Slashdot Mirror
Will Cellular Swamp WiFi?
hhutkin writes "Sure, Wi-Fi is great for my home network. But what else can it do? After reading this article, I'm convinced that cellular is becoming more ubiquitous with wireless networking than wi-fi will ever be. Just look at all the devices that are coming on the market using cellular technology. I can send email and pics, browse the web, plus listen to MP3s all on one cellular device. It makes the notion of a hotspot almost meaningless." But 802.11x is high-bandwidth, and often unmetered ...
Cellular will win out, because it can be metered. Companies want to be able to charge per byte (or Kbyte). WiFi doesn't let you do that, and overall WiFi may be regulated to mostly home/SOHO use. While cellular becomes the mobile alternative!
WiFi is high bandwidth because they don't have to deal with MANY users - and users who are MOVING AROUND.
3G solves that, with acceptable bandwidth for portable devices (and 4G is planned)
WiFi will _never_ be a threat to 3G - and people who think it is really need to learn about the technical differences.
it's in my head
I am no cellular or Wi-Fi expert, but it seems like security is something to consider when mulling over this question. My Wi-Fi network is secured by me, but the cellular network, being a public/private venture, seems to lack the ability for personal protection. Perhaps this is a good thing for my parents, as the cellular network would probably have some sort of built-in controls, but I like to control my own network. Any thoughts or secuirty tidbits that anyone can share would be appreciated.
And be prepared for a whopping 4 or even 5 digit phone bill. I know of somebody I worked with who thought it was cool to surf the web from his laptop, until he got his next cellular bill.
My rights don't need management.
until the price of cellular more closely resebles the cost of Wi-Fi this kind of apples to oranges comparison is irrelevant.
I love my cell phone, but I need more than 10MB of data for my $20.
TW
Cellular access right now is like (early) dialup:
-slow transfers
-disconnects often
-high latency
-expensive
I have a brand new phone (it says 3G and GPS on the back) but if i use any of these features i get charged up the ass!
Ok...how many people actually DO have the 400-in-1 cell phones that include the garage door opener, TV remote, phone, PDA, camera, MP3 player, GPS, 802.11g router, video game system? Yeah these are cool, but I'm not going to go running with my cell phone so I can use the MP3 player. I'm not going to base my entire network in my house on my phone. I (and a multitude of others) want a device for each purpose. My network will have my WiFi router. My phone will place calls. I'm not going to be screwed when one thing breaks, or when Verizon goes down, or when Verizon decides to jack up the prices, or when...
"He uses statistics as a drunken man uses lampposts...for support rather than illumination." - Andrew Lang
The major US cell phone companies are actually looking at WiFi as a way to increase their service.
They are looking at techniques to 'hand off' a cell phone call from the 2.5G or 3G networks to WiFi networks to maximize call volume.
Imagine that you are walking down the street, chatting on your cell phone and enter a Starbucks. The phone could switch to use the tmobile WiFi instead of using the broader cell phone network.
I saw a presentation by the cafe.com guys where they think the WiFi hotspot companies like them and boingo will get bought up by the Telcos to get access to WiFi for expanded phone service.
nuclear iraq bioweapon encryption cocaine korea terrorist
You have to remember the geograpy of much of the world. In the US, WiFi will never take off outside of cities because of the incredible expense of giving total coverage. Here in the UK, or over in japan however, our population is close to ten times denser, making WiFi a much more attractive prospect. Japanese cell networks have far more transmitters than US networks because of the density of cellular activity, with transmitters often less than 100ft apart, so it's no great leap of the imagination to see those transmitters being supplimented or even replaced by WiFi.
This article reminds me of opposite day. Except for the last sentence.
You see; telecos (in Europe at least) are very, very nervous about 802.11. They paid, literally, billions of dollars for UMTS licenses. Some other poster mentioned a $50/month unlimited plan on GPRS, well, GPRS is really, really slow. UMTS is supposed to be 2 megabit/second, albeit shared with all users in a cell. UMTS cells will however be smaller than conventional GSM cells.
GPRS can be rolled out pretty much instantly using the existing GSM infrastructure. UMTS will need entirely new basestations to put into place. Why build a network with near-100% coverage, if handsets will drop to GSM if necessary? Well, the VERY expensive UMTS licenses require it.
Then comes along wifi.. Speeds of 3 or 10 Mbps. Today. No big network. Free spectrum. Yes, only at hot spots, but where else do you use your laptop? In the mall? In your car? Hardly bloody likely. Hotels, airports, maybe Starbucks. And they ALL have hot spots already! Using the net on a little phone or pda/subnotebook gizmo? You don't need megabits of speed, existing data or GPRS (2.5G) will do.
So 3G is a really big liability. The license is use it or lose it; don't build a network (again, imagine a billion crisp green bills vanishing in thin air) and your other invested billions will never yield any return on investment (a write-off).
mmO2 already took a write-off selling the Dutch O2 branded operator (which will now rebrand to it's *old* pre-O2 brand, telfort). The company was bought for 75 million Euros, and mmO2 wrote off the rest of the company's worth (the billions it put into the UMTS license, which never materialized any revenue so far). The new Telfort says it won't do anything with UMTS.
Show me a device that ordinary people will buy that accesses a service that ordinary people want to use that uses megabit per second speeds AND that can be used anywhere (so, pocketsized, not a big ass laptop). Show me that, and I might be able to get you a very nice consulting gig. There is no killer-app for 3G.
Go wi-fi!
(Though always remember, wi-fi is a commodity, you won't make insane profits (maybe none) and competitors or kiddies can simply jam your signal by using a big ass microwave or other disruptive ISM equipment in the same band..)
SCO employee? Check out the bounty
Yes, WiFi is much faster, than real-world speed of either GPRS or 1xRTT. You're lucky to get 22.8/33.6kbps modem speeds on GPRS. But GPRS brings the service to me, and with Wi-Fi, I have to go to where the service is. We're really comparing wireless LAN (wi-fi) to wireless LAN (802.11a/b/g) technology.
As wireless WAN technology becomes better, it will inevitably win out, so long as the provider has decent coverage, a reasonable pricing plan *and* a strong revenue model. My blissful few years with a Ricochet modem made that clear (R.I.P. Metricom)-- the price was right, but coverage was too limited to attract a mass audience, and the buildout costs were too high for Metricom to turn a profit before running out of cash. But the cellular providers have a big advantage over Metricom, since they can subsidize the buildout of data services on top of their existing voice customers, and they already have the tower permits which they'll need to deploy 2.5G/3G data-- Metricom had to build a data network essentally from scratch, and the process of obtaining the local permits (different in each locality) was a killer right there.
And the big news, T-Mobile has just started an unlimited data GPRS plan, at $20/mo for voice customers with at least a $30/mo voice plan, and $30/mo as a stand-alone offering-- no service commitment, no contract. Schweet...that's the kind of pricing that will attract more users, and AT&T Wireless, Verizon, etc, are going to have to match that or lose huge numbers of data customers.
So now my Sony-Ericsson T68i is a full-time Bluetooth-enabled wireless hub for my Palm Tungsten T, my G4 Powerbook, and the phone-based data services. No, it's not fast, but it's better than doing data via regular cell dialup, and it's available all the time, anywhere, and for $20/month with all-you-can-eat service, it's the thing for me. When I'm at a hotspot, I'll use Wi-Fi, but for the rest of the time, GPRS does the job.
Cellular networks are run by companies that most emphatically do not have your best interests at heart. These are the same companies that, to pick the most contemporaneous example, spent millions of your talk-time dollars fighting number portability.
But getting soaked on per-byte and per-minute rates isn't the real issue. The bigger issue is that the cellular networks are end-to-end controlled by the telcos. No service flows over their waves or wires without their approval. Nothing happens on "your" handset that they didn't explicitly pre-authorize -- the exact opposite of the Internet.
If you want to experiment with a new idea in networking, and you're on the Internet, just park a machine on the net with an open port, and try it out. You don't have to get anyone's permission.
On the other hand, if you want to try something out on the cellular network, you first have to get "permission" to write software to run on your own damn phone, then you have to get the Service Provider du Jour to agree to send you the bits you're interested in (for a nominal fee, of course, assuming they know what the hell you're talking about at all). For an example of this idiocy, check out the API for BREW (Basic Runtime Environment for Wireless). It's the only API I've ever seen that expressly has a political layer in its networking stack.
802.11b, OTOH, is just Ethernet-like frames over the air (typically used to carry Internet protocol, but can be used for any other Ethernet-y thing). No permission, no politics, no fellating the local telco to give you "permission" to experiment. Just squirt packets out the interface and see what happens. Yeah, it carries data, but where is it written it can't carry voice as well?
Cell phones are darned useful, but believe me, you do not want the cell phone providers to become the dominant force in wireless data transport. They will screw you.
Schwab
Editor, A1-AAA AmeriCaptions
But 802.11x is high-bandwidth, and often unmetered ...
Yeah, but it's also a pain in the 4$$.
Currently, I have a cheapo Verizon cellular phone that with the over-priced mobile office kit, gives me 19200 max Kbps service - and for my needs, it's actually enough.
The cell companies are just now beginning to experiment with umlimited plans, they'll be common in a few years, as will the 2.5G/3G networks that make high speed Internet possible.
Combine the two trends, and in a few years you'll get 128 Kbps or better speeds on a cell phone with unlimited minutes for probably $50-$75/month.
One of the wonderful things about cellular technology is that there is still actual competition which drives innovation and tends to push costs down. Compare to land-line telephone companies, or Microsoft
Nice to have competition, eh?
802.11 can and will continue in places where the consistency of the user experience can be assured. This means in company buildings, warehouses, people's homes, and the like.
Just to indicate how *bad* 802.11 is, I tried to connect a neighbor into my home network with 802.11, and to go (I kid you not!) 75 Ft through the wood construction house in the middle, we had to get special directional antennas.
My cellular phone, however, gets reasonable service just about anywhere, even standing in the middle of our metal-and-plaster construction local hospital, and the nearest cell tower is over a mile away. (I know where it is)
I have no problem with your religion until you decide it's reason to deprive others of the truth.
What the article and poster are both missing is that cellular phones and WiFi networks have entirely different scopes of operation.
WiFi is intended for short-range use and provides large amounts of bandwidth (54Mbps) completely unmetered.
Cellular is intended for long-range use and provides small amounts of bandwidth which are typically (although not always) metered by the phone company.
It's the old adage: bandwidth, distance, cost. Pick any two.
Furthermore, while I don't see cellular overlapping in scope with WiFi, I see WiFi overlapping cellular in many areas. WiFi *can* be made long range with the proper equipment, and can replace some of the functionality of cellular.
The one thing you are missing here is relative scales. Cellular products working on 900mhz or whatever are just the same sorts of antennas that WISPs use. I worked on the deployment of a NAN in downtown Toronto and the technology of WiFi is equally viable as throwing cellular antennas up on towers. There are a couple of actual issues here.
1) Tower space is very expensive. The room for antennas is quite limited for new WiFi installation (on the same scale) as 9xxmhz or other cell equipment generally occupies the ideal antenna locations. Additionally the LOS requirements of 2.4+ghz WiFi make this a little bit less ideal than does lower frequency (thus better penetrating) technologies.
The lack of open source or even inexpensive management and network operations software / hardware make it less feasible (at the moment) than cellular. Just the general lack of cheap units with dual radios makes long range roaming fairly impossible. This will change in the short term, provided that WiFi doesn't go the route of cellular and become a per-minute per-packet large-corp-only business.
The real advantages of WiFi come with its robustness as well as its bandwidth potential. Once you start using Technologies Everyone SHOULD use But Don't(tm) like ipsec you start to have somewhat higher bandwidth needs. Because larger bandwidth with present technology =! higher costs (in fact SHOULd equal lower costs if it wasn't for strong telco resistance) little things like crypto and routing overhead should not make a difference. Nor should p2p traffic, streaming, etc.
The BIG advantage of using WiFi or mixed WiFi / hardline service with good QoS and intelligent ToS / cost-based routing / p2p filtering (i.e. slow down foreign packets / encouraging mature protocols (i.e. torrent, edonkey) is that the major costs of P2P become much much easier to swallow. Hint: multicasting p2p over WiFi is *really* efficient. = )
For those out there who have $100K/year jobs -- a economically unsustainable and generally damaging concept in and of itself -- go ahead and blow your packet charges on cellular. Enjoy 19.2k connections.
For those who still have a glimmer of the spirit of WiFi, the spirit of a free internet and people who enjoy being able to downloads hundreds of gigs a month of whatever they please -- WiFi has much, much MUCH more potential.
Gregory
--- ask me about nihilism, I will have nothing to tell you.