if you go to Scott's article, you'll see the RBOCs' long-term future being cloudy. Revenues are not keeping up.
>>>>>>>>>>>
Good, might get them off their collective duff ...
fear has been effective in motivating other private enterprises. A little panic in HQ could be just what we need.

>>>>>>>>>>>>>>>>
Powell's industrial policy was to have needless overbuilds, with each CLEC and ISP expected to own its own wire on the poles, hoping to get enough market share to pay off. That does help the bucket-truck-rental industry and the glass industry, but it's still an economic waste.
>>>>>>>>>>>>>>>>>
Time will tell. The companies doing that are businesses run by and for people who will feel PAIN if they fail, joy (or whatever moves you) should they succeed. This is what moves markets. Let it work.
Let's not pretend this paradigm applies to govt enterprises. Take a look at what you'll get out of social security if you've more than a few years to retirement for a graphic reminder.

>>>>>>>>>>>>>>>
Again, the idea is to not repeat the problem of the last divestiture, when the spun-off Baby Bells decided not to stick to their knitting, and ended up abusing their monopoly power.
>>>>>>>>>>>>>>>>>>
Agreed. One very good way of avoiding this problem is to make sure no one has the monopoly power. Let them earn their way just like the rest of us.

-kh
fgoldstein
50%
50%
fgoldstein,
User Rank: Light Sabre
12/5/2012 | 3:28:30 AM
re: WiMax Guide
Keel, I'll answer both of your notes.

- defeat any RBOC regulatory initiatives aimed at crippling the cable tv/internet/telephony space

That's fine, but the truth is that the RBOCs are doing bo diddley against those spaces anyway. Cablecos don't need the ILECs, except to interconnect, and have the clout to retain it. So that's kind of like asking to not let elephants into the dining room.

-loosen up regulations on RBOCs vis-a-vis line sharing. Let them charge whatever they want for colocated equipment & line access. In exchange for this give them a 5 year timeline: get real broadband (e.g. 10 Megabits per sec) to a high percentage of US households (e.g. > 75%) or the regulations will come back along with any public canings deemed appropriate for RBOC mgt.

First off, the RBOCs are no longer required to offer line sharing anyway -- where have you been? They sometimes offer it commercially (not as a UNE) because it's incremental revenue and takes away from cable. In any case, you're right back in the Industrial Policy area that I referred to earlier. But getting ADSL2+ to 75% of homes within a decade is trivial. And what good will it do the consumer, if it's on the RBOCs' terms (no ISPs except their own)? You'll end up with a simple duopoly, one CATV and one ILEC, each with its own ISP, and no other ISPs. Then those ISPs will offer filtered, walled-garden services (think "Great Firewall of China", only for commercial reasons as well as to curry political favor). Much worse than we have now. Who gives a goat's bzadeh if the ILEC can stream Fox News at you in HDTV?

Now, WRT pubic ownership, you're reciting the precise ideological argument that I was expecting. The government is actually pretty good at some things. Who else provides roads, bridges, sewers, flood control structures, and similar infrastructure? And for that matter, Social Security has an extremely low overhead, far better than the mutual funds. So I don't think letting them own a LoopCo would result in a bad job being done.

Nor am I calling for it to be a monopoly. It is by its nature a natural monopoly, which means that there won't be a lot of competitors even if it's allowed. Now if you go to Scott's article, you'll see the RBOCs' long-term future being cloudy. Revenues are not keeping up. Powell's industrial policy was to have needless overbuilds, with each CLEC and ISP expected to own its own wire on the poles, hoping to get enough market share to pay off. That does help the bucket-truck-rental industry and the glass industry, but it's still an economic waste. Sort of like the king whose answer to a depression was to pay people to build walls and tear them down, over and over. There are better uses of a resource.

Since it is a natural monopoly, and real competition isn't going to happen (duopolies don't count -- they collude), the answer is to regulate the monopolist. I'm perfectly happy to see it be a joint-stock company, with some fiscal incentives to be efficient (but not rate caps). Again, the idea is to not repeat the problem of the last divestiture, when the spun-off Baby Bells decided not to stick to their knitting, and ended up abusing their monopoly power.
keelhaul42
50%
50%
keelhaul42,
User Rank: Light Beer
12/5/2012 | 3:28:31 AM
re: WiMax Guide
As a followup, here's my admittedly minimalist, partially tongue in cheek (and NOT thoroughly considered ...) proposition:

- defeat any RBOC regulatory initiatives aimed at crippling the cable tv/internet/telephony space

-loosen up regulations on RBOCs vis-a-vis line sharing. Let them charge whatever they want for colocated equipment & line access. In exchange for this give them a 5 year timeline: get real broadband (e.g. 10 Megabits per sec) to a high percentage of US households (e.g. > 75%) or the regulations will come back along with any public canings deemed appropriate for RBOC mgt.

Comments, flames, whatever, welcome. I just want to stimulate some thinking.

-kh
keelhaul42
50%
50%
keelhaul42,
User Rank: Light Beer
12/5/2012 | 3:28:32 AM
re: WiMax Guide
fgoldstein wrote:
...
Frankly the main problem with your proposal is that it has no hope of going ahead at this time. Government ownership of anything -- even Social Security -- is ideologically noxious to the "conservative" wing of the Republican Party.
....
>>>>>>>>>>>>>>
It ought to be noxious to anyone who wants his money's worth of anything.
I understand the frustration with the RBOCs. A good spanking seems to be in order. But in seeking a solution let's not overlook the dismal record of govt run enterprises in general. Sure, there are some bright spots - the seeds of the internet, for example - but they are few and far between.

Let's also not forget that a philosophy of regulated monopolies is what got us into this mess in the first place. Ok, in fairness, better ideas may not have been obvious but will a govt run monopoly somehow be better?

How about thinking of some alternatives now instead of complaining about those right wing Republicans?
We all think it's worth doing, let's try to conceive of a free enterprise way to get it done.
Haven't the cable companies been spending heap plenty bucks to do just that? That shows that _someone_ is willing to invest private capital to make it happen -- it doesn't have to be the telcos who can just die on the vine if they don't want to get off their butts.

I'll grant you that not every govt project turns into a disaster but the record is bad enough that we ought to think long and hard before following that path again.

-kh
fgoldstein
50%
50%
fgoldstein,
User Rank: Light Sabre
12/5/2012 | 3:28:33 AM
re: WiMax Guide
Scott, you make some very good points. We agree on the fiscal problems facing the Bells and the folly of "more monopoly" as a solution.

Frankly the main problem with your proposal is that it has no hope of going ahead at this time. Government ownership of anything -- even Social Security -- is ideologically noxious to the "conservative" wing of the Republican Party. They'd probably privatize the Interstates if they could, even if it meant that the renamed "Ford Freeway" was closed to GM, Toyota, and DaimlerChrysler vehicles.

Municipal ownership is less noxoius to those moderate Republicans who, as Christy Todd Whitman said on a radio interview today, hold that if the people can't do something themselves, the government closest to them should, with the feds as a last resort. But then Whitman was being interviewed about her new book whose topic is the alienation of moderate Republicans like her from the ruling wing.

That's why, for practical reasons, a regulated private utility model is the closest to being both doable and ideal, though perhaps not ideal in either regard. I suppose it could happen in two ways. Going back to your excellent 2002 article, I suggest that a smart RBOC boss would do it voluntarily. If I had Ed's or Ivan's job, I might create the LoopCo voluntarily, keeping say 49% of the equity if allowed, but saddling it with debt which, as a utility, it could repay from cash flow. The second option is antitrust: Trinko protects some demonopolization from antitrust, but the Bell's current and desired behavior strikes me as crossing well over the line. Antitrust led to the 1984 breakup so it's the obvious weapon of choice. And of course divestitures like AT&T's and Standard Oil's before it led to increased, not decreased, total shareholder value.
Scott Raynovich
50%
50%
Scott Raynovich,
User Rank: Light Sabre
12/5/2012 | 3:28:35 AM
re: WiMax Guide
fgoldstein,

Clearly you've gone into depth on this and thought through many of the details. I've been a big fan of some sort of LoopCo model for a while.

http://www.lightreading.com/do...

The most popular analogy is the federal highway system. As you have pointed out, subsidizing transportation can bring whole new industries to life -- such as what happend in the U.S. to the car and tire industry after the Interstate system was created (the tire industry still fights mass transit programs because it would shift money away from cars).

But I think an even better analogy to LoopCo and a proposed municipal or quasi-governmental takeover of the access networks is the INTERNET. The Internet was a gigantic federally funded R&D project that was eventually spun off to private interests. It helped create a gigantic multi-billion dollar industry overnight.

Why can't we do the same with the access network?

What was the key to the Internet? First, of course, was the fact that it was engineered properly and up to the highest quality specs. Secondly, it was based on standards. It was standards that later enabled the commercial Internet to explode.

Think if the government followed this model, developed a transitional, government sponsored fiber communications network using new telecom standards (ATCA, IMS, IP, e.t.c.). The government RFPs for access equipment would be available to all and would ensure standard components and access widgets were used throughout the network. I think eventually this would have massive economies of scale and reduce the prices of access equipment by miles. Yes, a few access goliaths and components suppliers would end up winning, but that's going to happen anyway.

Your LoopCo could be funded by revenue bonds backed by future cuts of the network. Structurally, the only way it can happen is if the government "buys out" the RBOCs.

People say -- but would the RBOC shareholders ever allow this?

I say yes, if the government made the shareholders happy.

Think about it. Do these carriers REALLY want to be in this business? It looks like a pretty miserable business to me. If the government offered to pay them to take over their debt, would that really bother them?

Let's take Verizon. It's got $40B in long-term debt. It's market cap is about $100B, yet it's enterprise value (takeover value including debt is $140B). So clearly shareholders would be eager for somebody to aquire it at enterprise value, which represents a 40% premium. They would be eager for the government to take the debt off its hands. The market is saying it's not worth that much clearly... but the government could suck it up in the belief that eventually, standardization and open access networks would create new growth (and tax revenue).

Now, currently, shareholders get a nice 4% dividend on Verizon. But that's really the only reason to own the stock. So, in the end, it behaves like a municipal bond. Let's face it Verizon is not a growth company. Wouldn't we all be better off if the government bought all these RBOC access networks, repackaged them as quasi-municipal assets backed by tax-free revenue bonds, and sold them to the public that way?

Now the RBOCs are a municipal asset paying the bondholders tax-free yields. If the government rolled them all together and created a standard "Access Internet," and then adjusted regulation acccordingly, the whole thing could then represent a new platform for growth.

Obviously these are ambitious ideas and the lawyers, lobbyists, and Congressmen probably don't want to have anything to do with it. But the whole thing is worth mulling over.

--Scott
fgoldstein
50%
50%
fgoldstein,
User Rank: Light Sabre
12/5/2012 | 3:28:36 AM
re: WiMax Guide
Fred, in a couple of your posts you mention that wireless is decoupled and irrelevant w/r/t natural monopoly last mile considerations. I might buy that. Not sure.

Can you elaborate why/how in more detail?


Wireless systems have a fairly linear cost-capacity relationship. If you are supporting 10,000 subscribers, it costs X to build. If you are supporting 20,000 subscribers in the same geography, and the system is fairly well built out (i.e., current take rates, not early-stage development), then it costs roughly 2X to build. Each Erlang of phone call capacity, or its data equivalent, needs the same amount of transmitter power, bandwidth, etc. Therefore there is not a lot of an economy of scale, which is the defining characteristic of a natural monopoly. A small startup is at a disadvantage, of course, but it is nothing like trying to run parallel local loop wires. And while spectrum is tight, it has more room for entrants than the poles have vacanat gains. (Most urban poles are already full, or very close. Overbuilders pay a lot for make-ready, which the incumbent did not have to in order to install its plant.)

There are minor economies, of course, in having one big carrier rather than two small ones. When Cingular bought out AWE, they could lay off HQ staff, consolidate billing, etc. And towers need to be shared; however, that's already done. American Tower, Pinnacle et al own the towers, not the wireless carriers. The FCC has in the past made rules against tower exclusivity contracts.

But those are irrelevant to the "LoopCo" issue. Seven had postulated that wireless companies should lose their plant to a LoopCo, thus replacing a competitive marketplace (no one wireless company has more than about 50% market share in any major market, though the Powell FCC has been moving the market from the Hundt-Kennard 6-carrier-per-market model to a 3-carrier model) with a monopoly, which would presumably treat existing wireless carriers as MVNOs. That is just nonsensical as it goes entirely against the pro-competitive direction of the LoopCo proposal.

Also, the LoopCo proposal serves the ILEC shareholders' interests, because it jettisons debt and allows the slowly-depreciating LoopCo assets to be capitalized differently (more debt) from the (one hopes to be) faster-moving ServiceCos (more equity).
strands555
50%
50%
strands555,
User Rank: Light Beer
12/5/2012 | 3:28:38 AM
re: WiMax Guide
Fred, in a couple of your posts you mention that wireless is decoupled and irrelevant w/r/t natural monopoly last mile considerations. I might buy that. Not sure.

Can you elaborate why/how in more detail?
strands555
50%
50%
strands555,
User Rank: Light Beer
12/5/2012 | 3:28:38 AM
re: WiMax Guide
brookseven says: "The problem with local control is: It fragments the market so dramatically that there is no economy of scale. Prices of equipment [do not drop]dramatically."

Oh that's such dinosaur economist bunk. The economies of scale exist across advances in silicon and optical device processing in general, affecting all equipment designs uniformly, for all intents and purposes. 500k units of "emerging market" EPON chips doesn't cost any more than 500k units of mature 10/100 Ethernet controllers, unless there's a huge gate count, I/O, or packaging difference (which has nothing to do with current or future market size of the specific function the device implements).

Even with standards-based silicon products, every vendor's mask set is different. They might as well be devices for completely different markets.

Vendor underbidding on equipment using dinosaur economist models that presume some gigantic reduction in cost, from "economies of scale" theories from the industrial revolution (even though most of the market growth spurt happens within the same generation of process technology) should not be confused for "marke
rjmcmahon
50%
50%
rjmcmahon,
User Rank: Light Beer
12/5/2012 | 3:28:44 AM
re: WiMax Guide
[Local control] fragments the market so dramatically that there is no economy of scale. Prices of equipment dramatically.

The majority of the expense is paying the labor to install the new fiber. Even free equipment doesn't help with that. So it's not a problem "economy of scale" can solve.
rjmcmahon
50%
50%
rjmcmahon,
User Rank: Light Beer
12/5/2012 | 3:28:44 AM
re: WiMax Guide
I don't care about a bunch of lawyers and pedantics discussing things in the minutiae, but this is bordering insanity.

Yeah, I get frustrated too.

rjmcmahon
50%
50%
rjmcmahon,
User Rank: Light Beer
12/5/2012 | 3:28:44 AM
re: WiMax Guide
Social issues of haves or have nots will cause this to be a national issue. Why should Palo Alto have a network but the South Bronx not?

Both Palo Alto and South Bronx deserve modern communications infrastructures. Today, neither has one. The challenge is figuring out how to make it so.

The "federalist" approach of looking to Wall St. or to DC isn't solving the problem. Instead we get statements from SBC CEO Whitacre that nobody needs 100Mbs (and he doesn't even mention 1Gbs) on the one hand and on the other hand we get capitalists selling us 1.5Mbs wireless. Neither is adequate nor acceptable.

It's got to start somewhere. Unfortunately, that does mean there will be a period of time of haves and have nots. If that raises the issue such that Congress passes the equivalent of the Rural Electrification Act for real broadband than that's a good thing. It means we're progressing.
jim_smith
50%
50%
jim_smith,
User Rank: Light Beer
12/5/2012 | 3:28:47 AM
re: WiMax Guide
How spineless and pathetic do you have to be to use a new moniker to attack someone on a message board?

Ceoati, will you please crawl back into the hole that you came from.
fgoldstein
50%
50%
fgoldstein,
User Rank: Light Sabre
12/5/2012 | 3:28:48 AM
re: WiMax Guide
By the way, rate cap has been the paradigm for over 15 years. That is what causes lack of investment in infrastructure.

You're contradicting yourself. Sometimes your message is the rate-of-return would cause a lack of investment. Now rate cap is the problem? Make up your mind. I suggest that neither is necessarily pro-investment or anti-investment; the devil is in the details.

You are clearly a bigot. Make the Wireless companies unbundle. Explain why they must not.

"Bigot" is a term generally used to refer to those who show prejudice against some persons based on personal distinctions such as race or nationality. Your application is bizarre.

Okay, here's the deal. Incumbent wireless providers must unbundle their copper and fiber plant, where it is the principal provider.

(And I'll sell you all my kosher pork chops for a buck a pound.)

Wireless, I point out, has nothing to do with the wireline plant, nor the incumbent common carrier plant, nor broadband common carriage in general. It's a red herring, something irrelevant that you're trying to drag in just to distract from the discussion at hand, the natural-monopoly loop plant.

You have now killed new services by the way. Lets say that a new service costs $20 Million to create (lets say ADSL to introduce cost $100M dollars). Now, your company must pay that up front before there is any development. Get it? Your model is completely and utterly useless unless its a government body.

Again, you are not describing my proposal. I'm suggesting that LoopCo should be required to introduce services that are likely to be widely accepted, as ADSL was, though of course ADSL is (except for the remote-terminal case) not a LoopCo service at all. If LoopCo claims that a service is not likely to be widely accepted, and therefore its costs cannot be amortized into its rate base the way these things have worked since the 1880s, then it becomes a BFR. Regulators should make the call, guided by input from the carriers who would be LoopCo's wholesale customers.

There is, of course, a school of thought expressed in this thread that LoopCo should be municipally owned. I am agnostic on that point. It would work, but I also agree that economies of scale could work against it. With regard to equipment purchases, though, I note that small providers often band together. For instance, little cable companies usually join the Cable TV Cooperative and thus get decent rates on both programming and equipment. Similarly, a bunch of municipals could create a purchasing consortium, as many retail stores do. There are however real regulatory economies with larger entities. At the extreme, I doubt the New Jersey BPUC would want to regulate 566 new municipal LoopCos.

Tell me, where does the revenue growth from LoopCo come? How does it get 10% or so growth per annum? If it has no growth, who are its investors?

That's simple. We put Bernie Ebbers in charge. I have a feeling he's not doing much travel for the next few years. But seriously, the whole point is that not all companies follow the same fiscal mantra. LoopCo is not supposed to be a growth company; it's a utility, and it's supposed to pay a steady dividend and retain value. That's where the investors come from -- it inherits bonds, and the stock should pay a decent yield. GTE, not LDDS. Plenty of investors prefer steady income to risky growth. Yes, I've read Management by Compulsion and while it explains Worldcom perfectly, it's exactly what LoopCo should not be.

By the way the answer already exists, there are now 3 competitors driving massive investment in the network.

Not relevant to the end users -- there is only one common carrier, and it doesn't even want to be a common carrier any more. ISPs and other competitive information providers depend on common carriage. And the ILECs are abusing their dominance in telephony to squeeze out CLECs -- even interconnection is in danger.

Non-Facility based CLECs should die and let cable replace them for residential.

Speaking of prejudice... I frankly don't care that much about UNE-P either, though I find it on balance beneficial. But UNE-Loop is a real facilities-based operation, "Smart Build" being its other name, and it's that type of investment that I'm out to encourage. Encourage modern electronics at the CO end of the wire, but the wire itself, copper or glass, has a longer useful life than the electronics.

For businesses, gee we had competition before 1996 and continue to do so. So, for businesses there is no problem.

For a few thousand large buildings in mostly NFL cities. The vast majority of business locations are not on CAP fiber, and are not on cable. Read the FCC record in the TRO Remand, if you want evidence. For businesses there's a huge problem. And since businesses rarely use ILEC ISPs, it's a really big problem if say Verizontal gets its way and is allowed, as they've petitioned the FCC, to kick all competing ISPs off of their >56kbps (DSL, DS1, DS3, ATM) network.
rjs
50%
50%
rjs,
User Rank: Light Beer
12/5/2012 | 3:28:49 AM
re: WiMax Guide
Think about it .... let us not get lost in details.
Go back in time to 1950's during Eisenhower times.

Think about Verizon and SBC and RBOCs laying down the highways and FCC making and enforcing the regulations on the use of highways!

GET IT!!

The FORDs, GMs, TOYOTAs and HONDAs would not be what they are today, nor the UPS or FEDEX or the numerous other shipping companies and services that make up the economy of this country.

I don't care about a bunch of lawyers and pedantics discussing things in the minutiae, but this is bordering insanity. It is quite obvious and clear that something is very wrong with the system the way it is (for whatever reasons in the past, hind-sight is always 20-20). This stupid little band-aids over massive head trauma ain't going to solve this. WAKE UP! The emperor is walking naked.


Thoroughly disgusted ...

-rjs
paolo.franzoi
50%
50%
paolo.franzoi,
User Rank: Light Sabre
12/5/2012 | 3:28:49 AM
re: WiMax Guide

By the way, rate cap has been the paradigm for over 15 years. That is what causes lack of investment in infrastructure.

You are clearly a bigot. Make the Wireless companies unbundle. Explain why they must not.

You have now killed new services by the way. Lets say that a new service costs $20 Million to create (lets say ADSL to introduce cost $100M dollars). Now, your company must pay that up front before there is any development. Get it? Your model is completely and utterly useless unless its a government body.

Tell me, where does the revenue growth from LoopCo come? How does it get 10% or so growth per annum? If it has no growth, who are its investors?

By the way the answer already exists, there are now 3 competitors driving massive investment in the network. Non-Facility based CLECs should die and let cable replace them for residential. For businesses, gee we had competition before 1996 and continue to do so. So, for businesses there is no problem.

seven
paolo.franzoi
50%
50%
paolo.franzoi,
User Rank: Light Sabre
12/5/2012 | 3:28:49 AM
re: WiMax Guide

rj,

The problem with local control is:

- It fragments the market so dramatically that there is no economy of scale. Prices of equipment dramatically.

- Social issues of haves or have nots will cause this to be a national issue. Why should Palo Alto have a network but the South Bronx not?

seven
rjmcmahon
50%
50%
rjmcmahon,
User Rank: Light Beer
12/5/2012 | 3:28:51 AM
re: WiMax Guide
rj is correct about this aspect, if you want LoopCo then the US Government has to buy it, run it and maintain it.

Uhmm, I never advocated that.

Local control seems like the only way to enable progress in this arena. In that regard, I believe modern fiber networks should be built and maintained by municipal utility districts and/or be customer owned. Something along the lines of water districts or public power could do the trick.
rjmcmahon
50%
50%
rjmcmahon,
User Rank: Light Beer
12/5/2012 | 3:28:51 AM
re: WiMax Guide
Do any of you industry guys see a PRACTICAL way out of this? (PRACTICAL meaning private enterprise funded, owned, and operated service entities doing the work)

If practical is defined that way, where one places blind faith in "the private markets", I believe wireless networks with spotty access is the best we can hope for. Future generations will inherit a subpar communications infrastructure. It won't be much of a legacy.

Now if we define practical as the public sector and private industry working together to achieve maximum potentials, then the sky is the limit ;-) Let's hope folks like LUS can lead the way.

http://www.lus.org/site.php?pa...
fgoldstein
50%
50%
fgoldstein,
User Rank: Light Sabre
12/5/2012 | 3:28:53 AM
re: WiMax Guide
You seem to want to stop all investment by wireline telcos. Old wires old rules (UNE-L), new wires (or fibers) should have freedom - just like wireless and HFC. They did not have an unbundling requirement because they were viewed as new wires.

The "new wires new rules" mantra is ILEC propaganda. It wasn't part of the Telecom Act. The Act doesn't require new construction in order to fill UNE requests, but it doesn't limit UNEs to plant in service before 1996 either.

The LoopCo plan would in fact enhance investment. There is no way that most ISPs and CLECs can duplicate the loop plant. The ILECs, by harming their access to the OSP, are killing their ability to invest in the switching and information-processing facilities that they provide. UNE-P is a red herring and under my plan it doesn't exist so don't bring it up again. I frankly think that Verizon's FTTP is a sham, being done now to get regualtory concessions, but its expansion will be stopped before long. And their plans are to make it such a crock anyway that people will miss their old copper, which at least wasn't censored more strictly than China's nets.

So, make all 3 modes exactly the same from a regulatory standpoint. You have given no reason other than the fact that some companies were built under fake regulatory arbitrage to keep unbundling around. That is all it is. FAKE - REGULATORY - ARBITRAGE. There is no incentive to invest in the network.

No, it's the VoIP industry that's built on arbitrage. My "re-divestiture" plan does not particularly support UNE-P -- the ServiceCos keep their switches, subject only perhaps to transitional antitrust requirements.

- Must Loopco develop the access method for your 1,000 customers or,

Bona Fide Request should be cost-based. On a global basis, regulators should guide LoopCos to provide a current list of standard offerings. If they don't suffice, then a LEC could BFR something else, and the cost would be negotiated within cost recovery parameters.

- Do you require approval for this wonderful service from a Regulatory Body?,

LoopCo would be regulated.

- If this latter is true, is it a state body or a federal body?

State.

- How is the cost of the development, deployment, and maintenance of these 1,000 lines amortized?


If it's a custom offering, then to the requestor on a BFR basis. If it's to become a new standard offering, then more widely. The latter case might be handled via regulation -- a CLEC doesn't want to pay BFR, so it asks the regulator to have LoopCo develop a new tariff. LoopCo, however, doesn't have to act at a loss.

- If you (fgoldstein) pay for all this work and I decide to build a CLEC on this same offering, do I have to kick back the cost of developing this stuff to you?


Depends on the terms of the BFR. ILECs have done that for years. Adding a line link pulser to a 5XB cost about $100k, charged to the company making a request for DID service, but if a second company wanted DID, they'd pay $50k which would go to the first requestor. Standard regulatory constructs would apply.

- Would LoopCo return to a Rate of Return carrier with an 11% ROI. Would this not cause folks to dig up every street and lay a fiber to every home whether they wanted fiber or not, at our expense?

Yes, they'd be Rate of Return. No, they wouldn't dig up every street without demand, because proper regulation of an RoR carrier includes prevention of rate-base padding. We'd probably have good old-fashioned general rate cases, like we did before 1993.

- Would LoopCo not make a case to replace every bit of fiber every year to make sure that as much was invested as possible to get a big a check as possible from all of us?

See my previous answer. Trust me, RoR-regulated GTE certainly did not do more updating than necessary! As the folks in Santa Monica. The plant in much of VZ in the midwest is still decrepit.

- If they are not allowed to lay fiber to everyone, then who gets to decide who gets fiber and who does not?

Wherever a carrier (wholesale customer) commits to pay for enough to make it worth LoopCo's while.

As a rate of return carrier, then we are back to the utility model where things move very very very slowly.

Some things work better under the utility model. Outside plant does move slowly. Fiber has a long lifespan. So do poles and copper. Too much deregulation is a bad thing. The right wing wants to deregulate transmissin, to let monopolies extract monopoly rents and restrict its use, but they regulate content, putting their efforts into protecting America from the scourge of Janet Jackson's mammary. I don't want the wire to be controlled by someone who owes Fox News a favor.
paolo.franzoi
50%
50%
paolo.franzoi,
User Rank: Light Sabre
12/5/2012 | 3:28:54 AM
re: WiMax Guide

Absolutely I am being facetious.

You seem to want to stop all investment by wireline telcos. Old wires old rules (UNE-L), new wires (or fibers) should have freedom - just like wireless and HFC. They did not have an unbundling requirement because they were viewed as new wires.

So, make all 3 modes exactly the same from a regulatory standpoint. You have given no reason other than the fact that some companies were built under fake regulatory arbitrage to keep unbundling around. That is all it is. FAKE - REGULATORY - ARBITRAGE. There is no incentive to invest in the network.

Now you have a cute little CLEC and lets say you devise a new service. This service is big to you because it will add 1,000 new customers to your network. Of course, nobody else wants this service so,

- Must Loopco develop the access method for your 1,000 customers or,
- Do you require approval for this wonderful service from a Regulatory Body?,
- If this latter is true, is it a state body or a federal body?
- How is the cost of the development, deployment, and maintenance of these 1,000 lines amortized?
- If you (fgoldstein) pay for all this work and I decide to build a CLEC on this same offering, do I have to kick back the cost of developing this stuff to you?
- Would LoopCo return to a Rate of Return carrier with an 11% ROI. Would this not cause folks to dig up every street and lay a fiber to every home whether they wanted fiber or not, at our expense?
- Would LoopCo not make a case to replace every bit of fiber every year to make sure that as much was invested as possible to get a big a check as possible from all of us?
- If they are not allowed to lay fiber to everyone, then who gets to decide who gets fiber and who does not?
- If they are not a rate of return company, why would any human being invest in them? How are they supposed to make an ROI for their shareholders?

See there is a reason that the days of LoopCo (pre-1984 AT&T) were rate of return companies and are now rate cap companies. rj is correct about this aspect, if you want LoopCo then the US Government has to buy it, run it and maintain it. There is no private money for a 0 growth, 0 income business (except for people who invest in Corvis). As a rate of return carrier, then we are back to the utility model where things move very very very slowly. Imagine AT&T deciding that it wanted its money back on its ISDN investment instead of us having DSL.

seven

PS - KT gets the ability to get premium money for its content as part of its bundled offering (it does not have to open this up to others). It also gets a cut of other premium content being run over its network (say 1 penny for every iTunes download?). Gee....wonder why broadband is such a good idea for them.
fgoldstein
50%
50%
fgoldstein,
User Rank: Light Sabre
12/5/2012 | 3:28:55 AM
re: WiMax Guide
Seven,

I'm not sure if you're serious or facetious.

You need to roll into this the HFC network and all the wireless networks.

These can be open just as you describe for the wire.


Cable companies have never had an unbundling obligation, nor a common carriage obligation, so they have technically been model citizens. Although I think it would have been in everyone's intrest, including their own, except the ILECs', had they adopted a more "open" stance towards their HFC.

My proposal allows but does not require the cable companies to work with the LoopCo. An overbuilder could, however, work with LoopCo even if the incumbent cable operator didn't.

Then we stop all investment in copper and HFC and transition directly to FTTH.

Maybe, maybe not. HFC has a lot of legs -- I'm involved in designing new HFC systems this year. In one case, the "competitive advantage" that the overbuilder is selling is "more analog TV", because the largely-elderly, rural population in his isolated burgh is resistant even to ordinary "digital cable". Don't even think of Video-over-IP there! (But there's a video-over-IP overbuild a county or two away. Choose your poison.)

Copper, of course, is in decline, though FTTH is still more expensive than short-throw copper (such as CSA or FTTC). I doubt a LoopCo would be asked to do much large-scale investment there, except maintenance.

We stop all development of all non-3G Wireless and move everybody to a single wireless access infrastructure.

Preposterous. Wireless is totally separate, and not in LoopCo's domain. There is no natural monopoly over wireless systems, nor even the stiff economies of scale that apply to wireline. Mandated tower-sharing (to deal with NIMBY) is enough. If one operator wants CDMA2000 and another WCDMA and another oldeGSM, and another wants to dance with Flarion, let'm duke it out on their own frequencies.

The tricky area, quite frankly, and here I am venturing back to the nominal topic of this thread!, is the relationship between loops and wireless in high-cost areas. I'm on record at the FCC proposing that USF high-cost recipients be offered free AWS spectrum to do WLL in lieu of future high-cost loop construction. WiMax might be just the ticket to keeping that price down for a faster-than-voice product. Of course the FCC staff misquoted me grotesquely in their Order, ridiculing it on false grounds. (Again, there's an article about that on the Ionary web site.) Their real issue, I suspect: WTB owns AWS spectrum, and WCB owns USF, and the WTB folks have no interest in solving WCB's problems. That's more evidence that Powell's FCC re-org was a miserable flop, which will further tarnish his "F-grade" legacy.
keelhaul42
50%
50%
keelhaul42,
User Rank: Light Beer
12/5/2012 | 3:28:55 AM
re: WiMax Guide
The link works ok. Not a bad idea but, as you note in the article, the RBOCS will resist -- fiercely.

I recall reading elsewhere (can't recall where, though) of similar propositions. There have been others calling for something more radical, such as abolishing the FCC and letting the chips fall where they may.
Given the enormous (gi-normous as I'm hearing from the kids these days) lobbying clout of the RBOCS in the political arena it's not easy to see any of these ideas gaining traction.
What is easy to forsee is the RBOCs finding a [regulatory] way to cripple the rising competition from cable and wireless.

We've already behind S. Korea in residential broadband. I hear the usual RBOC rationalizations about population density, easy to reach apartment bldgs, etc. One wonders how long before we fall behind N. Korea. (and the rationalization for that!)

Do any of you industry guys see a PRACTICAL way out of this? (PRACTICAL meaning private enterprise funded, owned, and operated service entities doing the work)

-kh
paolo.franzoi
50%
50%
paolo.franzoi,
User Rank: Light Sabre
12/5/2012 | 3:28:55 AM
re: WiMax Guide

You are not going far enough.

You need to roll into this the HFC network and all the wireless networks.

These can be open just as you describe for the wire.

Then we stop all investment in copper and HFC and transition directly to FTTH. We stop all development of all non-3G Wireless and move everybody to a single wireless access infrastructure.

seven
fgoldstein
50%
50%
fgoldstein,
User Rank: Light Sabre
12/5/2012 | 3:28:56 AM
re: WiMax Guide
In your mind, how would the access networks (LoopCo) be defined such that they supported the service providers in a neutral way and also supported end consumer choice between those providers?

I actually posted my position as a little article on my web site two years or so ago, called "Breaking up the ILECs: A Win-Win Proposition": http://www.ionary.com/ion-redi... . (Does LR permit links?)

No, it does not use IP! I see IP's role roughly where ISO puts it in ISO8648, defined above common carriage. That's what "internetwork" means, after all; common carrier may provide the "networks" that IP interconnects. The current fashion of recursing IP into serving as Layer 1 is, in protocol-weenie terms, an abomination.

What I suggest for structural separation, in brief:

LoopCo takes over the entire ILEC outside plant, as well as the wire center buildings. LoopCo owns DLCs, copper, and glass. LoopCo owns intraLATA interoffice facilities, up to the fiber terminal and core DACS (ie., whatever an MSSP does). It is a carriers' carrier, not selling to its customers' customers. (A line has to be drawn somewhere.) It can sell to cable companies, who might see value in outsourcing work to them.

ServiceCo (the nominal successor to the ILEC) keeps the customers. It keeps the switches, including the data switches used for data services. It leases loops from LoopCo at the same price that the CLECs pay. It leases collocation space from LoopCo at the same price that the CLECs pay. (This allows CLECs to collocate swithes too.)

LoopCo prices are regulated, on a permanent utility rate-of-return model. (I haven't decided whether I think LoopCo or ServiceCo should get the USF funding.) LoopCos upgrade plant on request from any carrier. They don't have to do anything for a major loss, modulo the usual rate averaging; they are regulated to minimize risk. Essentially, the Bells end up buying UNEs from them, though TELRIC might not be the primary basis of prices. LoopCo would provide FTTH or FTTC or whatever it was paid to do, but that plant would be non-exclusive; more than one carrier could buy fiber loops from them. A given drop might be exclusive though.

LoopCos gets most of the ILECs' debt, while the ServiceCos get a more volatile capital structure, without much debt burden. Their stock could be a good widows-and-orphans issue; ServiceCo stocks probably wouldn't be, but would have more upside potential too. Labor and concominant obligations are divided with the tasks performed.

Your position on VoIP (or voice over connectionless data networks) seems to lead to the conclusion that generally accepted "convergence model" may be a faulty goal and rather that voice, broadcast video and data networks should remain on separate networks. Is our industry being led astray with the assumption of convergence?

Yes. To the extent that the IP layer is the convergence one, that is. The physical media, however, can certainly converge (think HFC or BPON), and one can conceive of a higher layer (than that) convergence protocol taking IP's place, or going under IP (which is what ATM was intended to be, though its builders were a bit unsure about how to do it). Just because it isn't in use yet doesn't mean that it can't happen.

BTW, I saw a great quote this week that seems germaine to what some of the posters have said.

"Research is the process of turning money into knowledge; innovation is the process of turning knowledge into money."

I wish I knew who said it. (Google credits Scott Steedman.)
ceoati
50%
50%
ceoati,
User Rank: Light Beer
12/5/2012 | 3:28:57 AM
re: WiMax Guide
Does Jim-Smith have nothing better to do than post buffoonish statements to every discussion on this site? I suggest he/she/it get a job and astop suffering constant rubukes.
rjmcmahon
50%
50%
rjmcmahon,
User Rank: Light Beer
12/5/2012 | 3:28:57 AM
re: WiMax Guide
So now we have the same hubris, just from the other side. Treat voice like data, rather than data like voice.

In your mind, how would the access networks (LoopCo) be defined such that they supported the service providers in a neutral way and also supported end consumer choice between those providers? Some suggest that the "/" between TCP/IP is the natural place to break things. That does make the assumption that IP can carry all bits in a neutral way and still meet all service level requirements. Do you see a different way to achieve this separation?

Your position on VoIP (or voice over connectionless data networks) seems to lead to the conclusion that generally accepted "convergence model" may be a faulty goal and rather that voice, broadcast video and data networks should remain on separate networks. Is our industry being led astray with the assumption of convergence?
fgoldstein
50%
50%
fgoldstein,
User Rank: Light Sabre
12/5/2012 | 3:28:59 AM
re: WiMax Guide
Also, it seems to me that those that are screaming the loudest against IP technology as the defacto access technology are really trying to protect the incumbents. CLECs are dependent upon the incumbents for their existence. Your position of advocating for CLECs may be biasing your technological views. I don't really know. But it seems fairly obvious that IP is what is being used to connect computers together and those are the access networks that need to be built.

Well no, RJ. And I'm no talmid chocham either; I have just been in some interesting places at interesting times. I actually like your idea, the "howstuffworks" approach.

I do agree that IP is what is being used to connect computers together, and indeed for that purpose it is usually the best technology. It's using IP for everything else that ticks me off. And I'll tell you why. It's not because I don't like IP. It's because 20-25 years ago, we went through the same damn thing, "integrated voice and data" this and that. But it didn't work then, because voice and data were different. And they're still different!

In 1980, voice guys were trying to cobble together data networks. You want to see pathetic? Look at Bellcore's 1984-1985 ISDN presentations. Made the worse because they ignored the failure of 1980-1982's "PBX supercontrollers". (There's a decent bit about that era in my new book, The Great Telecom Meltdown, btw. I chronicle all sorts of failures.) So I spent years at standards committees trying to make networks like ATM usable for data, when the dominant telco guys there couldn't understand why "VBR" wasn't just a perfect model of data traffic, or why a 10-bit CRC (look up AAL3/4 for a laugh) wasn't good enough.

So now we have the same hubris, just from the other side. Treat voice like data, rather than data like voice. Well, I studied traffic engineering and congestion management from the best in the business (Raj Jain and KK Ramakrishan, mostly) and now I see people ignoring those things and trying to brute-force VoIP and TDM over IP and ATM over IP and whatever. And it demos nicely, but it is not fundamentally stable. Not many nines.

So while IP is fine, if the market wants a single one-size-fits-all architecture, I suggest that it should be one that was designed for that, and not, like ATM, a bunch of bad committee-driven compromises. (Even after which, ATM works pretty well. It's just not fashionable.)
jim_smith
50%
50%
jim_smith,
User Rank: Light Beer
12/5/2012 | 3:28:59 AM
re: WiMax Guide
If you assume that IP is perfection incarnate...

Never said that. Go back and read my post.

I'm prefer a solution that will offer: ...

I didn't see anything new in that list. These ideas have been studied to death from a research point of view. Pick your choice of last 20 years of IEEE ToN, JSAC, ToC, Networks, SIGCOOMM, ToISS, INFOCOM, etc. papers and you will find enough work in most of those areas.

The real challenge (and this is an engineering challenge more than a research challenge) is to find ways to implement these "beaten to death" ideas in an existing network, which usually happen to be based on IP. And if you can make money off it, more power to you.

And yes, I do know of people working on this, clean slate (not tweaked ATM or IP or OSI or SNA or whatever), though it's an uphill battle when the conventional wisdom thinks the problem is solved.

Oh ya? And where do these "people working on this" publish their papers? In an obscure journal in Transylvania? I hope our government is not giving them any money to waste!

But in your world, you must believe Gates' alleged quote of Charles Duell, "Everything that can be invented has been invented." Duell didn't say it, Ellsworth didn't say it, and it still isn't true.

I guess it depends on what you call incremental research. I don't blame you. I have never ever seen a Professor or a researcher (or someone like you who is neither a prof or a researcher but is sympathetic to their cause) claim that their work was incremental. It is always the best thing since sliced bread. But most of research is incremental. And most of it will always be.
rjmcmahon
50%
50%
rjmcmahon,
User Rank: Light Beer
12/5/2012 | 3:28:59 AM
re: WiMax Guide
I do agree that IP is what is being used to connect computers together, and indeed for that purpose it is usually the best technology. It's using IP for everything else that ticks me off.

Yeah, I'm ticked off too.

If we laid the fiber communications line the users could decide if IP works or not. Doing so wouldn't break the existing so-called broadband access technologies.

Unfortunately, the incumbants are so afraid of a fiber, ethernet, IP (read internet-style) access model that they fight like crazy to stop it. Hence the bogus legislation in places like PA amongst other things.

Somewhere, some group has to step up to the plate and drive this issue - prove the status quo is wrong and rather is holding back our society. Getting the power to accomplish this is very difficult and maybe even a bit delusional. But hell, so are the attempts at converting the Middle East to democracy (which has cost over $200B to date).

Markets are failing. Regulation is failing. Government leadership is failing. Industry leadership is failing. Wall St. is, well, Wall St. Too few are pulling together. Somehow we have to change this.

I believe it to be an achievable goal and a legacy worth leaving.
rjmcmahon
50%
50%
rjmcmahon,
User Rank: Light Beer
12/5/2012 | 3:29:00 AM
re: WiMax Guide
Nobody is designing networks from first principles any more!

I'm not an expert in what's going on in the R&D area of network design, but folks at Internet2 and Canarie seem like a good starting point to me.

http://www.internet2.com
http://www.canarie.ca

I'm just a layperson, without any patents to my name, or any advanced degrees from brand named universities, so most of their stuff is beyond my comprehension ;-). If it's unable to be explained on places like http://www.howstuffworks.com it's hard to see how it scales to the masses.

The right way is to identify general needs, figure out their actual requirements for capabilities (semantics), and craft a design to solve them (syntax).

Agreed that network design is required but I believe there are many, many experts all over the world designing modern IP networks that meet their requirements. The places to look would be in areas where regulatory affairs don't skew their results.

Also, it seems to me that those that are screaming the loudest against IP technology as the defacto access technology are really trying to protect the incumbents. CLECs are dependent upon the incumbents for their existence. Your position of advocating for CLECs may be biasing your technological views. I don't really know. But it seems fairly obvious that IP is what is being used to connect computers together and those are the access networks that need to be built.
fgoldstein
50%
50%
fgoldstein,
User Rank: Light Sabre
12/5/2012 | 3:29:00 AM
re: WiMax Guide
fg>>No, I'm talking about progress and innovation.

js>Really! Please tell me the "grand open problems" that you are working on :-)

If you assume that IP is perfection incarnate, handed down to Moses on Sinai, then there's obviously no open problem. I don't.

I'm prefer a solution that will offer:

- more reliability, from simpler code paths resulting from a cleaner architecture
- more security, from a design that kept it in mind from the beginning, vs. being designed for the closed world of the ARPAnet.
- more efficient scalability, because it wasn't designed for the ARPAnet's Type 208 and 303 modems.
- horizontal recursion, so small networks can transparently roll up inside big ones (this is related to scalability)
- vertical recursion, so layers that have similar behaviors can gracefully re-use the same code and syntax
- QoS, selectable as required
- low overhead, suited to the type of traffic being carried
- topological addressing, so routing is inherent rather than hideously complex

And yes, I do know of people working on this, clean slate (not tweaked ATM or IP or OSI or SNA or whatever), though it's an uphill battle when the conventional wisdom thinks the problem is solved.

js>I don't know what the view is from your hill, but from the mountain that I live on, I don't see any real innovation left to be done in networking.

This reminds me of the famous misquote in Bill Gates' book, which really dates back to ...Patent Office Commissioner Henry EllsworthGÇÖs 1843 report to Congress. In it he states, "The advancement of the arts, from year to year, taxes our credulity and seems to presage the arrival of that period when human improvement must end." But Commissioner Ellsworth was simply using a bit of rhetorical flourish to emphasize the growing number of patents as presented in the rest of the report. He even outlined specific areas in which he expected patent activity to increase in the future.

But in your world, you must believe Gates' alleged quote of Charles Duell, "Everything that can be invented has been invented." Duell didn't say it, Ellsworth didn't say it, and it still isn't true.

http://www.ideafinder.com/gues...
jim_smith
50%
50%
jim_smith,
User Rank: Light Beer
12/5/2012 | 3:29:01 AM
re: WiMax Guide
I do know core ATM technology very, very well; I have a few patents in the area.

Please don't try that line with me. It may impress (or scare) others, but it doesn't impress (or scare) me. Besides it sounds very desparate.

No, I'm talking about progress and innovation.

Really! Please tell me the "grand open problems" that you are working on :-)

I don't know what the view is from your hill, but from the mountain that I live on, I don't see any real innovation left to be done in networking.

I've seen some good work done in SONET...

I'm all ears. Indulge me. I really want to see how the "innovation" (or lack thereof) in SONET is any different from the "innovation" in MPLS (which you rightly pointed out is FR redux).
fgoldstein
50%
50%
fgoldstein,
User Rank: Light Sabre
12/5/2012 | 3:29:01 AM
re: WiMax Guide
Get over it. IP has completely and totally beaten ATM.

Oh, I'm sorry, I forgot to mention your favorite straw horse! You're assuming that ATM is my one-and-only solution to the "under alles" network? Please find where I said that.

I do know core ATM technology very, very well; I have a few patents in the area. And because I was on the committee, I know why the camel's so lumpy. The 48 octet payload is indeed asinine. The US and most of the world wanted 64 octets, while France (mainly concerned with voice) demanded 32. Both wanted a power of two. 48 octets was a mutually-undesirable compromise arrived at instead. But that's water under the dam. I am not positing ATM as "the" answer. It's one tool. I happen to find it useful: A CLEC I work very closely with is using ATM to deliver voice+data EELs via IADs at the customer site. It provides better QoS (CBR) and lower overhead than VoIP. And don't forget that the ILEC DSL networks are primarily ATM (though frankly it strikes me as silly to use ATM on the actual loop; frame makes more sense).

No, I'm talking about progress and innovation. Remember that? Remember when people actually invented stuff -- even in America? Nowadays it's more like "how can I use IP and WiFi and other buzzword-compatible things to impress the funders?" I've seen some good work done in SONET, and WiMax isn't exactly chopped liver (though it's not a loop substitute either), but the data networking space is oh-so-stagnant. And if I hear one more jerk (like Carlyle Group whore Bill Kennard, once a good guy at the FCC before he cashed in) tell me that the innovation is at the "edge of the network", I may be strongly tempted to get violent.

Dynarc had some clever ideas a few years ago, but couldn't get an audience, and tanked. Not even the universities are doing much good work. CompSci is bogged down in eunuchs, datacomm and telecom in eye pee. MPLS isn't useless -- it's the IETF's attempt to rewrite frame relay with its own jargon -- but it's not really progress vs. what was doable 15 years ago.

Nobody is designing networks from first principles any more! The right way is to identify general needs, figure out their actual requirements for capabilities (semantics), and craft a design to solve them (syntax). Instead, we take an ancient syntax and look for bigger hammers. It may all look like a nail to you but the Primitive Pete method isn't necessarily the best. And it certainly doesn't deserve government favoritism, lest the gravy train be upset by real progress or innovation.
jim_smith
50%
50%
jim_smith,
User Rank: Light Beer
12/5/2012 | 3:29:02 AM
re: WiMax Guide
Access networks need to be digitized and packetized. Eye Pee seems to do the job nicely.

Proof by vigorous assertion? Eye pee does it badly -- no QoS, huge overhead, cockamamie addressing (none, really, just a screwed up multi-layered naming system) and (as a result) routing. It's better than, uh, what? DECnet Phase II with Poor Man's Routing? Bisync? X.25? Gimme a break -- we can do better.

I'm not the one to argue with legitimate success. And regardless of IP's shortcomings, it has achieved what none of the other protocols you mentioned have ever come close to achieving.

Get over it. IP has completely and totally beaten ATM. That too in a very lop sided fight - ATM had the backing of all the big telcos and standards bodies. And while we are on the topic, please explain what is so "perfect" about ATM's 53 byte cell and 48 byte payload? 53! 48! Sounds like a compromise between idiots!

Even now there are huge "I pee on ATM" networks. Dude, that's a lot of peeing... on ATM.
rjs
50%
50%
rjs,
User Rank: Light Beer
12/5/2012 | 3:29:02 AM
re: WiMax Guide
" ....What is the point of a CLEC selling the RBOCs service? What value does it add to the network? Are you declaring this is the way to fund a business? "

Brookseven, an answer to one of your rhetorical
questions which I have pasted in quotes is that
with UNE-P and CLEC resellers, the ILECs cannot
rape the public with price gouging. There is a good reason for this, the ILECs and for that matter any monoploy, cannot be allowed to police itself. They do not have any altruistic bone in them!

With UNE-P the government has more or less valued how much it costs the ILEC provide the line and how much extra they can charge. Think of the CLEC with UNE-P as a balance to gouging power of ILECs. Remember, ILECs used OUR money to build their monopoly. They have not, repeat, HAVE NOT, earned the right to gouge. On the other hand, if they had built the monopoly on their own, they would be justified in complaining about UNE-P. In that respect, Microsoft and the cable companies have a better justification for gouging the public. Atleast they built their monopolies on their own.

UNE-L is a different story altogether since the CLEC invests in equipment.


-rjs
jim_smith
50%
50%
jim_smith,
User Rank: Light Beer
12/5/2012 | 3:29:03 AM
re: WiMax Guide
And better yet to spin off the entire ILEC Outside Plant to a neutral carriers' carrier, allowing true competition.

I couldn't agree more. Unfortunately, they should have done that in 1996. The government missed its chance. It's going to be impossible to do that today. Only way to get this done today is to link the ILECs with terrorism.

OK. So what can we do now to fix the problem?

Here are some crazy thoughts:

The Eye-Pee layer can be used to decouple the service provider from the connectivity provider (who is equivalent to a "carrier's carrier"). The problem is that someone has to make sure that the end-to-end bandwidth is sufficient for the services to work...

Maybe the government (FCC) can help here: force the connectivity providers (ILECs and Cable cos.) to interconnect with one another and other ISPs with sufficiently big pipes - the size can be based on the number of subs etc.

How about that? Now we have sufficient end-to-end bandwidth to make phone calls and do other stuff.

Fred - before you start pounding me with Eye-Pee's lack of QoS features and how that is going to prevent me from making a phone call regardless of how I dimension the links, please note that IP is here to stay and it is much much better to try and work with IP than trying to replace it with an "ideal" protocol. If that means using yet another hack like IP QoS bits or MPLS - so be it. But that is still better than trying to reinvent the wheel.
fgoldstein
50%
50%
fgoldstein,
User Rank: Light Sabre
12/5/2012 | 3:29:04 AM
re: WiMax Guide
And again, perhaps you didn't notice every time the FCC has rules requiring things like UNE-P the 8th circuit knocks it down. You can complain that the FCC should do a better job of defending the rules, but the 8th circuit is in the middle of this as well.

The 8th? They were active in the 1990s, but the latest rounds weren't there. The TRO itself -- adopted two years ago and the remanded text is still not out -- was in part a remand from the DC Circuit. The TRO was appealed to the 2d, 9th and DC Circuits, IIRC, and consolidated in the DC, which remanded it. The DC Circuit, like the 8th, has been very pro-ILEC, while the 2d and 9th have been pro-CLEC. The majority of this Commission did not want the UNE rules to stand -- I wonder where the now-resigning Powell turns up -- so they didn't appeal, and while the CLECs did appeal, the FCC ended up speaking for them. That was a disaster, of course.

Non-ILEC loops (replacing UNE-L) are not required. UNE-L is still available. What is required is you invest some equipment. What is the point of a CLEC selling the RBOCs service? What value does it add to the network? Are you declaring this is the way to fund a business?

You're spouting the Big Lie party line. First off, I'm not a big fan of Resale. I do have clients who do it and it is handy for certain markets, like credit-challenged subs who can't get ILEC service. UNE-P is not my main bag either but it does have some advantages. Mainly, it keeps the rates honest. VZ-NY was charging what, 15c+ per minute for a "local zone" call from NY to LI, when competitive "LD" rates to China were much lower. UNE-P forces monopolies to have a cost-based price available, even if only via an intermediary. I see that as good, yes.

But the big lie is that UNE-L is always available. It is still available hither and yon, but it's getting harder and harder to do. There are various exceptions to UNE-L now. For one thing, if the ILEC pulls FTTP, FTTC, or anything that Kevin Martin thinks is broadbandish enough, then the ILEC can cut the UNE loop to subs in that area. So DSL investments in that CO are lost. Not that the ILEC has to even offer high-speed common carriage -- they can offer Universal DLC low-grade POTS, but if it's via an FTTC SLC, then the loop is cut. That's part of the TRO.

Second, in order to use UNE-loop, you need to get into the CO. The FCC just took away interoffice facilities between big COs, and all UNE entrance facilities. Given the preposterous gouge that is Special Access, in many cases the cost of getting bandwidth into the CO is just insane. The original TRO allowed the states to determine that a route was unimpaired if the ILEC demonstrated that there were competitive alternatives. The DC Circuit struck that down, claiming irrationally that "route by route" was an improper analysis. (Why, there's a ton of bandwidth to 33rd Street Manhattan, so why do you need UNEs to Bayside Queens? Reroute the bus?) The FCC's TRO Remand Order sticks to route-by-route anyway, but instead of demanding a showing of competition on the route, it uses a pair of proxies (either/or) for competition, and assumes that if either proxy condition is met, then that route is competitive, even if there is no other carrier! Totally insane.


You want to define the competition as the telecom wire when it is obvious that the loss to wireless and cable creates a situation of competition.

Wireless creates no competition at all, except for very low usage voice -- maybe 5%. It's farcical to assume that a cell phone is competition for wired telephony or wired ISP access. WISPs can only operate in limited geographies.

Cable provides no competition at all at wholesale, and only limited competition at retail. A duopoly is not a competitive market. The FCC said that many times, before Powell at least. The ILEC-cable duopoly, minus common carriage (which the Bells have petitioned to do away with), provides exactly zero avenues of mass market access for independent ISPs. And VZ Online is already shamelessly blocking incoming email from most of the world. This is what you want for everyone?

Better to follow the rest of the world's model, using showings of Significant Market Power to maintain regulation. And better yet to spin off the entire ILEC Outside Plant to a neutral carriers' carrier, allowing true competition. And no, that wouldn't require UNE-P using the old ILECs' switches. But anyone could then sell services over the Broadband plant, paying the LoopCo its economic (including a fair rate of return) cost.
paolo.franzoi
50%
50%
paolo.franzoi,
User Rank: Light Sabre
12/5/2012 | 3:29:07 AM
re: WiMax Guide

fgoldstein,

32k x 32k DS0 switches are nothing new. Now being able to buy a 40 Gb/s switch chip for say sub-$300 is new. Thats a 200k x 200k voice switch. Bigger ones are available as well.

And again, perhaps you didn't notice every time the FCC has rules requiring things like UNE-P the 8th circuit knocks it down. You can complain that the FCC should do a better job of defending the rules, but the 8th circuit is in the middle of this as well.

Non-ILEC loops (replacing UNE-L) are not required. UNE-L is still available. What is required is you invest some equipment. What is the point of a CLEC selling the RBOCs service? What value does it add to the network? Are you declaring this is the way to fund a business? And hello, every cable company built a network as did every wireless carrier. So, 2 other competitors for virtually every human being in the US built a network. Is it that selling the exact same service that an incumbent provider does is simply a horrible business idea (From the Grisham book "The Rainmaker" - "You must be stupid, stupid, stupid.")? The other 2 facilities based carriers made different services and got money to build them.

You want to define the competition as the telecom wire when it is obvious that the loss to wireless and cable creates a situation of competition.

seven
fgoldstein
50%
50%
fgoldstein,
User Rank: Light Sabre
12/5/2012 | 3:29:08 AM
re: WiMax Guide
In terms of technology selection, I think the FCC is simply watching what everybody is doing and is a follower here. The semiconductor world and their investments dictate the range of solutions that we can choose from. So, the push to IP and Ethernet technologies comes from that investment. Nobody is building 90nm ATM or circuit switching engines. So, this choice is made whether the FCC wants to do anything or not.

It's not their job to pick winners, whether or not they think that the trend is going that way. It's their job to let progress continue. And don't write off TDM or ATM. Look for instance at the Zarlink MT90869 or ZL50075 to see what circuit switching engines look like nowadays. It's not your father's 1ESS. I'm not sure about the process size, but a 32k by 32k matrix of 64kbps calls in one chip, eating 2 watts at 3.3/1.8 volts, is pretty efficient.

The problem with IP is that it's extremely cool, which is a marketing thing. (Or "k3w1" for the k1dd13z.) The fact that it's a mid-1970s lab hack is lost in worshipfulness. And because it's worshipped -- this being a country full of fundamentalists who are habitually seeking the One True Answer, whether in religion, investment, or regulation -- nobody is looking to do any better. That's a problem. But it's a bigger problem when the government essentially bans progress by favoring IP over alternatives. Doing Zarlink little good, of course, trying to sell its TDM engines. And doing little good for those who have better ways to ship data around, too.


In terms of how the government should involve itself, this is also not an FCC decision. The FCC is trying to implement the rules it was given to live with. It is trying to do so in the middle of a massive financial and technological restructuring of the industry it is supposed to manage. Both the financial and technological changes are outpacing the regulatory changes that it is making.


The FCC is not following the rules it was given -- Powell's actively calling for the repeal of the Telecom Act and its replacement by a propertarian model. Such a scheme would be financially disastrous. The point of the Telecom Act was to bootstrap competition by allowing a competitor to start with UNEs, even all-UNE (Platform), and then phase over to its own facilities where and when it became economically feasible. For most CLECs and ISPs, non-ILEC loops are never going to be feasible. WiMax won't replace the copper, nor will BPL, nor will UWB, nor will any other pi-in-the-sky. Powell's holding them up to the public to cover his desire to massacre those sectors, while giving the Bells the kind of unregulated monopoly that a Mobutu Sese Seko or Ferdinand Marcos only granted his best cronies.
fgoldstein
50%
50%
fgoldstein,
User Rank: Light Sabre
12/5/2012 | 3:29:09 AM
re: WiMax Guide
Our country needs a communications infrastructure that supports modern data networks.

Duh.

The attempts of retrofitting these infrastructure needs on top of the legacy voice or videeo infrastructure isn't working technically, business wise, nor regulatory wise.

Defend that. You obviously don't do high-speed data across voice-grade dial-up. But as attendees to LR's SONET seminars know, there's huge growth in Packet-over-SONET, which allows Ethernet+IP and TDM to coexist comfortably. Lots of technological progress there too. And cable modems do a bang-up job for consumer access -- I have clients planning big investments in HFC for triple play.

The markets aren't working either. Markets rarely build infrastructures such as roads, electrical systems, sewage systems, etc. on their own.

I agree -- there's a natural monopoly down at the pipe layer, but that has nothing to do with eye pee. That's copper or glass.

Government has to step up to the plate and use its power wisely. The current trend of leaning on free market rhetoric and ideology is just an excuse for abdication of responsibilities. This needs to change. We need to start demanding better because we deserve better.


What does that mean? Use its spending power to subsidize things? Fat chance, at least at the federal level, except for some outlying USF. The Powell method of using power is to kill off competitors so that monopolies make huge profits which, he pretends, they'll use to invest more. As if monopoly spurred investment more than competition, which is preposterous.

PS. Access networks need to be digitized and packetized. Eye Pee seems to do the job nicely.

Proof by vigorous assertion? Eye pee does it badly -- no QoS, huge overhead, cockamamie addressing (none, really, just a screwed up multi-layered naming system) and (as a result) routing. It's better than, uh, what? DECnet Phase II with Poor Man's Routing? Bisync? X.25? Gimme a break -- we can do better. Stopping progress at the Model A stage of technology is a bad idea, and comparing IP to the Ford Model A (much better than the T!) is, perhaps, a bit of an insult to Ford.
paolo.franzoi
50%
50%
paolo.franzoi,
User Rank: Light Sabre
12/5/2012 | 3:29:11 AM
re: WiMax Guide

I think we have 3 different slants trying to achieve the same idea. Modernizing the US telecom infrastructure. I would say that many of agree in some areas and disagree in others.

In terms of technology selection, I think the FCC is simply watching what everybody is doing and is a follower here. The semiconductor world and their investments dictate the range of solutions that we can choose from. So, the push to IP and Ethernet technologies comes from that investment. Nobody is building 90nm ATM or circuit switching engines. So, this choice is made whether the FCC wants to do anything or not.

In terms of how the government should involve itself, this is also not an FCC decision. The FCC is trying to implement the rules it was given to live with. It is trying to do so in the middle of a massive financial and technological restructuring of the industry it is supposed to manage. Both the financial and technological changes are outpacing the regulatory changes that it is making. I would like to see it move faster, but any movement pisses off someone.

In terms of cable CLECs, as long a Interconnection and LNP exist a cable company acts like a facilities based CLEC. I think this is part of the problem in particular with the UNE-P carriers. UNE-L carriers at least put transmission and switching investment into the network, even if they do not put access investments. UNE-P carriers invest no capital to offer the service. Now the problem is under a rate cap structure, the role of the regulator is to spur investment. That is why UNE-P was such a horrible idea - the RBOCs did not invest and the UNE-P carriers did not invest. The consumer won in the short term (cheaper phone rates) but nothing was added to the network.

As to the government getting involved in investment, I think that is so highly unlikely at the federal level that it is not worth talking about. The current administration has communications at the bottom of its agenda. As to municipal and state investment, these are local decisions and the structure to work these within already exists (look at the co-ops within the IOC segment). The problem is that solving the access problem does not really help, because it simply shifts the problem. After the access is completely resolved, there will need to be a metro and then a long haul retrofit. Problem is with all of this, is that there is no economic model (other than faith) that shows that broadband or super broadband (call that 100Mb/s per subscriber) has any economic impact at all. The technology is simply too new. Teledensity can be directly traced to economic growth. Broadband density...no idea.

seven
rjmcmahon
50%
50%
rjmcmahon,
User Rank: Light Beer
12/5/2012 | 3:29:12 AM
re: WiMax Guide
I do agree with you on what Powell is doing, but I think it is both technically wrong and illegal. He is pursuing a totally misguided industrial policy, playing technological favorites at a time when the government's role is to be neutral and let the market iron it out.

Our country needs a communications infrastructure that supports modern data networks. The attempts of retrofitting these infrastructure needs on top of the legacy voice or videeo infrastructure isn't working technically, business wise, nor regulatory wise. The markets aren't working either. Markets rarely build infrastructures such as roads, electrical systems, sewage systems, etc. on their own.

Government has to get involved. The question becomes exactly who in government can and will help? The government I believe should participate is the idealized one we preach about. It's the government "of the people, for the people, and by the people." Unfortunately, most of the state and federal representatives and bureaucracies, currently in power, spend their efforts acting as a sausage factory spitting out special interest legislation protecting the status quo. This will never get the job done but only leads to the US falling further behind. And yes, we are falling behind.

Government has to step up to the plate and use its power wisely. The current trend of leaning on free market rhetoric and ideology is just an excuse for abdication of responsibilities. This needs to change. We need to start demanding better because we deserve better.

PS. Access networks need to be digitized and packetized. Eye Pee seems to do the job nicely.
fgoldstein
50%
50%
fgoldstein,
User Rank: Light Sabre
12/5/2012 | 3:29:16 AM
re: WiMax Guide

One thing you are not noticing is that even when the FCC tries to protect what it can for an IXC, the 8th circuit is saying that the rules do not meet the 1996 Act and is forcing the FCC to move the rules again.


The different Circuits have done different things over the years. The 8th was the one that originally, in the Iowa Utilities Board vs. AT&T case, overturned the FCC's original Rules that implemented the Telecom Act. However, the 8th was largely overturned by the Supremes. That is when UNE-P began -- it was the FCC's rule, and it was approved by the Supremes. The 9th, 2d, and I think other circuits have been friendlier to the CLECs. The DC Circuit at present is highly hostile to both the FCC and CLECs, in kind of a hate-triangle relationship.

However, it's pretty clear to me that the DC Circuit's decision of last February did not comport with either the Record of the case, which it mis-cited repeatedly, or the FCC's own rules in Iowa Utilities Board. The FCC defended its own rules with all the enthusiasm of a condemned prisoner working for the prosecution. Sort of like Bre'r Rabbit, if you remember the tale, "Pleasea don't take away those UNEs!" And the Solicitor General took his cue. Remember, Powell's closest advisor at the time, Jane Mago, was the wife of BellSouth's VP-Regulatory. Fair? No, it stinks to high heaven, but then what in this regime doesn't?

I think Powell is interested in moving the country to an IP based infrastructure with fiber access and is pushing rules in this favor. I wonder if AT&T had envisioned this, whether they would have sold their cable operations. With these, they did not need a CLEC - they could be offering local VoIP connections over cable today. AT&T has now sold AT&T Wireless as well. I think they hoisted themselves up on their own.


I do not disagree that AT&T has been mismanaged. They paid too much for their cable properties and didn't manage them well. However, they would still need to be a CLEC in order to offer local dialtone over the cable -- I did get my dialtone from them (previously MediaOne, now Comcast) and they are a CLEC. There is some doubt about whether a parasitic VoIP provider needs to be a CLEC, but it needs to have a CLEC to provide it network connections. Your typical Vonage subscriber is actually using Focal or Paetec or some other CLEC.

I do agree with you on what Powell is doing, but I think it is both technically wrong and illegal. He is pursuing a totally misguided industrial policy, playing technological favorites at a time when the government's role is to be neutral and let the market iron it out. Had Powell been in charge during the early years of digital cellular and PCS, he might have picked a favorite technology and given it special treatment, the way the Europeans did with GSM. I think time proved Hundt's approach right. The Telecom Act does not pick favorites. Powell, of course, has called for the Act to be rewritten, presumably so it can be explicitly favorable towards the incumbents, following Gingrich's extremist propertarian philosophy.

I think I understand both IP and the PSTN a whole lot better than Powell does -- he's just a lawyer, after all, and I'm a network geek, the expert witness on the lawyers' team. Unlike most geeks I know, I've actively worked on both sides (IP and telephony), and as such much of what I do is translate between the two (they really don't understand each other). I understand why his "IP under alles" vision, which was handed to him by some equipment vendors looking for an excuse to sell new stuff into a saturated market, is not a good idea. But it's hard to sell reality when the press and Wall St. promoters (the spirit of Grubman lives on) are flogging buzzwords like "IP", "WiMax", "fiber", "wireless", "seamlessly enabled enterprise scalable solutions", etc. -- and of course has not a klew what they're talking about.
paolo.franzoi
50%
50%
paolo.franzoi,
User Rank: Light Sabre
12/5/2012 | 3:29:18 AM
re: WiMax Guide

1. Depends on your implementation. Big cell covering SF might have say 1,000 users on it. This would give an apparent rate of 40Kb/s per sub if they were all transmiting. Similar to what a DSLAM does today.

2. Yep
3a) Depends on what services your trying to offer.
3b) See 3a.
3c) Who said it was easier to maintain? It is less cost to construct. Maintain is a different thing.

4) The 2nd 3 - Ever had 2 FM (or AM) radio stations interfere with each other on your radio? Transmitters capable of the power to go 30 - 50 km are not the same power as those to go 30 - 50 meters. By the way depending on your placement of devices, 802.11, 2.4 GHz cordless phones, and Microwaves can interfere with one another. Generally not enough to lose connectivity on the 802.11, but enough to lower the connect rate.

4a) So, the big deal is you place a tower 3km from my tower and we both have 30km reach. If we choose the same band then whoops...no workie.

4b) If they all chose different frequencies this would work. If not, then it won't.
4c) The receiver on your tower will receive my transmissions. This will generally swamp out any users further away than I am. Even if theuse is closer, noise is added to the signal - raising the bit error rate.

This is why the FCC allocates radio stations in an area to assure minimal interference. This has nothing to do with WiMAX and is a general wireless issue.

seven
vk_lite
50%
50%
vk_lite,
User Rank: Light Beer
12/5/2012 | 3:29:19 AM
re: WiMax Guide
I'm trying to better understand this 802.16 space (and in some cases wireless period). I thank those who offer some constructive advice in this area.

Just a few questions.

1) Would 802.16 offer a kind of "cell" coverage, up to the distances mentioned (ie 30-50 km), or is this distance only achievable with directional antenas, focused on an access point with a shorter range technology behind that.

I'm confused because (often just "because"), frankly, from the literature (some have called it hype), I would expect that this is a cell type coverage, which should cover essentially the city of San Francisco as 1 big cell (which would require no mobility), with obvious degradation of bandwidth depending on the number of users and the load or bandwidth intensive operations they are actually performing. If this were the case, sharing the load with more 802.16 BTS's would solve this.

2) 802.16 could be used in both licensed and unlicense spectrum, right?

3) In the licensed spectrum:
a) Which is cheaper to deploy - an 3G+ network (like HSDPA), or a WiMax network?
b) Which is cheaper to maintain, 3G+ or WIMAX
c) Why is wireless cheaper to maintain than cable or other fixed line communications.

3) Now remember this is "lite" reading. So this may not make sense, but here is my issue. I have a WiFi set up at home (encrypted), and so does half my neigbourhood. I have never had a problem with interference. I can always log onto my network, and seem to have the bandwidth I expect.

a) Having said this, what is the big deal with interference in the unlicense spectrum?
I never seem to run into it, meanwhile, half the things in my house are wireless or use the same frequencies (microwave, cordless speakers, cordless phones...)

b1)Is there an issue with, for example, twenty 802.16 providers in the same area?
b2)What is the problem?
lighten up!!
50%
50%
lighten up!!,
User Rank: Light Beer
12/5/2012 | 3:29:21 AM
re: WiMax Guide
To deliver WiMax appear to have been grossly understated. The "Hypesters" are at it again with promises to deliver hundreds of megabits over 10's of miles. All these fancy antennas, demodulation schemes and OFDM, comes at a price that's not going to be competitive with existing DSL/Cable. Even if the prices came down, does this stuff really work in real life? What real data does anyone have to substantiate all these claims by WiMax? So far I have only seen a whole lot of marketing hype and very little tangible proof. Having been burned by the Lambda Hype, I don't believe all the B.S. that is coming out of WiMax. If it sounds too good to be true then it probably is...
paolo.franzoi
50%
50%
paolo.franzoi,
User Rank: Light Sabre
12/5/2012 | 3:29:21 AM
re: WiMax Guide

Jim Smith,

I agree with you. I view the market you described (basic connectivity) as a great vehicle in less industrialized nations. There, I DO believe in the potential of a mass market.

fgoldstein,

One thing you are not noticing is that even when the FCC tries to protect what it can for an IXC, the 8th circuit is saying that the rules do not meet the 1996 Act and is forcing the FCC to move the rules again.

I think Powell is interested in moving the country to an IP based infrastructure with fiber access and is pushing rules in this favor. I wonder if AT&T had envisioned this, whether they would have sold their cable operations. With these, they did not need a CLEC - they could be offering local VoIP connections over cable today. AT&T has now sold AT&T Wireless as well. I think they hoisted themselves up on their own.

seven
fgoldstein
50%
50%
fgoldstein,
User Rank: Light Sabre
12/5/2012 | 3:29:23 AM
re: WiMax Guide
What's your opinion on the Intel and AT&T collaboration in the area of WIMAX and VoIP? Have these guys been boxed out by the feds and hence this unlicensed wireless approach is the best (only??) access technology they can hope for? Why didn't the feds take their interests into account when creating the "industrial policy?"


Intel's a beneficiary of the industrial policy -- for them, WiMax is an incremental source of chip sales. They're the main cheerleader, it seems, for WiMax, and thus least interested in reminding people to "pick one".

AT&T is the hated victim of this FCC. It's almost as if Powell remembers the old AT&T Bell System days and is punishing them for it. Of course he's too young, and it's the RBOCs who have inherited the bulk of Bell System monopoly genes. So it's more like young thugs saying "let's roll the old coot". Since AT&T sold its cable operation and its wireless operation, it has essentially no way to reach potential subscribers without going through the Bells. (TCG lights a few thousand buildings, if that.)
So out of desperation, they will talk about WiMax. Probably more for the sake of Wall Street than anyone else, though -- the banAnalysts on The Street don't understand "pick one". AT&T failed once on WLL with Angel and knows how hard it is.
fgoldstein
50%
50%
fgoldstein,
User Rank: Light Sabre
12/5/2012 | 3:29:25 AM
re: WiMax Guide
Why might we need WiMax?

The answer, flat out, is industrial policy, government favoritism towards specific industrial sectors. It's the antithesis of classical free-market economics, but it's all the rage in Bushland, where they have picked favorites. The ILECs are favorites. Their traditional suppliers are favorites. ISPs, CLECs, and scrappy little suppliers who don't play the Telcordia-spec game on a large scale are the losers.

However, that is so blatant a power grab that the Bush-Powell FCC needs cover. So they preach "intermodal competition", something not even mentioned, AFAIK, in the Telecom Act. And since the only two modes with any bandwidth are cable and the ILEC wire (especially if they get around to putting a little glass in), they need somewhere to push everyone else.

Hence wireless: A third medium that theoretically allows consumers to get their Internet service via someone other than the ILEC or cable company. Of course it's rarely suitable or competitive, and the total spectrum capacity is extremely limited -- there's lots of sand but only one spectrum. But it's a cover story that they love to trot out every time they cut off CLEC access to loops, or cut off ISP access to ILEC telecom networks. And hey, if there were infinite money with no need for profits (think 1998, but ten times more so), and the laws of physics could be waived as easily as the Geneva Conventions are ignored, then indeed there might be more sales of "high tech broadband" wireless equipment.
rjmcmahon
50%
50%
rjmcmahon,
User Rank: Light Beer
12/5/2012 | 3:29:25 AM
re: WiMax Guide
Why might we need WiMax?

The answer, flat out, is industrial policy, government favoritism towards specific industrial sectors. It's the antithesis of classical free-market economics


What's your opinion on the Intel and AT&T collaboration in the area of WIMAX and VoIP? Have these guys been boxed out by the feds and hence this unlicensed wireless approach is the best (only??) access technology they can hope for? Why didn't the feds take their interests into account when creating the "industrial policy?"

http://nwm.mobilepipeline.com/...

AT&T, Intel Explore WiMAX Together
Courtesy of TechWeb.com

As the Federal Communications Commission prepares to hammer another nail in the coffin of AT&T's efforts to compete in local-landline telephony markets, the long-distance company has begun to reveal the fruits of its R&D deal with Intel. The companies are focusing on WiMAX for much of their cooperative efforts.

AT&T is assisting Intel to develop the semiconductor firm's Roseville family of WiMAX chips scheduled for release next year. As AT&T has retreated from consumer landline-telephone markets, it has increasingly directed its attention to voice over Internet protocol (VoIP) and WiMAX, the wireless wide-area technology.
jim_smith
50%
50%
jim_smith,
User Rank: Light Beer
12/5/2012 | 3:29:27 AM
re: WiMax Guide
brookseven,

I basically agree with your analysis, except that some people only want connectivity - they don't want video over the 100Mbps pipe.

I'm talking about a niche market. WiMAX can allow you to serve these customers very cheaply and make a profit.

JS.
paolo.franzoi
50%
50%
paolo.franzoi,
User Rank: Light Sabre
12/5/2012 | 3:29:28 AM
re: WiMax Guide

Again, I point to the US.

The RBOC network may be expensive and wrong, but it has two things going for it:

1 - It exists. Only thing that needs to be added are linecards when incremental subscribers exist. Basically, there is 95%+ broadband coverage in the US. The places that there is no broadband are extraordinarily rural (call it pop density of 1 person per square mile country).

2 - Its commodotized. I think there is a belief that suddenly a 1Mb/s/sub infrastructure is going to be replaced by a single WiMAX base station in say San Francisco. More likely this will be replaced by 1 on every corner. Thus, the buildout is nowhere near as cheap and easy as stated.

So again, I see no case for this technology in the US outside of mobility. In areas where teledensity went wireless (the 3rd world as an example), WiMAX makes sense. The US (and most of the industrialized world) simply does not have a connectivity problem. This problem is solved. We are now moving onto a bandwidth increasing per subscriber problem (thus all the talk of deep fiber and architectures of sustained 100 Mb/s/subscriber). Wireless simply is too noisy of a channel to support these rates over any distance with a wide coverage area.

seven
djkstra
50%
50%
djkstra,
User Rank: Light Beer
12/5/2012 | 3:29:28 AM
re: WiMax Guide
>Basically, there is 95%+ broadband coverage in the >US.

>WiMAX makes sense. The US (and most of the >industrialized world) simply does not have a >connectivity problem. This problem is solved.
>We are now moving onto a bandwidth increasing per >subscriber problem (thus all the talk of deep >fiber and architectures of sustained 100 >Mb/s/subscriber)

Not sure how accurate these statement are.. However, I agree that notion that Wimax is not quite necessary technology exists predominantly.
I beleive Wimax is not just meant to solve bandwidth problem which it does to some extent.. To-Be-Mobile completely can be 10X factor on the fast paced businesses..
djkstra
50%
50%
djkstra,
User Rank: Light Beer
12/5/2012 | 3:29:29 AM
re: WiMax Guide
The question is why would I want to build yet another broadband High Speed Internet infrastructure. ???

Brookseven,

Internet service market is under-leveraged. According to Mckinsey, every month 35Million users wordwide are migrating to broadband.. Last-mile problem is huge to solve.
jim_smith
50%
50%
jim_smith,
User Rank: Light Beer
12/5/2012 | 3:29:31 AM
re: WiMax Guide
The question is why would I want to build yet another broadband High Speed Internet infrastructure.

Because the present RBOC infrastructure is too costly to maintain.

For example:

SBC has approx. 150K employees for approx. 50M subs. Comcast has approx. 70K employees for 20M subs. Essentially a 3:1K ratio.

Cingular (after merger) has 60K employees for approx. 60M subs. A 1:1K ratio.

Why? Partly because wireless is a "cleaner" technology. I know that Cingular leases its core network from SBC and that is one reason why SBC has more employees, but it still cannot account for the huge gap.

Bottom line: a wireless infrastructure lowers the capex and opex. Mobility is just the icing on the cake.

You can't beat that.
paolo.franzoi
50%
50%
paolo.franzoi,
User Rank: Light Sabre
12/5/2012 | 3:29:32 AM
re: WiMax Guide

I guess my view is a bit different. The question is why would I want to build yet another broadband High Speed Internet infrastructure. In most places in the US (not ALL but MOST) DSL and Cable Modems are available. In virtually the entire US, either DSL or Cable is available.

So, introducing this technology would have to have an application that did not work today or in places that service was not available today. To build out a network to support commodity High Speed Internet seems highly unlikely, since there are already embedded competitors. The question is what is that application that requires a WiMAX buildout if you already have a significant penetration of broadband deployed. Mobility seems like a logical answer.

seven
djkstra
50%
50%
djkstra,
User Rank: Light Beer
12/5/2012 | 3:29:34 AM
re: WiMax Guide
Question that has been unanswered in minds of carriers in West is whether market is ready for Wimax to the extent where they could gain a potential momentum in selling Wimax services and migrate to this technology.
Post bubble effects have changed ISPs' attitude towards technology and made them measure technology critically in terms of returns on equity, capital, and assets.They are to answer whether they intend to continue build up
debt based capital structure where take the risk of deploying new technology even with larger debt.

Whereas, in my personal opinion, Wimax seems to be perfect fit in eastern countries like India and China who are still in the process of expanding carrier networks and opt for the latest technology. Market is these countries price driven and at the Wimax does not answer whether it can be cost effective soln to ISPs in these countries.

Time to watch..
ggoyal
50%
50%
ggoyal,
User Rank: Light Beer
12/5/2012 | 3:29:36 AM
re: WiMax Guide
Hi,

What is the maximum inter-node distance/range in a WiMAX mesh mode theoretically and practically?

Has anybody compared the mesh part of the two technologies (WiFi & WiMAX?

vinay
bdixon23
50%
50%
bdixon23,
User Rank: Light Beer
12/5/2012 | 3:30:11 AM
re: WiMax Guide
In the section, "In these [unlicensed 2.4 and 5 GHz] bands, WiMax operates at a much lower wireless power and has a very limited range (roughly the same as for WLAN, which also uses these frequency bands)," do you mean to say that the effective range of WiMax in unlicensed bands is no greater than Wi-Fi?
Peter Heywood
50%
50%
Peter Heywood,
User Rank: Light Beer
12/5/2012 | 3:30:33 AM
re: WiMax Guide
This note came in from someone working for a major carrier, who requested anonymity:


QUOTE
One question that the article did not answer was how many simultaneous users a typical WiMax implementation could handle. WLAN can handle up to 230 simultaneous users per node, at lower bandwidth speeds of course, but what is the theoretical limit of users and lower bandwidth constraints and what would be a better quality of service from a bandwidth and simultaneous user perspective (typical implementation)
UNQUOTE
Peter Heywood
50%
50%
Peter Heywood,
User Rank: Light Beer
12/5/2012 | 3:30:33 AM
re: WiMax Guide
Tim Hills has taken a crack at addressing this issue:


As far as I understand it, from digging in the material I have:

-----

IEEE Std 802.16.2T-2004 (Revision of IEEE Std 802.16.2-2001)

IEEE Recommended Practice for Local and metropolitan area networks

Coexistence of Fixed Broadband Wireless Access Systems


Table 22-Parameters for 3.5 GHz systems with a PMP (point-to-multipoint) architecture (this is a brief sample only - the whole table is much bigger)

Number of terminal stations per megahertz per transceiver per cell - Up to 70.
Number of cells in a system - 1 to 25 (typical range).
Channel bandwidth - 1.5, 3, 6, 12, 25 MHz (North America). 1.75, 3.5, 7, 14 MHz (Europe) (use 7 MHz for coexistence calculations) Typical sector arrangements and frequencies - Typically 4 sectors per cell,
4 frequencies. Vertical and horizontal polarization both used. Some systems will use AAs, pointing at individual users. FDD and TDD used.

-----

This suggests that a single cell transceiver in a 3.5GHz system could support a theoretical maximum of between 105 and 1750 terminal stations per frequency channel for the range of channel bandwidths contemplated in North America. Multiply by 4 to give a cell maximum for the typical 4-sector 4-frequency cell - so 420 to 7000 terminal stations per cell.

But there are lots of complications arising from the highly specific nature of radio propagation in a given environment and the need to avoid interference between different WiMax service providers in the same area. And mesh WiMax architectures have different parameters - and so on.

And I am very prepared to be corrected. Perhaps someone else is better qualified to answer?
Peter Heywood
50%
50%
Peter Heywood,
User Rank: Light Beer
12/5/2012 | 3:30:34 AM
re: WiMax Guide
We would prefer you post messages about this article on this public board, but if you want to send a private note about it, please email [email protected] and include "wimax guide" in the subject field.
Peter Heywood
50%
50%
Peter Heywood,
User Rank: Light Beer
12/5/2012 | 3:30:35 AM
re: WiMax Guide
In explaining Wimax, we need to explain its potential, and I guess this is what you're describing as over optimistic.

Take a closer look: We actually start the report saying that 2005 could be the year when Wimax sinks or swims and then go on to quote a poll that indicates that the industry is evenly split on which way it will go.

I don't call this overly optimistic. I call it giving both sides of the argument.

hills@lightreading.com
50%
50%
[email protected],
User Rank: Light Beer
12/5/2012 | 3:30:35 AM
re: WiMax Guide
I'm not sure I would describe the report as "overly optimistic". There is no intention to suggest that WiMax will sweep away wired/fibred broadband - clearly, it won't. No freespace radio system can match the space-division multiplex capabilities of wires and fibres (nor the THz frequencies of optics, for that matter), without even considering all the propagation and similar issues the posting points out.

The point is rather that the broad industry (telecoms, IT and consumer electronics) has understandably got very interested in the idea of wide-area wireless broadband access as a new capability, service and revenue opportunity - and so there is a lot of pressure to try to make this one fly. How it will turn out is anyone's guess at the moment (the report points out a fair number of issues), but it is all part of the technical innovation needed to keep telecoms advancing. It seems a bit unfair to write off a technology just because it doesn't offer the peformance of another technology - aircraft have not yet replaced cars.

Tim Hills
fgoldstein
50%
50%
fgoldstein,
User Rank: Light Sabre
12/5/2012 | 3:30:40 AM
re: WiMax Guide
A lot of the press reports on WiMax, and I'm afraid including this one, are apparently written by people who are somewhat familiar with data networking, but are not skilled in the art of microwave radio. So they overstate the case, making it sound simpler than it really is. WiMax is good technology: It is a smarter spec than, say, WiFi (which to be sure is optimized for a lower price point and reflects implementation capabilities of some years ago) and it may well be the best hope for a mass-produced data radio. But it's still a radio.

Think of cellular. CDMA is, like WiMax, very efficient, even more so than analog or GSM. But in most areas it either has lots of cell sites or lots of dead spots. WiMax is similar: Path loss cannot be overcome by clever modulation or even fancy signal processing. Shannon's Law applies with a vengeance; it is not trumped by Moore's Law.

WiMax will find markets; it will find applications. But it won't run DSL and fiber out of business. It won't be the magic bullet that creates Gilder's phantasmagorical Telecosm. It won't widely replace the local loop. It will require careful path engineering, or it will work on an opportunistic basis, like many hot spot services do now.
Featured Video
Flash Poll
Upcoming Live Events
September 17-19, 2019, Dallas, Texas
October 1, 2019, New Orleans, Louisiana
October 2-22, 2019, Los Angeles, CA
October 10, 2019, New York, New York
November 5, 2019, London, England
November 7, 2019, London, UK
December 3-5, 2019, Vienna, Austria
December 3, 2019, New York, New York
All Upcoming Live Events
Partner Perspectives - content from our sponsors
Transform Beyond Borders to Lead the Innovation
By Ben Zhou, CEO, Whale Cloud
Reject Limits. Build the Future.
By David Wang, Huawei
China Telecom & Huawei Jointly Complete the World's First End-to-End 5G SA Voice & Video Call
By Jay Liu, Senior Marketing Manager, Cloud Core Product Line, Huawei Technologies
All Partner Perspectives