Can Mesh WiFi Solve The Net Neutrality Issue?
from the not-likely dept
We’ve pointed to two separate, but equally insightful, pieces on the real issues concerning network neutrality (which is, basically, that network neutrality isn’t the issue — a lack of competition is). The first piece was by Tom Evslin and highlighted why you can’t trust either the telcos or the regulators to do things right. Then, yesterday, there was the piece by Andy Kessler, suggesting we go after the telco-owned networks to force the telcos to generate some real competition. Evslin has now come back and written another long post about the situation, in part responding to Kessler’s piece while also filling it in with some ideas from an event he attended. He comes up with a two part solution to the competition issue and net neutrality. First, instead of talking about net neutrality, he suggests focusing the debate on America’s Antiterrorist Network. The idea being that, if we actually had better, faster, more reliable broadband networks we’d actually be able to do a much better job protecting the country (economic boosts would just be a side benefit). Of course, in some ways this seems on the same level of propaganda as the two main players in the debate over net neutrality keep rolling out.
His second suggestion is perhaps more problematic. He’s betting on mesh WiFi as the answer. In fact, some of the telco supporters (the same folks who used to bash muni-WiFi) now point to such mesh WiFi efforts as proof that there is real competition already in the broadband market. If only it were so. As some of us have been pointing out for years, WiFi technology really is not designed for this sort of usage — and early results have supported this position. Story after story after story highlight how wide-area WiFi is a lot more complicated than many in the industry (and the press) would have you believe. However, even if these networks worked flawlessly up to expectations, it still wouldn’t be an answer to the net neutrality issue — because even a perfect WiFi network can’t handle the same level of usage and doesn’t come close to providing the bandwidth of a fiber network. We can hope that other wireless technologies will be the solution — but so far none of the upcoming technologies have been shown to be even remotely effective in competing with a fiber line. Cellular networks are cutting off users who use over 10 gigs a month of bandwidth, claiming it degrades network performance. If it can’t handle that much, it’s not a real competitor to DSL, let alone fiber. WiMax still has a long way to go before it shows it’s a viable technology at the same scale — and even then there are spectrum allocation issues that need to be dealt with. Meanwhile, Tom Evslin, himself, explains why satellite also doesn’t represent real competition in the space. We can still hope that future generations of these technologies will be much more competitive — but we’re talking many, many years (and even then, it will still depend on spectrum allocation issues). So, using anti-terrorism as a peg for fighting for net neutrality may be useful in stirring up some debate — but betting on mesh WiFi seems like a losing bet.
Comments on “Can Mesh WiFi Solve The Net Neutrality Issue?”
yes, yes, yes, it’s about competition.
I wish the US commentators on this subject would actually look at what’s happening in the rest of the world instead of just blindly dismissing it as “of course it could never happen in the USA”. The telco/cable hold over the last mile is an artificial government sanctioned monopoly. It’s time for government to change the rules slightly and force it open. If that means LLU and forcing them to sell wholesale bandwidth, then so be it.
Over the last century we’ve built a water, sewage, electricity, phone and cable TV networks. Why is it now impossible to layer another network on top using fibre? It’ll take players with deep pockets and good lobbying skills but it can be done, surely?
mesh doesn't work?
I have looked mesh networks more than the average tech person. Often people say they will not be able to handle the traffic to compete. A mesh network would not have to work the way the current internet does. A mesh network could be part of an overall reconfiguration of the internet. One example of how it could work differently and better would be to localize where information is stored and retrieved. This could involve utilizing free storage space near by and dupilcating a great deal of files. Commonly sought files could be stored in your neighborhood instead of on the other side of the world there by making downloading an easier task for a mesh network.
What about instead of using wifi as we know it, we boost the power, and get the fcc to require a lisence only from those who run the towers, then you connect the towers either by long distance wireless links or wired links. Do it more like radio stations but two way, if thats possible. I have no idea on how any of this would work, but I figured if I asked here, I’d get people telling me why it couldnt work, or get them thinking on a way to make it work.
We already have this. It is called the cell phone network.
Do it like radio?
Sniff Sniff. (pun intended) I smell mass DoS.
Even if Wi-Fi, satellite, and other technologies remain niche products, they can provide a useful basis for benchmarking the actions of cable and telco operators. So the fact that they may not rise to the level of being a potential total replacement for those networks does not mean they won’t be useful.
Mesh and Radi
In reply to comment #1, what you are saying is true, but don’t you think saying “mesh will work fine if we just change the Internet” is a bit of a long shot?
As an answer to the quasi-questions in comment #2, one of the biggest problems of Muni WiFi is that the signals are ALREADY too high powered. There are only 3 distinct, non-overlapping channels in WiFi b and g at 2.4GHz. Boosting the power of public networks beyond the legal limit, which they already use, would cause a great deal more intereference on all other devices using the 2.4MHz spectrum. Think of cordless phones, baby monitors, remote surveillance cameras, and existing WiFi LANs inside houses and offices. The current FCC laws (and a fair use of the unlisenced spectrum) would not permit that kind of destructive interference. Other logistical problems would come up, like who gets the licenses. Well, auctions could solve that, but who wants to bid for the right to broadcast high-power, yet over unlicensed spectrum? That’s still a risky investment. Next problem is that even if the towers are very high power, the CPE is not. That means that although you can hear the tower, the tower doesn’t “hear” your laptop. Your laptop will report a strong signal, yet you won’t understand why you can’t connect. Jack up the output of every CPE out there, and you have significant cost increases, battery life decreases, and you’d have a cacaphony of interference in the three 2.4GHz channels, and basically nobody would be satisfied.
What’s the fundamental, underlying problem here? The techology is using unlicensed spectrum. To do so inside your own four walls with low power works great. To do so outside of your four walls, and at higher power means interference beyond your control.
Well, there are rudimentary phased array Wi-Fi sets already turning up in the computer supply catalogs, eg, a router with seven antennae for $130, and a matching PC adapter for $110, as against $50 each for conventional Wi-Fi units. Now bear in mind that telephone bills run in the ballpark of several hundred dollars a year or more, and it seems reasonable for a long-distance Wi-Fi unit to cost a thousand dollars or more. I’m not sure exactly what that would buy, but it should be fairly impressive. By my calculations, a phased-array Wi-Fi antennae with an effective diameter of two feet should have an angular resolution of six degrees at 2.4 Ghz, and three degrees at 5 Ghz (“802.11a” band). 20 Ghz would be better, of course (0.75 degrees), but that might involve another round of politics.This would be a set suitable for mounting on a home rooftop. A set mounted on a communications tower could of course be rather bigger, with a corresponding improvement in resolution. Phased array antennae have superior ability to reject off-axis signals– that is one reason why they are favored for military radars which have to resist enemy jamming. Effectively, what something like this gives you is a microwave relay with automatic search and lock-on. Of course, the seemingly intimidating price tag will diminish by annual Moore’s Law increments, to the escalating panic of the telephone companies (“First they ignore you/ Then they laugh at you/ Then they fight you/ Then you win!”).
Unlicensed spectrum is not a defect, it’s a feature. You just have to use it intelligently, in a highly focused way.
Anyone remember Microsoft citing Linux as proof that competition existed in the OS market during their Antitrust trial. What makes the Telcos citing no existent competition any different, saddly most regulators just take what hey say as truth.
WIFI net neutrality
WIFI is the last mile solution for those who cannot get anthing else. It does not provide access to the backbone of the internet AT&T MCi ect.. You have to use the fiber from the Long Distance lines at some point. The WIFI for the masses is very complicated and has yet to prove it’s reliability over time. To low powered, Not enough bandwidth, unregulated frequencies, interferance from othe 2.4ghz stuff
means dropped packets and frustration.
Wifi Outdoor Long Range Solution
I dont know if this answers some of the questions or wonders that were raised on the posts. But i would like to get your opinion on this latest antenna technology.
Re: Wifi Outdoor Long Range Solution
Sorry forgot to add the URL
re: phased array
in response to comment 6 about how phased array will save the day:
wifi technologies have been around for a number of years, as have dsl / cable. although true that wireless has improved over this time, and speeds have improved, interference issues are largely where they always have been, and prices have not significantly dropped (except for clearance equipment, which does not include the features you say will save the day)
incidentally, dsl and cable have also improved during this time, and at roughly the same rate. because dsl / cable are wired technologies, they do not have to have complicated signal processing to deal with free-space interference and losses. and because they are wire based, they have access to as much analog bandwidth as the wire can physically transmit, which is quite significant compared to how much analog bandwidth you can use in a wireless setup without expecting interference from others.
in other words, wireless may get better, but it does not get better faster than wired technologies, and therefore will remain a niche market because of its high price, low reliability, and low bandwidth.
you could say “but once this awesome technology / fcc regulation is passed, wireless will rule for all ever!”
or you could say, that once there is a wide scale deployment of fiber, that wireless will be forever relegated to niche markets.
i dont see the killer wifi technology development or regulatory hurdles being passed happening near as rapidly as fiber is being deployed.
I have to agree that currently Mesh networks are probably not a viable solution due to bandwidth and reliability.
however, I believe wireless technology is the future. Why lay down fiber lines when you can just send a signal through a different medium, the air/space. After all, a wireless signal is just a lower amplitude wave than light, and one day we will figure out reliable and long distance wireless communication.
Let’s all dig our back garden and mesh fibers. If you haven’t reached four neighbours, just keep digging, just keep digging…
wifi as an ISP is effective in rural communities as it is. But in my opinion, it needs to be modified in order to act efficiently as a mesh. If I were in charge of building a WiFi Mesh I would do it like this.
Use either 5GHz or a licensed proprietary band for the backbone.
Each “tower” would have its own channel to recieve on, while it must be able to broadcast on up to 4 or 5 other channels simultaneusly.
Each “tower” connects to at least 2 and preferably 3 other towers.
“Towers” would act as routers seeking around down towers and congestion.
Each “tower” would have a high power (around .5 – 1.0 watt) omidirectional 2.4GHz radeo.
Each “client” would have standard hardware equipped with high gain directional antenna’s
said system would have little overhead and could support around 100 – 250 users per radio / per tower (according to specs of 802.11b) while maintaining a dedicated back bone. backbone equipment would require special programming and perhaps proprietary hardware.
We have been looking at mesh for the past few months for use by a single entity(Our District employees). For small number of users with specific needs and a narrow bandwith (5-11 Mgb) this appears to be a reasonable solution. That being said we are also looking at several 802.16 (WiMax) solutions as well to cut down on the number of AP’s required.
All in all the cost for a 54 sq mile area is still beyond the realm of most organizations and is really not justifiable at this time.
Your skepticism is healthy (and comments on this post are great). I’ve posted some more on America’s Antiterrorist Network – this time on distributed data storage which, not incidentally, helps make net works more practical.
new post here
Mesh WiFi correction
should have said that distributed storage makes MESHED networks more practical.
Phased Array For Strategic Results
You are not going to get public funding for mesh WiFi networks in the foreseeable future. What you need to carry the project forward is a body of enthusiasts. The obvious candidates are the Ham Radio Operators. Many of them will already have their own radio towers, and they will have the skills to install additional equipment. This is where phased-array comes in. You don’t want to try to sell the ham radio operators something mediocre and cheap. On the contrary, you want to sell them “heavy iron,” something awesome and expensive, though not impossibly expensive. Five thousand dollars is probably about right. There has to be room for “homebrew” systems.
That is where Robert T. Morris has gone wrong with his Roofnet Project at MIT. The Roofnet Project is too democratic, and insufficiently aristocratic.
Morris is spending too much time worrying about how to hook people up for a hundred dollars each, and his core constituency seems to be the impoverished student living in an apartment rented from an urban slumlord. That is all very well, but it leaves Morris bottled up in the urban slum district. Roofnet as presently constituted is tactical, not strategic.
For strategic results, you have to make it fun for the participants. Given the right circumstances, phased array WiFi equipment might be able to reach out for twenty miles. Again, given the right circumstances, a very small number of elite enthusiasts, a hundred or so, might be able to run a megabit link from Florida to Quebec, probably along the line of the Appalachian Mountains.
As you say in the original piece, WiFi was never designed to do this job, so why do people insist on adding bells and whistles to something that cannot fundamentally provide municipal area coverage?
Althoug WiMax may not be perfect, it is much better suited to performing this task and (in my experience as a WiMax provider) does it not ony well, but with ease.
Well I have had my WiFi ISP for about two months now and it recieves from the outdoor antennea from line of site. I pay $35/month for it and it works very well for the most part.
Mesh WiFi is only as good as the ISP’s backbone. Lets say the ISP is operating on a T-3 or T-4 network – then it would be really nice! Mine is operating on a T-1 line. Still it isnt that bad – about like low grade DSL.
My ISP operates on B protocol which in my opinion isnt as good as N protocol which overlaps and gives the same range – all with the added speed of N. I think that if he internet were entirely wireless then N protocol would be the way to go. There are a lot more cordless phones and other wireless equipment that operate on 5 ghz spectrum. If we could reserve the 2.4ghz for WiFi – that may rid the problem of high powerd licensed towers for the backbone.
For the home user to get onto the backbone – they would still rely on the 2.4ghz low power towers but these arent really that bad especially if I had my way and they broadcasted on wireless N!
There are a ton of mesh wifi implementations now. Sadly, the Wikipedia article is only showing a small fraction of the solutions.
If you make another article on this topic, can you please send me a notice at my email address? Thanks, Joe