Kuro5hin.org: technology and culture, from the trenches
create account | help/FAQ | contact | links | search | IRC | site news
[ Everything | Diaries | Technology | Science | Culture | Politics | Media | News | Internet | Op-Ed | Fiction | Meta | MLP ]
We need your support: buy an ad | premium membership

[P]
802.16: Medium distance wireless networking that could change the world?

By bslade in Technology
Sun Dec 12, 2004 at 03:11:35 PM EST
Tags: Technology (all tags)
Technology

The (sort of) new IEEE 802.16 Metropolitan Area Network standard (MAN) promises to break the cable modem/DSL monopoly on the "last mile" of Internet connectivity to peoples' homes.

With ranges up to 30 miles and speeds up to 70 mbps it seems like a pancea.

While I suspect the actual usable speeds and distances will be considerably less, my real question is where is it? This standard has been in the works for years.


Here's a summary of what 802.16 is from an old Intel 802.16 press release:

The 802.16 standard, which the IEEE modified this [past] January [2003] in its 802.16a amendment covering the 2 GHz to 11 GHz frequencies, is a wireless metropolitan area network technology that will ... provide a wireless extension to cable and DSL for last mile broadband access. It provides up to 31 miles of linear service area range and allows users to get broadband connectivity without needing a direct line of sight to the base station. The wireless broadband technology also provides shared data rates up to 70 Mbps, which is enough bandwidth to simultaneously support more than 60 businesses with T1-type connectivity and hundreds of homes with DSL-type connectivity using a single sector of a base station. A typical base station has up to six sectors.

And some discussion of the implications of 802.16 can be found in a non-online article from Technology Review titled "Why WiMax?" (11/04 Vol. 107, No. 9, P. 20; Roush, Wade) [Note, WiMax is a consumer friendly name for the IEEE 802.16 standard]:

The forthcoming Worldwide Interoperability for Microwave Access (WiMax) metropolitan-area wireless communication standard is expected to put Wi-Fi in the shade. Wi-Fi can transmit signals across up to 100 meters indoors and 400 meters outdoors, but WiMax boasts a maximum transmission range of 50 kilometers at a peak data transfer rate of 70 Mbps.

Furthermore, once industry consensus is reached on such details as WiMax data encryption, [regulated and unregulated] frequency allowances, and multiple-user frequency access, companies will be able to mass-produce WiMax-enabled chips and make WiMax receivers affordable to consumers; the end result could be the replacement of current ISPs with broadband Internet connectivity.

WiMax promises to facilitate wireless communication for new small and mid-sized businesses, the construction of mobile-computing hot spots in areas that lack phone lines, and the expansion of broadband Internet access to impoverished regions. The instigator of the WiMax movement is Intel, which saw a need for Wi-Fi to develop into a carrier-like technology as well as use more, as-yet untapped frequencies. In addition to designing communications processors to exploit these frequencies and delivering the chips as samples to manufacturers, Intel is promoting the WiMax Forum as an industry organization for certifying WiMax-compliant equipment, and is making investments designed to demonstrate WiMax's profit potential through Intel Capital.

The high cost of building a WiMax transmitter network could complicate the technology's rollout. In addition, WiMax equipment manufacturers must address the challenge of achieving the economies of scale necessary for enabling WiMax hardware in the consumer price range.

A more recent and detailed discussion can be found at WiMax starting to make its move (nwfusion.com, June 04)

My comments:

Well, maybe finally, Intel's ventures into non-core technologies will pan out.

It seems like the planning for this is taking forever, especially since it should be a boon to the tech industry. But remember, all government coordination is bad, private enterprise is always more efficient (that's sarcastic folks).

Sponsors

Voxel dot net
o Managed Hosting
o VoxCAST Content Delivery
o Raw Infrastructure

Login

Poll
Do you think the 802.16 wireless technology will change the world?
o Yes 25%
o No 29%
o Whatever 25%
o It's all a CIA/NSA/NRO/DIA plot! More aluminum foil for my head! 20%

Votes: 44
Results | Other Polls

Related Links
o Intel 802.16 press release
o WiMax starting to make its move
o Also by bslade


Display: Sort:
802.16: Medium distance wireless networking that could change the world? | 74 comments (46 topical, 28 editorial, 0 hidden)
WiMax or 3G (2.50 / 2) (#4)
by Stuart Ward on Fri Dec 10, 2004 at 02:52:05 PM EST

You havent commented on the obvious comparisons between WiMax and 3G technology. Both use licenced spectrum as opposed to unlicenced for WiFi, so these will need to be brought to market by someone. Who will do this? well not the Cell (mobile) phone companies they have invested far too much in 3G technology. Perhaps the fixed line telcos and the broadband ISP will step up to the mark? This may happen in the USA but not in europe, all those billions spent on 3G licences only to be surplanted by a diffrent technology.


Could you please elaborate on this? (none / 1) (#24)
by mcc on Sat Dec 11, 2004 at 05:04:31 PM EST

You havent commented on the obvious comparisons between WiMax and 3G technology.

You seem to be familiar with this technology already. Could you please elaborate on what you mean by this? Since I do not know anything about either WiMax or 3G besides what is in this article, the comparisons are not very obvious to me.

Incidentally according to this WiMax/802.16 is capable of operating both in licensed and unlicensed frequencies.

[ Parent ]

WiMax or 3G (none / 0) (#59)
by Stuart Ward on Mon Dec 13, 2004 at 02:44:39 PM EST

Ok apologies I was not aware of the unlicensed spectrum option for WiMax. Not sure how that will compete with WiFi.
I was commenting on the comparisons with Mobile phone technology as developed for data. We already have extensive GPRS worldwide rollout, this gives 20k to 40k data rate in actual usage (theoretical 171.2k) and is even being adopted in the USA. EDGE pushes the typical rate to around 100k. Then as 3G is adopted around the world these data rates approach those touted for WiMax. The difference being that:
  • 3g standard is agreed and equipment available.
  • 3G has a sound security model enhanced from 2G
  • oh yes it can be used for voice calls as well
So after all the billions spent on 3G licences and technology who is going to back another wireless data standard (outside the standards mess of the USA). We will have to wait and see but given my bias obvious from above, I leave it to the proponents of WiMax to refute these arguments.

[ Parent ]
Re: WiMax or 3G (none / 1) (#51)
by bslade on Sun Dec 12, 2004 at 03:30:41 PM EST

Here's some relevant info from the 2nd page of the WiMax starting to make its move article (square brackets and italics added by me):

Unlike Wi-Fi, which has been deployed using primarily one band of spectrum (2.4 GHz) that is unlicensed nearly everywhere, WiMax is based on standards that allow for any frequency band between 2 GHz and 11 GHz [of which some bands are licensed and some are not]. The WiMax Forum is narrowing that by developing profiles for specific spectrum bands, says Francois Draper, vice president of sales and marketing at WiMax chip developer Wavesat and chairman of memberships at the WiMax Forum.

The problem is I don't understand exactly what "narrowing that [the 802.16 standard] by developing profiles for specific spectrum bands" means. Fewer frequency bands? Licensed and unlicensed? Different usage on different bands? What the heck is a profile?

At the very least, my statement that "Wi-Max is just a consumer friendly name for 802.16" is not true. Apologies.

So it sounds like Wi-Max is some consumer friendly subset. I will continue to research to figure out exactly what that means.

I hate standards that are so flexible someone has to create another standard to permit people to use the first standard.


Ben in DC
PublicMailbox at benslade.com (put 030516 anywhere in the subj to get thru)
"It's the mark of an educated mind to be moved by statistics" Oscar Wilde
[ Parent ]

Cell phone companies do invest on WiMax in Europe (none / 0) (#72)
by IsoToop on Tue Dec 14, 2004 at 04:36:18 PM EST

TeliaSonera will invest on WiMax in (Northern)Europe. Same time they build their own 3G network. TeliaSonera to help build first WiMax network in Scandinavia and Sonera 3G covers 20% of Finnish population

[ Parent ]
this will be good (1.00 / 2) (#9)
by auraslip on Fri Dec 10, 2004 at 07:46:35 PM EST

I'm moving in to a new house and i can't afford to pay for broadband. I was worried that no one near enough would have a WiFi network for me to steal. But 30 miles!! I"M NEVER PAYING FOR THE INTERNET AGAIN!
124
Dead in the water, chum (2.55 / 9) (#10)
by trhurler on Fri Dec 10, 2004 at 08:46:17 PM EST

First of all, in any remotely "urban" or even "suburban" area, that's not NEARLY enough bandwidth for a thirty mile radius. Not even for a ten mile radius. So, the only place this will be of much use is out in the boonies. But guess what? There isn't that much money in the boonies. Not only are there fewer people per square mile(and remember, cheap though end user equipment may be, base stations are going to be expensive, so this translates directly to less profit per dollar invested,) but they are on average poorer, less educated, and less interested. Yes, there are certainly plenty of exceptions - well to do, well educated, sensible people who live out in the boonies because they like it that way. That is, there are plenty of them to demonstrate that no generalization is always correct. There are NOT plenty of them in any one area to make a technology like this worth having.

Yes, you might be able to deploy it on the fringes of suburbia. Problem being, that's not a big enough market to be worth most companies' choosing to employ a whole new technology over it.

Dead, I tell you. Hopeless. Whoever thought 70Mbps was even remotely sufficient for a "metropolitan area network" was having a Bill Gates moment. ("640k is more than anyone will ever need.")

Add to this the fact that all the big established players have a good reason NOT to see this deployed(namely, it starts to remove a need for their bigness and establishedness,) and now you've got to hope that some startup will come along and deploy this despite the obvious cash flow problem it will have. Be serious; even if some "visionary" is dumb enough to try it, it won't last.

In the long run, cable is going to suffer the same fate, for much the same reason. As user density goes up, quality goes down, and there's not much cable companies can do about it that is affordable to them. Their infrastructure is inherently unscalable in a way that is not true of infrastructures built on more "traditional" telecom technologies.

Of course, telecom as an industry is fucked, but in a technical sense they've got things at least half right, and have a shot at success over the long haul.

But really, give up on this wireless MAN crap. It isn't delayed so long because of problems with the technology. It is delayed so long because it is an unprofitable idea that won't actually provide worthwhile service in its intended role.

The only way wireless is going to work as an infrastructure is if it is done mostly the same way mobile phone networks operate today, but with more bandwidth. And in fact, I think we're only a few years away from seeing a resurgence of the "connect your computer to your phone" method of connectivity, except digital and with mobile phones. People already do it(some people, that is,) but it will become more common as airtime gets cheaper and bandwidth goes up.

--
'God dammit, your posts make me hard.' --LilDebbie

wireless is the best option for many users (3.00 / 3) (#11)
by jsnow on Fri Dec 10, 2004 at 10:08:30 PM EST

Smaller, denser networks can be created simply by reducing transmit power. (There's a lot of research currently being done in the use of adaptive transmit power settings in wireless ad-hoc meshes. Lookup "topology control" on citeseer or google.)

Furthermore, there's a heck of a lot of people in this country who's only option for broadband is satelite. It's very easy to compete with ~$70 a month for heavily rationed, high latency network access, with setup fees of hundreds of dollars.

My parents in the boonies just got wireless internet access from a local ISP using waverider equipment. It's a proprietary tdma system a bit more primitive than 802.16. If you buy straight from the company, I believe the radios each cost about $400, and the access points cost a few thousand. I think the total throughput is about 2mbps, and they use unlicensed spectrum. Sure, it's slow if the network is heavily used by a lot of people, but it's vastly superior to dialup. It doesn't take long to recover equipment costs when you can charge ~$50 a month or so, and there are a lot of people who live just beyond the reach of wired broadband.

802.16 will also provide some competition for the local wired internet monopolies.

The real solution is for everyone to have their own line of single mode fiber pair and be able to run an OC-192 or gigabit ethernet (or multiple channels of each) straight to the telco, but this isn't likely to happen any time soon. In the mean time, people will pay for whatever gives the best quality at the cheapest rates.

[ Parent ]

Be serious (1.00 / 2) (#16)
by trhurler on Sat Dec 11, 2004 at 02:59:41 AM EST

First of all, reducing or increasing the power output dynamically and rearranging network configurations as you do so is a problem that's actually MUCH tougher than IP routing, which is still done in ways so primitive that it is a pathetic joke. After 30 years, IP routing protocols and related materials still SUCK ASS and are based on methods a teenager could probably dream up, because there's nothing better! Sure there's research - because it employs researchers and draws grant money. It isn't going to work out though - at least, not anytime soon.

Second, 4-6 miles? In urban areas, a 4-6 mile radius may have several hundred thousand people. Do you think they can all be served with a bandwidth limit that's actually only about 45 T1 lines? (That 60 number is someone's absolute raving fantasy, even WITHOUT any contention issues. Add in contention issues, and probably you won't get more than the equivalent of maybe 30 T1s in effective bandwidth.)

Third, proprietary schemes will take off, yes. Your parents may be well served by them. STANDARD schemes for wide area wireless, on the other hand, are not going to take off for reasons that ought to be obvious. You DO understand the difference in the business model, right?

--
'God dammit, your posts make me hard.' --LilDebbie

[ Parent ]
I am serious (none / 0) (#18)
by jsnow on Sat Dec 11, 2004 at 04:10:42 AM EST

First of all, reducing or increasing the power output dynamically and rearranging network configurations as you do so is a problem that's actually MUCH tougher than IP routing

For mesh networks, that may be true. See this paper for an example topology control algorithm that maintains a biconnected mesh. For 802.16, most people are going to be using a simple star topology: one base station, many clients. Multiple base stations can be connected by landline. In that kind of network, topology control is very easy: set all the clients to the minimum power they need to reach the AP, and set the AP to whatever power creates an optimally sized cell. Not rocket science. Most current research is in mesh networks and their associated routing protocols (AODV, DSR, TORA, DSDV, etc...), and topology control algorithms (LINT, LILT, CEC, etc..) which is a much harder problem. We've known how to make "cellular" one-hop networks for a long time now.

Sure there's research - because it employs researchers and draws grant money. It isn't going to work out though - at least, not anytime soon.

Research rarely works out anytime soon. Someone has to advance the state of the art so that one day we can build networks of thousands of self-organizing wireless nodes that can move around without bringing the network to its knees with routing protocol overhead. This is a real problem that a lot of people with deep pockets would like to see solved, and some groups have already seen some success with real networks.

Second, 4-6 miles? In urban areas, a 4-6 mile radius may have several hundred thousand people. Do you think they can all be served with a bandwidth limit that's actually only about 45 T1 lines?
Not everyone lives in urban areas (most of whom are already served by cheap broadband). Also, ISPs can usually get away with heavily oversubscribing their users without anyone noticing. 70mbps could serve over a thousand 1mbps users. That would be about right for most small towns (many of which don't have broadband). If that's not enough, set up more access points on different channels, and/or shrink the coverage areas.
Third, proprietary schemes will take off, yes. Your parents may be well served by them. STANDARD schemes for wide area wireless, on the other hand, are not going to take off for reasons that ought to be obvious. You DO understand the difference in the business model, right?
Let's see, Intel or whoever makes money selling a reasonably low-cost hardware solution. ISPs buy it because it's faster and more featureful than competing products. They can either use unlicensed spectrum, and fight for bandwidth with the other unlicensed ISPs, or buy spectrum from the FCC and not have to worry about interference. Sounds like a reasonable model to all involved. It's also not much different than a proprietary-hardware solution, except you can buy equipment from multiple vendors and have it interoperate (assuming the equipment works on the same frequency and implements the same subset of the standard). As far as I can tell, Waverider just sells hardware - they don't control the network or demand a portion of the ISP's revenue or anything like that.

[ Parent ]
Good one! (none / 0) (#32)
by trhurler on Sat Dec 11, 2004 at 09:27:06 PM EST

For 802.16, most people are going to be using a simple star topology: one base station, many clients. Multiple base stations can be connected by landline. In that kind of network, topology control is very easy: set all the clients to the minimum power they need to reach the AP, and set the AP to whatever power creates an optimally sized cell.
Step one: steal underwear. . . step three: profit! "Whatever power creates an optimally sized cell" becomes your key routing decision. How do you make it? What if it changes frequently(as it likely will, what with commutes and so on?) How do you coordinate with neighbors to make sure your changes don't create coverage gaps? Yadda yadda. Totally infeasible.
We've known how to make "cellular" one-hop networks for a long time now.
Yes, but only by using multiple frequencies, substantial overlap, and so on(or just putting up with gaps in coverage.) The sizes are fixed, we deliberately undersize them to ensure capacity, and we pay through the nose for it. Thing is, if you're going to do that, you don't NEED 30 mile radius, a dedicated protocol, and whatever else. There are already other standards for doing what you're describing, if it is just a cell network.
Research rarely works out anytime soon. Someone has to advance the state of the art so that one day we can build networks of thousands of self-organizing wireless nodes that can move around without bringing the network to its knees with routing protocol overhead.
Agreed. However, this isn't going to happen until LONG after 802.16 is so obsolete it isn't even funny.
70mbps could serve over a thousand 1mbps users. That would be about right for most small towns (many of which don't have broadband).
HAHAHA. Um... yes, if you want to be the most hated company in existence.

I'm not even going to go into your description of the business model, because it is completely wrong and I have to go get drunk right now. Gotta arrive before the cover starts, you know. Suffice for it to say that the existence of "competitors" who will swoop in and take your customers away after you've done all the expensive "marketing" and making sure the customers have hardware and educating them on how to use it and so on and then undercut your prices is a proven problem in the telecom space, and that in this case, it is a HUGE risk that no sane investor is going to take without the promise of bigger profits than you can possibly honestly suggest he might make.

--
'God dammit, your posts make me hard.' --LilDebbie

[ Parent ]
file under (none / 1) (#53)
by Harvey Anderson on Sun Dec 12, 2004 at 03:47:31 PM EST

'dork overcompensating by acting the hard-nosed business tough guy'

[ Parent ]
Obviously (none / 1) (#54)
by trhurler on Sun Dec 12, 2004 at 04:22:05 PM EST

After all, since YOU are a dork and don't understand business, that must be true of everyone you meet. Right?

--
'God dammit, your posts make me hard.' --LilDebbie

[ Parent ]
Ya know, aside from my armchair psychology... (3.00 / 4) (#56)
by Harvey Anderson on Sun Dec 12, 2004 at 05:59:35 PM EST

You have a greater than average (for K5ers anyway) tendency to cast about factoids and analyses as if you are some sort of knowing authority but you offer nothing to back up your implicit assertion of "I know what I'm talking about; listen to me!"

Examples:

"Not only are there fewer people per square mile ... but they are on average poorer, less educated, and less interested." (Probably technically true but on your mythical index here it is unknown by you how true it is and thus it is unknown by you whether the different is large enough to be relevant to what you were discussing)

"Of course, telecom as an industry is fucked, but in a technical sense they've got things at least half right, and have a shot at success over the long haul." (This is said with authority, as if you have long experience with these issues, or is it a tidbit gleaned from back issues of Money?)

"A good rule of thumb in technical fields is, if you can't make at least $100,000/year per salary you're paying, you shouldn't even bother to start it up, even in the boonies." (What rule of thumb is this?  Is this in Chapter 2 of How To Start a Business and Get Rich?  Is this a whispered secret among multi-billionaires?  Econ 101?  Where is this coming from?)

"As for mom and pop ISPs, they're still quite common in the REAL boonies. Turns out big companies think exactly as I predict they would think: it isn't worth it." (How do you know any of this?  What is your definition of 'real boonies'?  Once you have that, have you broken down what portion of these are served by various sources of connectivity?  Are you making educated guesses?)

"I know, you're a self-described geek, and you don't understand business." (Making totally unfounde personal digs at someone you do not know, apparently because he dares to challenge your omniscient regarding The Cold, Hard Real World)

"HAHAHA. Um... yes, if you want to be the most hated company in existence." (regarding 70mpbs serving 1,000 cable subscribers) (Do you know how much cable companies oversubscribe?)

"After all, since YOU are a dork and don't understand business, that must be true of everyone you meet. Right?" (A clear over-reach.  'If I question the mighty trhuler, I must not be a dork and not understand business.'  Puh-leeze.)

And so it goes.  Please note that I'm not saying that you're automatically wrong, but you come off sounding like a load of hot air.  If you're sure of what you claim, it shouldn't be hard for you to back it up every now and then.

[ Parent ]

Um... (none / 0) (#63)
by trhurler on Mon Dec 13, 2004 at 08:11:19 PM EST

I do have experience in these industries. My figures are estimates, but I can in fact back them up. Finally, I don't give a damn. If you think I'm full of it, quit bothering me and go bother someone else. I'm not here to satify you.

--
'God dammit, your posts make me hard.' --LilDebbie

[ Parent ]
Bothering? (none / 0) (#65)
by Harvey Anderson on Mon Dec 13, 2004 at 08:41:33 PM EST

You've just made more vague assertions.  Whatever.

The question is why do you bother providing your 'expert opinion' if you aren't willing to back it up?

[ Parent ]

Well, (none / 0) (#66)
by trhurler on Mon Dec 13, 2004 at 08:46:42 PM EST

First, because I cannot provide any "proof" that is likely to convince anyone who wants to believe that I'm full of it, for two reasons. One, that they will demand an entirely unreasonable standard of proof(far higher than they normally accept for MUCH more radical claims day to day,) and two, that while there is method to my madness, nothing about something this complex is ever more than a very strong probability.

Second, because I am not here to convert the world. That is not possible. I am here to say what I believe, and if someone gets something out of it, good for him. If some people think I'm not worth their time, I lose nothing by that.

--
'God dammit, your posts make me hard.' --LilDebbie

[ Parent ]
I hear you. (none / 0) (#69)
by Harvey Anderson on Tue Dec 14, 2004 at 08:37:24 AM EST

You could at least say something like, "Hey, I've been in the telecom industry for X years".. I am not asking you to jump through hoops.

You say you are here to say what you believe.  I'm saying that's fine but it's pointless unless you want to say what grounding you have to even have a belief in the first place.

And yeah, there are suckaz who will never believe anything anyway, but screw them.

[ Parent ]

what's your problem with fixed wireless? (none / 0) (#55)
by jsnow on Sun Dec 12, 2004 at 05:43:35 PM EST

"Whatever power creates an optimally sized cell" becomes your key routing decision. How do you make it? What if it changes frequently(as it likely will, what with commutes and so on?) How do you coordinate with neighbors to make sure your changes don't create coverage gaps? Yadda yadda. Totally infeasible.

Just pick a reasonable number and stick with it. I suspect most 802.16 users are going to be fixed wireless with directional antennas permanently installed. If no one's moving around, adaptive power control isn't a problem.

Yes, but only by using multiple frequencies, substantial overlap, and so on(or just putting up with gaps in coverage.) The sizes are fixed, we deliberately undersize them to ensure capacity, and we pay through the nose for it. Thing is, if you're going to do that, you don't NEED 30 mile radius, a dedicated protocol, and whatever else. There are already other standards for doing what you're describing, if it is just a cell network.

That's probably true. I don't know why 802.16 is considered any better than, say, wireless DOCSIS or MMDS, which are also TDMA protocols. Maybe it has higher capacity and/or more efficient spectrum use. The question "is 802.16 useful?" is different than "is 802.16 better than existing standards?" I would say yes to the former, but I'm not familiar enough with 802.16 or its competitors to answer the latter.

When comparing to cell phones, remember that you pay a huge price for mobility. Dirctional antennas can boost the recieved signal, and suppress unwanted noise coming from other directions. Increases is S/N translate into a proportional increase in information capacity.

I'm not certain of this, but I suspect that data connections over CDMA cannot take advantage of spare capacity on underutilized channels. 802.16 can do this. If you're the only one on at 3:00 in the morning, you get all 70 mbps of throughput.

HAHAHA. Um... yes, if you want to be the most hated company in existence.

70mbps / 1000 users = 70kbps, which is still better than dialup (which is what we're competing with), even in the extremely unlikely case that everyone is using the network simultaneously. Keep in mind that most people just surf the web and check email. If 14.28:1 oversubscription is too high, market it as 512k or 256k internet. People will buy it if it's the best service available.

Suffice for it to say that the existence of "competitors" who will swoop in and take your customers away after you've done all the expensive "marketing" and making sure the customers have hardware and educating them on how to use it and so on and then undercut your prices is a proven problem in the telecom space, and that in this case, it is a HUGE risk that no sane investor is going to take without the promise of bigger profits than you can possibly honestly suggest he might make.

In other words, wireless internet access will be a commondity, like bread. Anyone with a leased line and an access point can sell it. That's the way it should be. (The FCC complicates things a bit, but 802.16 can work in unlicensed space.)

[ Parent ]

Hmm (none / 0) (#64)
by trhurler on Mon Dec 13, 2004 at 08:27:16 PM EST

Just pick a reasonable number and stick with it. I suspect most 802.16 users are going to be fixed wireless with directional antennas permanently installed. If no one's moving around, adaptive power control isn't a problem.
If that were practical, it'd be great. Problem is, it isn't practical. It is going to mean you have to MASSIVELY overcommit resources most of the time(peak usage will probably occur for less than one fourth of the day in most areas,) thereby raising your fixed costs and capital expenditures dramatically and eventually raising your operating expenses even further through non-fixed expeditures such as maintenance on all of this. All of a sudden, instead of being a cheap alternative to tethers, you have an expensive alternative to tethers, which will only be affordable by companies that have no incentive to offer it.
When comparing to cell phones, remember that you pay a huge price for mobility. Dirctional antennas can boost the recieved signal, and suppress unwanted noise coming from other directions. Increases is S/N translate into a proportional increase in information capacity.
Only if everyone is required to use a directional antenna. If you even ALLOW other users, their higher transmission power and greater retry rates will screw everything up. If you DON'T allow other users(ie, mobiles, laptops, etc,) this isn't going to be very popular at all.
I'm not certain of this, but I suspect that data connections over CDMA cannot take advantage of spare capacity on underutilized channels. 802.16 can do this. If you're the only one on at 3:00 in the morning, you get all 70 mbps of throughput.
A wondrous misfeature if ever there was one. Aside from the fact that your less clueful users(that is, most of them,) will not understand this and will therefore find reasons to complain about it, networks without explicit bandwidth reservation end up paying a price in wasted bandwidth(turns out resource allocation costs resources.)
70mbps / 1000 users = 70kbps, which is still better than dialup (which is what we're competing with),
What you're competing with in Bumfuck, Idaho maybe. Certainly not in any "metropolitan" area in which you might put a network. But in the boonies, I still don't think this will work even with a 30 mile radius.
even in the extremely unlikely case that everyone is using the network simultaneously. Keep in mind that most people just surf the web and check email. If 14.28:1 oversubscription is too high, market it as 512k or 256k internet. People will buy it if it's the best service available.
Maybe. I still think you're just not reaching a big enough group of people that way to be worth it. A 4-6 mile radius in an urban area, and only a couple hundred customers? You'd be better off setting up cheap 802.11 equipment and supplying a few landlines to connect it all. The result would cost you less and probably provide superior service to a larger number of people.
In other words, wireless internet access will be a commondity, like bread. Anyone with a leased line and an access point can sell it. That's the way it should be. (The FCC complicates things a bit, but 802.16 can work in unlicensed space.)
Everyone says this. The problem, of course, is that nobody wants to invest in a commodity unless the scale is enormous. As just mentioned, the scale on this is closer to "pointless" than "enormous."

--
'God dammit, your posts make me hard.' --LilDebbie

[ Parent ]
802.16 is not an 802.11 replacement (none / 0) (#67)
by jsnow on Tue Dec 14, 2004 at 12:23:42 AM EST

  • If you want 802.11, you know where to find it.
  • If you want a medium distance microwave link or T1 replacement use 802.16.
  • Shared bandwidth connections are more scalable.
  • This is why circuit switched networks are being replaced by packet switched networks.
  • Living outside a metropolitan area does not mean you live 20 miles from your nearest neighbor, or that you don't have electricity and indoor plumbing.
  • 802.11 does not have QoS. In fact, it gives priority to the most recent sender. In effect, it imlements anti-QoS.
  • Large, variable latency is bad for VoIP and games.
  • 802.11 uses random backoffs and carrier sensing for collision avoidance. If the propogation delay is large, this works poorly, resulting in many collisions.
  • 802.16 channel access is scheduled in advance by the access point.
  • 802.16 has QoS capabilities.
  • "Not Qwest or Verizon" is not the same as "nobody". If the service is not tied to the hardware, anyone who can get a leased line can set up a wireless ISP.


[ Parent ]
isn't 30 miles the max possible radius? (none / 1) (#12)
by cryon on Fri Dec 10, 2004 at 10:27:29 PM EST

You say that within the area encompassed by a 30 mile radius, that 70G is not enough radius, implying, if you understand you correctly, that there would be so many users within that area, that the 70G would be eaten right up. Correct? OK, but 30 miles is a MAXIMUM radius. No law against using a 1 mile radius or even a 1000 ft radius, or an even smaller radius, in urban areas, that is. THe problem is the LAST MILE problem. OK, fine, just go smaller. Even if it requires much more infrastructure to have multiple 500 ft radii as opposed to one 30 mile radius, you still avoid having to run wires to everyone's home! Remember, THAT is the core of the problem. Someone help me out here. I am not a network person.
HTGS75OBEY21IRTYG54564ACCEPT64AUTHORITY41V KKJWQKHD23CONSUME78GJHGYTMNQYRTY74SLEEP38H TYTR32CONFORM12GNIYIPWG64VOTER4APATHY42JLQ TYFGB64MONEY3IS4YOUR7GOD62MGTSB21CONFORM34 SDF53MARRY6AND2REPRODUCE534TYWHJZKJ34OBEY6

[ Parent ]
correct (none / 0) (#13)
by jsnow on Fri Dec 10, 2004 at 11:01:41 PM EST

That's 70mbps, not 70gbps. According to this, Intel expects average cell sizes to have a radius of 4-6 miles.

[ Parent ]
completely off topic (none / 0) (#14)
by KnightStalker on Fri Dec 10, 2004 at 11:36:56 PM EST

Where was that photo on your cs.pdx.edu page taken? Can't be anywhere around here (Portland), as you appear to be sitting on actual rocks :-)

[ Parent ]
yosemite falls (none / 0) (#17)
by jsnow on Sat Dec 11, 2004 at 03:24:50 AM EST

That's a picture from a road trip I went on around the country with three of my friends. We made a web page for the trip. If you're really bored, you can read about how we were almost eaten by bears.

[ Parent ]
Extremely short-sighted. (3.00 / 6) (#19)
by mcc on Sat Dec 11, 2004 at 10:52:07 AM EST

First of all, in any remotely "urban" or even "suburban" area, that's not NEARLY enough bandwidth for a thirty mile radius. Not even for a ten mile radius

Were everyone to use this for home service, very possibly not. However home service is not the only possible application of this technology. Also possible is that people would want to use the internet when not at home-- people with laptops and PDAs and such. In order to get this kind of service people would be likely to put up with drastically lessened bandwidth from what they might get at home, and the very dense urban areas where the bandwidth problems you raise become a problem are the exact places where a sustainable number of people might be willing to pay for such a service just to use their portable devices.

This is entirely apart from the case of areas-- Universities, San Francisco-- where there is some party that is willing to just set up wireless internet and pay for everyone's connections. Since people are already widely using 802.11 for this exact purpose despite the protocol's extreme unsuitibility for wide area coverage, this indicates there is a decent demand for 802.16 right out of the gate.

(and remember, cheap though end user equipment may be, base stations are going to be expensive, so this translates directly to less profit per dollar invested,)

This does not seem like an entirely reasonable assumption to make. The base stations will probably be very expensive, yes, but the cost will almost certainly not be in any way comparable to that of running fiber, and the cost will be associated with an area rather than with every individual house. "Less profit per area" doesn't matter. What matters is whether you can make some profit per area consistently enough to make selling the service worth someone's while. The difference in cost between slapping down a big transmitter on the roof of a building downtown and running fiber to all the local buildings is very probably enough to make the difference between profitability and not.

Add to this the fact that all the big established players have a good reason NOT to see this deployed(namely, it starts to remove a need for their bigness and establishedness,) and now you've got to hope that some startup will come along and deploy this despite the obvious cash flow problem it will have.

The exact appeal here is that starting up with this technology does not require being an established or even particularly large player. The lack of cabling removes a huge barrier to entry and opens the possibility that businesses that are willing or capable to exist on small or risky margins will be able to spring up. This means that commercial uptake of this technology, if it's adopted at all, will be governed by the limits of the extent to which the market can support it, not the whims of large lethargic telco companies. Capitalism can do some neat things in those rare cases where it is actually given the opportunity to function.

The thing is all the problems you cite are only problems in comparison to an existing broadband alternative-- say, dsl or cable. However there still remain places and situations in the world where dsl or cable are not viable options, and services based on this new protocol look like a very attractive possibility in those cases, especially if 802.16 support begins to show up commonly in 802.11 cards. There is also the definite possibility this technology could prove itself sustainable at prices well below that of traditional broadband-- thus opening up an entire new market full of people that right now cable/dsl isn't touching-- but it is hard to predict that ahead of time.

[ Parent ]

No (none / 0) (#31)
by trhurler on Sat Dec 11, 2004 at 08:59:18 PM EST

First of all, 802.11 is NOT unsuitable for peoples' current uses for it. It is far more suitable than this crap we're discussing right now.

Second, no, really, the amount of profit you are going to make DOES matter, because your capital could be invested in something with a higher return on investment. (I know, you're a self-described geek, and you don't understand business. Well, I do, and I'm here to tell you, nobody with the expertise to do it right is going to go into a low-margin line of work unless the scale is enormous.)

Third, this is the kind of technology that either has to be deployed in a very widespread manner, or else won't take off much at all. People aren't going to buy compatible equipment unless there's a reasonable installed base of places you can use it, and there won't be a reason to deploy it unless there are going to be users. A company with an existing userbase can expand that userbase with a technology LIKE this, but it is in their interest to make that technology deliberately nonstandard because otherwise a competitor can come along after the first guy has done the hard work(awareness-raising, getting equipment into peoples' hands, etc,) and offer a lower rate!

Ain't gonna happen.

--
'God dammit, your posts make me hard.' --LilDebbie

[ Parent ]
Intruiging. (3.00 / 4) (#36)
by mcc on Sun Dec 12, 2004 at 12:53:03 AM EST

I didn't know you could be branded a "self described" anything by other people.

[ Parent ]
You can't have it both ways (none / 1) (#25)
by pyro9 on Sat Dec 11, 2004 at 05:49:27 PM EST

So you're saying that even if you can find 140-280 customers in the boonies willing to pay $50/month it's not worth it?


The future isn't what it used to be
[ Parent ]
Well, (none / 0) (#30)
by trhurler on Sat Dec 11, 2004 at 08:53:49 PM EST

Let me put it this way. Three hundred customers at $50/month is less than $200,000 a year. That might support a VERY tightly run business with two employees. With three, they'd have to be grossly underpaid or you wouldn't have any money left for operating expenses, let alone profits, after you did the payroll, paid taxes, paid for various permits, paid for benefits for employees, paid for insurance, and so on. But, even with three, you need one guy to go around doing installs, so now you have two, then you need one guy to answer the phone, schedule the first guy, handle complaints, do the books, interface with other(contract) professionals such as lawyers, insurance guys, and so on(your "business guy,") and at least one techie to run the base station. Somehow, you have to advertise this service to these people, and you have to convince quite a few of them to sign up(considering the population density, probably not achievable even if you had a marketing department, which you obviously do not.)

So no, it isn't worth it. A business guy, a techie, and an installer can easily find work that would pay better that that gig would(can you imagine a business guy with that kind of variety of skills who would work for way under a third of $180,000?! Yeah, right...) A good rule of thumb in technical fields is, if you can't make at least $100,000/year per salary you're paying, you shouldn't even bother to start it up, even in the boonies. In areas with higher cost of living, make that $200,000.

--
'God dammit, your posts make me hard.' --LilDebbie

[ Parent ]
Anything seems bad if you try hard enough (none / 1) (#47)
by pyro9 on Sun Dec 12, 2004 at 12:14:24 PM EST

300 people in 2800 square miles isn't really all that hard!, there's no law that says you can't operate two or more of these.

Self install is an option since there's not much to it. Send the unit by USPS and have the customer call when they get it.

All businesses are faced with bgetting the word out. Plenty manage to do that.

Running a single station as a sole source of income for 3 people is a bad idea. Running several or doing that as part of another business is more workable.

Consider, how many of your arguments couldn't as easily be applied to providing dialup service in a small town. Yet, mom and pop ISPs spring up all over at one time, they didn't go away until large ISPs with economy of scale moved in.


The future isn't what it used to be
[ Parent ]
Er... (none / 0) (#52)
by trhurler on Sun Dec 12, 2004 at 03:45:33 PM EST

First of all, for most of your proposed customers, no, you cannot just ship them the hardware and have them install it. If you want to know why, work a helpdesk for a few months.

Second, reaching them is a bigger deal than you think, because it will cost a lot of money that you aren't going to make back very quickly. An ad campaign that would work might easily cost several years' operating revenues.

Third, given that the primary initial operating cost in any given area is going to be marketing and given the scale of that expense versus your expected income, how exactly are you going to operate MORE than one of these?

Running this as part of another business might be feasible, but where's the incentive? You can dedicate those resources to this, in which case you're faced with MAYBE being profitable five years later if your service isn't obsolete and/or replaced by a lower priced competitor who didn't have to pay all the up front costs just to create the market, OR you can dedicate those resources to something that has a more favorable outlook. Guess what?

As for mom and pop ISPs, they're still quite common in the REAL boonies. Turns out big companies think exactly as I predict they would think: it isn't worth it.

--
'God dammit, your posts make me hard.' --LilDebbie

[ Parent ]
Word. (none / 0) (#74)
by valeko on Tue Dec 14, 2004 at 10:52:26 PM EST

First of all, for most of your proposed customers, no, you cannot just ship them the hardware and have them install it. If you want to know why, work a helpdesk for a few months.

Amen!

"Hey, what's sanity got going for it anyways?" -- infinitera, on matters of the heart
[ Parent ]

Yeah. (none / 1) (#33)
by valeko on Sat Dec 11, 2004 at 10:05:51 PM EST

I think you're right on with that one.

Speaking of the boonies, I work for an ISP in Georgia whose (steadily declining) revenue base consists of mostly dial-up customers out in the boonies. We're experiencing this problem of low ROI firsthand. A sizable group of them live in areas where they are unqualified for DSL, do not have cable TV service, and so on, and so they constantly clamour to have this remedied. But neither the cable nor the telephone company are bothering with them; there's just not much money to be made.

The company is in the process of rolling out this "accelerated dial-up" service that stuffs brutally compressed data down the same 56K pipe (for a theoretical maximum of "five times faster than regular dial-up"), but that's not going to work out. Some of their land lines are so poor they do not get anywhere near 56K as is, and in the grand scheme of things, it's just not going to work out well, I can already see.

"Hey, what's sanity got going for it anyways?" -- infinitera, on matters of the heart
[ Parent ]

Excuse me sir but are you a troll? (none / 0) (#41)
by Haunting Koan on Sun Dec 12, 2004 at 10:43:02 AM EST

Those ascii dudes were wondering.

[ Parent ]
Where's the killer app? (2.00 / 3) (#15)
by Dr Gonzo on Sat Dec 11, 2004 at 12:29:48 AM EST

Urban areas are already getting fiber run to customers by major bandwidth providers, a medium that provides orders of magnitude more bandwidth than a wireless link while not being susceptible to atmospheric interference and other problems inherent in wireless transmission.

On the other hand, rural areas don't really have the sort of money to justify running these sort of networks out there.

So where does this all fit in?

"I felt the warmth spread across my lap as her bladder let loose." - MichaelCrawford

where it fits in (3.00 / 2) (#23)
by jsnow on Sat Dec 11, 2004 at 03:48:47 PM EST

Urban areas are already getting fiber run to customers

Fiber deployment is a long way off for most of us (even in urban areas), and will probably be tightly controlled by the local ILEC when it is deployed. Sure, it's a lot faster (we can push several tbps over a single strand of singlemode fiber with current (though expensive) technology, with repeaters about every 100km), but we aren't going to see gigabit ethernet to the home anytime soon, even if it is economically feasable to install the hardware.

On the other hand, rural areas don't really have the sort of money to justify running these sort of networks out there.

By "these sort of networks", do you mean fiber or wireless? Rural folk are probably about equally willing to pay for broadband as city users, it's just not available. $50 a month might justify a fiber run, but it won't pay off for quite some time. (If you can charge $50 a month for a fiber line, and have $10 a month in upstream bandwidth costs, and assuming it's obsolete in 30 years (that's pessimistic), that's $14,400 per customer - not a bad return on investment.) At any rate, the ILECs don't seem in any hurry to deploy fiber to the un-networked masses, and wireless networks can be deployed very easily by just about anyone.

People will use whatever technology is cheapest, best, and most available, and setting up an access point to serve a small or medium sized town is very cheap compared to burying wires or fiber. Unless the hardware or spectrum licensing is prohibitively expensive, I expect 802.16 to be popular in areas that don't already have DSL or cable broadband. It might even provide some much-needed competition for the local broadband monopoly or duopolys.

[ Parent ]

Oops, accedental +1 (none / 0) (#29)
by Nyarlathotep on Sat Dec 11, 2004 at 07:39:18 PM EST

So you never said how it scales.  There are very good theeoretical wireless protocolls which automatically adjust to population density, and do message passing between local base stations, i.e. you buy a base station, just like with 802.11b, but you don't plug it into DSL, you just let it talk to your neighbors base station.  The point of such a system is that it causes the "last mile" transport to scale with adoption or population growth.. although your land line links from this system need to be manually added too.  The side point of such a system is that end users are paying for the last mile hardware, and hence have the freedom to buy cheaper bandwidth.

I'm guessing 802.16 is not such a protocoll, and hence not likely to be all that revolutionary.
Campus Crusade for Cthulhu -- it found me!

That's a mesh netowrking proto.. (none / 0) (#62)
by hyevoltage on Mon Dec 13, 2004 at 05:49:32 PM EST

What you are describing is a mesh network topology and no it is not supported by 802.16x. The IEEE has some preliminary study groups (802.11s i believe is one) but they are to be incorporated within WiFI/802.11 tech, not wimax. However, just because WiMax doesn't initially support mesh doesn't mean it's not a revolutionary tech. Maybe not in the innovative sense, but in the sustainable progression of wireless tech its a big leap. It has the potential to give a small ISP huge penetration and offer multiple services all over IP.This will create carrier competition which will drive the prices lower as effectively as your "buy only hardware at the last mile" scenario. What will be the killer serv (not killer app. hehe) is when 802.16e (or 802.20) gets mature and introduced. Same carrier for mobile phone, tv, internet, telephony, etc. Though Flarion is an available alternative ready right now as we speak. Only time will tell.

[ Parent ]
+1; Good idea, expect poor implementation (none / 1) (#34)
by darklordseth on Sat Dec 11, 2004 at 10:18:09 PM EST

Okay, this is how I picture things will go:
  1. 802.16 is quite complete and nicely getting along.
  2. Company gets massive funding from a VC.
  3. VC pushes company into releasing 802.16 tech prematurely.
  4. VC ascertains control over patents regarding the companies 802.16 products.
  5. Due to lack of competition and outrageous pricing ( as set by the VC ), the company will fold.
  6. VC withdraws, having made a few million on the company in question.
  7. 802.16 dies a slow horrible death; Any company trying something similiar gets patent-smacked by the VC, existing hardware poorly constructed with buggy software and no support or repair services available anymore.
  8. 802.16 fades into obscurity.
  9. Anything but profit.
This is a worst case scenario, mind you. More realistically, the VC would turn the company into a sockpuppet and charge outrageous prices for poor hardware and software and close to no support. The standard will become undead; alive yet no one uses it.

IEEE (none / 0) (#60)
by Aero Leviathan on Mon Dec 13, 2004 at 03:03:30 PM EST

I don't think you can patent an IEEE standard, can you?

~ Aero
[ Parent ]
Perhaps not. (none / 0) (#73)
by valeko on Tue Dec 14, 2004 at 10:50:46 PM EST

But it may be possible to patent various approaches to its implementation, specific devices and derivative technologies, etc. This is even worse if these technologies -- either through some perception of intrinsic merit or some coercive aspects -- become de facto dominant.

"Hey, what's sanity got going for it anyways?" -- infinitera, on matters of the heart
[ Parent ]

70Mb, huh? (none / 0) (#61)
by cdguru on Mon Dec 13, 2004 at 05:18:56 PM EST

Anyone looked into a rural/suburban 70Mbit connection?

T1 isn't much of a problem, but be prepared to pay $600+ a month.

You would need something more than an OC3 (48Mbit I think) and one huge problem out in rural/suburban areas is you can't get a OC3 there. Not at anything like a reasonable cost, anyway.

So, yes, you might be able to put an antenna on the tallest object in the area, or get a permit to erect a tower for it, but the cost per Mbit will greatly exceed what anyone is willing to pay for it. Way down below there is some arguing about how a business like this could work - except they forgot the connection to the Internet price. About all you're going to get in some rural, mom-and-pop ISP type of place is a T1, and that would be a joke on something like this.

Why is this on the shelf? Because there isn't much of a need, and where there is need it doesn't work.

Multiple hops (none / 0) (#70)
by mrogers on Tue Dec 14, 2004 at 09:52:55 AM EST

The base station doesn't need an OC3, it could forward traffic wirelessly to another base station. If you can really manage 30 miles per hop, you wouldn't need many hops to reach a built-up area.

[ Parent ]
Wireless MAN (3.00 / 2) (#68)
by jsnow on Tue Dec 14, 2004 at 12:31:18 AM EST

(I didn't make this up, it's originally from here.)

Wireless Man, Wireless Man.

Doing the things that wireless can. What's he like? It's not important. Wireless Man. Is he a standard or is he a spec? When he's in the last mile does he connect? Or does the mile connect him instead? Nobody knows. Wireless Man.

Fiber man, fiber man. Fiber man hates wireless man. They have a fight, wireless wins. Fiber man.

DSL man, DSL man. Size of an ILEC cable plant. Unusually cruel to CLEC man. DSL man.

He's got reserves you can't withstand. Anticompetative with shonky plans. And when they meet, it's a no win land. DSL man.

Pick one (3.00 / 2) (#71)
by isdnip on Tue Dec 14, 2004 at 03:00:13 PM EST

It's important to note that the press -- and I think the topic story here has this problem too -- often misses the reality of WiMax. They pass along hype and then the user is disappointed when reality sets in. WiMax offers:

70 Mbps maximum speed

50 kilometer range

Unlicensed or licensed operation

But what they forget to tell you is: PICK ONE.

Range and speed are a trade-off. The receiver needs a certain amount of power from the transmitter to get a given speed (Shannon's Law). So if you want 70 Mbps, your range is reduced, for a given path and power, vs. say 5 Mbps.

And most WiMax is for licensed use. Yes, it can be used on unlicensed frequencies, but power limits are lower and the noise is sometimes worse. A lot of WiMax gear now is aimed at the 3.5 GHz licensed band, which doesn't exist in the US. Some is aimed at the 2.5 GHz band, which is licensed in the USA but usually, at this point, to Nextel or Sprint. (Also a lot of educational licensees, including churches, who can lease some of their bandwidth to commercial users.) Licensed gives you more power. Unlicensed WiMax suffers the same power limits as WiFi, though it's somewhat smarter technology.

So don't get all too excited.

802.16: Medium distance wireless networking that could change the world? | 74 comments (46 topical, 28 editorial, 0 hidden)
Display: Sort:

kuro5hin.org

[XML]
All trademarks and copyrights on this page are owned by their respective companies. The Rest 2000 - Present Kuro5hin.org Inc.
See our legalese page for copyright policies. Please also read our Privacy Policy.
Kuro5hin.org is powered by Free Software, including Apache, Perl, and Linux, The Scoop Engine that runs this site is freely available, under the terms of the GPL.
Need some help? Email help@kuro5hin.org.
My heart's the long stairs.

Powered by Scoop create account | help/FAQ | mission | links | search | IRC | YOU choose the stories!