My previous posts on this blog have advocated fiberoptic internet technology. Some have mentioned the problems with wireless. The cellular phone companies, such as Verizon, AT&T and Sprint are transitioning over the next few years to a "4th generation" which some think might be a cure-all for all internet problems. I advocate the use of wireless in many bands and forms where it is appropriate, but I think for heavy-duty daily internet, fiber is far preferable.
Perhaps someone might like to offer other opinions?
Dale Carstensen
Subscribe to:
Post Comments (Atom)
7 comments:
I have no idea whether 4G is the correct solution for Los Alamos, but then neither do the Council or anyone else associated with the Community Broadband Network.
What I do know is that it has been used successfully in other small and rural communities.
I also know that, increasingly, Internet connectivity is being consumed by mobile devices, phones, tablet, notebooks - anything portable.
I would also expect that it would be lower cost than running fiber to every location.
So, it's a proven technology that's probably lower cost, covers more of the devices that the resident have at the same speed with an easier upgrade path.
I don't know if reliability is a problem or whether it can really deliver the performance it promises, or whether it would be lower cost for 100% coverage, but we never will unless we spend some time looking into it. We shouldn't just ignore it.
I found the anonymous comment above in the spam quarantine, and it had been there two and a half hours. Fortunately, no other comments that really weren't spam were there, but if a real discussion ensues, I may need to be more diligent about checking.
I haven't heard much about 4G deployment in small communities or rural areas. I have heard about it in Reno, Baltimore, Portland, New York City, and by now I think it may be in Albuquerque as one of 30 to 50 cities. Some of it is Clearwire using 802.16, the standard also known as Wimax, and I recall reading that many customers are not very satisfied with that. Some are Verizon's LTE (Long Term Evolution) and I recall some testimonials that transfers of around 15 to 20 Mbits are possible. Wimax promised 50 Mbits, but I don't think any actual transfers have matched that.
Wireless is appropriate for a true rural area or even a small (a few to several hundred residences and business computers in a square mile or three) community or true rural area (under 100 residences etc. per square mile), whether it is Wifi, Wimax or some cellular technology. Maybe Pajarito Acres would fit that bill, but the terrain there may not be all that friendly, probably a tall tower would be needed. Widespread hot spots can be useful for mobile use, either by pedestrians or in or on vehicles. But that's not where the money is, so the early deployments are in densely populated affluent areas.
Then there's the business model for cellular phones. You don't own your phone, you rent it, and you're bound by a contract to pay several times the true data subscription cost to pay off the expensive cost of manufacturing the phone. Charges for text messages can be hundreds of times the actual cost to process them. Your phone is obsolete before the contract expires. Your traffic is limited, although sometimes the carrier lies and claims it isn't. On and on.
Politicians discuss spectrum (just watch the weekly "The Communicators" on C-SPAN2) and how billions are spent on spectrum auctions and monopolies prevent deployment more than they do deployment. Sometimes it seems the cell carriers have twisted the discussion into thinking only they need be considered as a medium for internet. No fixed wireless, no point-multipoint systems other than cellular, no copper or fiber media. Fiber and copper merely feed the cell towers. And the market for radios for cellular is quite closed, almost secret.
On the other hand, the Los Alamos County discussion of fiber seems to be locked into ONT equipment, too, which isn't exactly a technology that is an inexpensive, readily available, commodity, either.
I was under the impression that the fiber backbone was to be 1Gb/s which means that individuals will be getting much less. That's the same model as 4G. Unless I seriously misunderstand, no single person is going to get 1Gb/s. I'd be extremely happy with 50Mb/s. We're talking 2013 for deployment. Even is the 1Gb/s wireless is not available, it's an upgradeable solution - fiber is not. And, given how reliable our wired electrical service is, I have my doubts about how well the County will maintain fiber.
You also seem to be considering a model where the provider would be one of the traditional cell companies. There's no reason why the management of a wireless network shouldn't be the same as fiber. Multiple ISPs, phones, etc. on the network. We're not begging one of the providers to build a network, we're building it, so we don't have to worry about "where the money is". (And I'm not talking about wifi hotspots anyway - those are a fiber thing - I'm talking about wireless everywhere).
You've also completely ignored mobile device. 500,000 unique Android devices are being registered every day. With 4G, these could (potentially) have data all around town - not just at a wifi hotspot. 53% of phones sold are smartphones - 60MM plus smartphone units per quarter. With fiber, these devices would only have data at your wifi hotspots.
I don't buy the 4G/LTE, etc. is unsuitable argument based on population density, if that were the case it would be absolutely useless in major cities.
The spectrum issue is the only one that might be valid, but we won't know unless we look.
I'd expect similar data rates, higher reliability (minimal cable maintenance) and wider coverage (mobile devices) at a lower cost because we're not running fiber.
But all this is just speculation, we just don't know.
I'm not advocating 4G/broadband wireless, I'm advocating that we should look at what it might be able to provide. Currently that hasn't happened.
The fiber itself is not limited to the speed available at a reasonable price at the time the fiber is installed. In 2000, and a few years before (I'm not sure what year Bill Cabral started running fiber), we used multimode fiber, because the equipment for singlemode was several times more expensive and multimode would run the 10 Mbits that was then affordable a distance of something like 10 miles. By 2004, 100 Mbits equipment was affordable, becoming cheaper than 10 Mbits equipment, and multimode runs the cheapest 100 Mbits equipment for about 2 miles. Now gigabit equipment is affordable but multimode will only run the cheapest gigabit for about 500 feet. The REDI (Regional Economic Development Initiative) and proposed CBN (Community Broadband Network) projects are using singlemode fiber, which is now cheaper for the fiber itself than multimode. It is more difficult to manufacture singlemode, but the market size for singlemode is now larger than for multimode, so by supply and demand, singlemode is cheaper to buy. The technology may flip again, though, since multimode is a bigger pipe, it may eventually be able to carry higher bandwidth a longer distance if the right technological tricks are found.
Replacing or adding to underground conduit is relatively inexpensive.
I expect the long hauls (over, say, five miles) will be at least 10 Gbits. There are already standards for 40 Gbits and 100 Gbits is becoming state of the art. WDM (Wave Division Multiplexing) can be used to get multiple 10 Gbits "lambdas" on a single fiber for up to around 70 miles. CWDM (Coarse WDM) equipment is affordably priced compared to running extra fibers, DWDM (Dense WDM) not quite so much so. But the technology is really just a precision prism at each end, just like the high school physics demo that shows the rainbow spectrum, so potentially with a large market, DWDM equipment could be really cheap, maybe under 100 dollars. I think running 80 Gbits 40 miles costs a few hundred thousand dollars for the boxes at the ends currently, but every few months prices drop.
Wireless will never keep up with fiber on bandwidth, and a few singlemode fibers are much more futureproof compared to wireless. And as hard as it may be to imagine what anybody would want with more than, say, 50 Mbits for their internet connection, IBM once thought the world market for computers would be maybe 5 or 10 forever. The CBN presentation for 11/01/2011 mentions exabyte. That's decades in the future, I suppose, but may be attained eventually.
"Unlimited" is an interesting concept at any scale. Although we'll never really reach it, we will be amazed at how small the "unlimited" of the past was.
Those big cities that have 4G also have fiber with fast connections, and without the fiber, the 4G would be just about useless. There would be too much interference on the limited spectrum available for wireless to be tolerable at high speeds. One thing that does help cellular, though, is that towers closer together running lower power can keep interference localized, but then you need fiber between the towers. This is why Wifi can work well around a house in Pajarito Acres or La Senda, but not so well in apartments or townhomes, and not well in homes with metal in the walls, big steel furnace ducts, etc. The aestheticists don't like any of those towers. Santa Fe seems to have collected enough people convinced they're allergic to radio waves to cause a big problem for wireless there.
If fibre is easy to upgrade, wouldn't it be easy to upgrade whatever fibre is running to the tower? Obviously, that would be either cheaper or a similar cost to upgrading the fibre to the distribution point.
While I agree that 256k is not enough for anybody, in the next decade, will 1Gb/s be any good? I agree that even Verizon's spec'ed 300Mb/s (not likely to be reached) would only last a few years, but the advantage of a few cell towers is they are easier to upgrade than re-running cable to every house.
The issue you raise about the availability of spectrum is an interesting point. I have no idea if there is any free cellular spectrum (possibly set aside by the government for this purpose) or what it would cost to rent/buy it from a company not interested in expanding to Los Alamos.
Thank you for the blog.
I got the link from your email following up on the Tuesday meeting at the LA library. Perhaps we can use the blog as a focus for organizing support for fiber in LA.
Andy Fraser
Thank you, Jim, anothernonymous and Andy for posting, too.
I must have given the wrong impression by stating that replacing or upgrading conduit fiber is relatively inexpensive. Probably everyone has read between the lines and realized that I mean replacing or upgrading the fiber inside the conduit, not replacing the conduit itself, so that's not where the wrong impression occurred. What I meant to convey is that adding fiber would only be necessary when we need to exceed 10 Gbits/sec, which is already over 10 times as fast as the fastest dreams for wireless. Or with spare fibers at the outset or cheaper WDM, maybe several times higher than 10 Gbits.
Really high speed wireless requires a really high frequency carrier, tens of Gigahertz. And that implies very short propagation distances, and big chunks of spectrum, and expensive equipment (yes, it will get cheaper, but so will fiber and fiber equipment). So while I agree that wireless is useful for mobility, I just don't see it ever being useful as my main internet connection, compared to what fiber can do.
I actually do have a wireless connection, though, and have had since about 1998. I've used two technologies. First was a 900 MHz band Wilan, which runs about 1.2 Mbits, and can still work as a backup. Second is 5.7 GHz 802.11A and I've had two different models of radios for that. These are supposed to be 50 Mbits or so, but in practice the first model was about 5 Mbits and the second is about 15 Mbits. Nobody else in the path between antennas seems to have been using the same band, and if they tried probably my signal would drive them to a different one since I'm there first.
This is for a link under half a mile. For longer distances, the inverse square law applies. Twice as far away is a quarter the capability.
The connection to Albuquerque has never been faster than about 9 Mbits for me, and is currently about 3 Mbits, so, there you go, the wonderful world of wireless, and why I'd rather have fiber. Fiber connections a few blocks away have been able to get around 30 Mbits to Albuquerque via an aggregation of DSL and Comcast connections.
I don't have a smartphone, but my voice phone (2G, maybe??) certainly has problems getting signal in lots of places, and drops calls sometimes, and battery life is a concern. I recall using a smartphone to try to look at google maps to find the Johnson Center on the UNM campus while a block from the Johnson Center on the UNM campus, and it was so slow it would have been faster to stop at a store and buy a map. Asking a campus police officer for directions worked much better. Those problems will be 4G problems, too, I guarantee it.
I don't know about running a cell system ouside the usual carriers, either. The big corporations do have the spectrum licensing tied up, I think. In rural areas, there is a company called Viaero that seems to be where Verizon, AT&T, Sprint, T-Mobile, etc. don't feel it worth being. In cities, there are Cricket and the like that are cheaper than the big boys. But there is some momentum behind the ideas of microcells, nanocells and femtocells, and you can get relay equipment to provide a local booster so a phone can work where a hill, tree or building otherwise blocks you, by adding an antenna on a roof or mast.
Post a Comment