Microsoft Sniffs Out Unused Wireless Spectrum 102
alphadogg writes "Microsoft researchers have designed a scheme for measuring whether licensed radio frequencies are actually being used so unlicensed devices can use it, something that may become necessary as demand for wireless applications grows. The architecture, called SpecNet, would sense and map where spectrum is being used and more particularly where it's not — so-called white spaces, according to a paper being presented next week at the USENIX Symposium on Networked Systems Design and Implementation in Cambridge, Mass."
Isn't this contradictory? (Score:2, Interesting)
Re: (Score:1)
Well, if home sites used a web proxy cache (eg: Squid) and sites didn't rely on massive amounts of bandwidth, we'd not have that problem.
Also, if wireless routers used directional aerials once a connection had been established, you'd have less risk of interference between devices and therefore more bandwidth available (since there'd be less sharing). In fact, you could probably get away with just having a bunch of fixed-position loop aerials (12 should be sufficient) and scrap the omnidirectional reception
Re:Isn't this contradictory? (Score:4, Insightful)
No, no, and no.
1) you're not going reduce the demand for bandwidth, just like you're not going to make taxes go down, for long.
2) the whole point of wireless is mobility, not fixed point-to-point multicasting. The 802.11a/b/g/n frequency allocation sucks, although the a/n that uses 5Ghz has more non-interfering channel allocations. But density is not in your favor no matter what wise-ass antenna you try to use. There's leakage and uncontrollable other-device-location that will always thwart your design. Some over come this, but it's an endpoint problem that's really not covered at all by this misadventure that Microsoft is embarking on, and
3) You have no clue what it takes to do rural broadband, nor the problems of how twisted pair networking operates, over what kind of distances, and to what degree of external signal problems.
Go fish.
Re: (Score:3)
There's a lot of dark fiber out there.
But wiring it up IS a problem. You need poles, easements, contractors, and terminations. None of it's cheap. And fiber is a great solution if you can afford it. Some people have used Hughes sat dishes, which work ok if you don't mind the latency and cost.
You need people to fix it after bad weather, that have pretty expensive rolling costs, like training, trucks, TDRs, repair gear, and you have to pay them, their insurance, and so on. It's not a trivial endeavor.
We need
Re:Isn't this contradictory? (Score:5, Insightful)
I've probably lived in more rural communities in more countries than you could shake a stick at, so cut the wise-ass remarks. If I say I know damn well that you can get CAT6 to people's houses, then I suggest you start by asking how, not telling me that it can't be done. I won't say the problem's not solvable unless I've actually done the work to know it is solvable and have the engineering skills to know what the limits of theory are in practice. The people who get things done are not the ones who say it can't be done. The ones who get things done are the ones who establish IF, WHEN and HOW -- questions you utterly fail to ask.
If you haven't asked those questions for each and every damn article you read and each and every post you reply to, you have failed.
Re: (Score:2)
If you're going to run a cable to ever house, don't, for the love of god, make it copper. The install costs would be the same and you could be putting fibre in the conduit rather than copper, that way you can upgrade it without ever digging it up again
Re: (Score:3)
I've probably lived in more rural communities in more countries than you could shake a stick at, so cut the wise-ass remarks. If I say I know damn well that you can get CAT6 to people's houses, then I suggest you start by asking how, not telling me that it can't be done. I won't say the problem's not solvable unless I've actually done the work to know it is solvable and have the engineering skills to know what the limits of theory are in practice. The people who get things done are not the ones who say it can't be done. The ones who get things done are the ones who establish IF, WHEN and HOW -- questions you utterly fail to ask.
If you haven't asked those questions for each and every damn article you read and each and every post you reply to, you have failed.
Why the shit would you run Cat6 rather than fiber? Cat6 is only good for 100 meters, it's not capable of anything close to the capacity of fiber, and it's susceptible to electromagnetic interference. The fact that you're suggesting running Cat6 to every home in a rural area shows that you have no idea what you're talking about.
Re: (Score:3)
At lower data rates, Cat6 is good for longer distances, but I also disagree that it makes a good choice. Even shielded twisted pairs just aren't good for rural distances. This is the reason why DSL doesn't go far from a central office: the signal degrades.
Re: (Score:2)
I'll leave you to re-read the bit about dark fibre. And then look for the bit in my original post about running CAT6 across rural distances (amazingly, you won't find it). And then I expect you to say something sensible.
Re: (Score:2)
I'll leave you to re-read the bit about dark fibre. And then look for the bit in my original post about running CAT6 across rural distances (amazingly, you won't find it). And then I expect you to say something sensible.
I don't see anything about dark fiber, and you seem to be wanting to run Cat6 across rural distances. If that's not your intention, please explain how else you would go about doing so. If you suggest that we do so by running fiber to nearby distribution boxes and then running Cat6 to the home, why not just skip the extra hardware and gain some extra capacity by running the fiber straight to the home?
Re: (Score:2)
Then I will paraphrase what I've said across the posts.
First, plot the rural end-points.
Second, group them such that each group can be fitted to a line, curve or elipse, such that the distance required to build a spanning tree* between the groups plus the sum of the length of the groups is minimal AND the shortest distance from the line, curve or elipse to any given point within the group is minimized.
Third, fill out your spanning tree and groups with fibre at a depth at which ground temperatures will remai
Re: (Score:2)
Allow me to present to you, your average picture of rural California [goo.gl]. Now how exactly do you propose that we run such a complex ring of fiber around such irregular areas with individual homes spaced out far more than the hundred-meter limit imposed upon gigabit-or-less Cat6 run lengths? The way I see it, every fiber tap will only be able to support one Cat6 drop simply due to the sheer distance between locations, which once again returns me to this question which you have yet to answer: Why bother with Cat6
Re: (Score:2)
I'm looking at that picture, particularly North May Road. I'm seeing houses that can be served by communal taps there. Fiber optic =modules= (never mind the rest of the router) can run upwards of $8000 a throw. Buy one for two or three houses and run cheap lines from the communal point, you end up saving a lot of money. Remember, the object is to prove this can be done cheaply. It may not be the best technologically, but if I can halve the costs without impacting bandwidth then my solution is (by definition
Re: (Score:2)
So now you have the task of convincing farmers to let you dig up their land to place telecom infrastructure, and ensuring that they're placed in such a way that they won't possibly be compromised by the activities on the surface of the actively-farmed agricultural land. And you again have yet to say how it would be better and cheaper to run Cat6 (which requires all this extra planning, equipment, and space) when you could just run fiber next to the road and not have to worry about setting up special hardwar
Re: (Score:2)
Re: (Score:2)
Rural American farmers are convinced by Sarah Palin. That doesn't show me that it's particularly difficult to convince them of anything. I'm willing to bet that many believe ghost stories and the moon landing conspiracy theories too. And these are the people you don't think can be convinced to put some cable in? Besides, look at the picture again. Those three houses on North May Road? Mid-point is not in any of their land. It's on the road. So what's your problem? Don't like losing the argument with your ow
Re: (Score:2)
I'm not arguing that it can't be done, I'm simply arguing that it's more expensive than sharing the higher-end gear. You can get a gigabit router for around $100. On the Google Maps link we were given, there's a village off Acampo Road with around 7 dwellings per block on average. One fibre router and 7 gigabit CAT6 routers will therefore run you $1700. 7 fibre routers runs you to $7000. There's about 15 blocks that you'd need to take care of. That's a savings of $79,500 by using communal routers. That's a
Re: (Score:3)
No.
If you ran Cat6 to a rural home than you made an expensive and WRONG choice at best, and used a hammer where a screwdriver should go. TSB-67 cable isn't designed for rural home runs at all; if you got any kind of data rate, it was through luck, not design.
I, too, get things done, and with a great deal of experience running an enormous variety of installations, I don't believe you. If you have the how, then why are you using the wrong stuff for long end runs?
Re: (Score:2)
Look, it's quite simple. I said nothing in my original post about running CAT6 across rural distances and made it explicit in my follow-up that you'd use fibre for the long-distance (including "last mile") and CAT6 only at the last moment (since routers that support fibre are generally expensive, so you don't want more of them than necessary). In rural towns and villages you find buildings next to each other. Why have fibre to both when you can have fibre to the midpoint between them and 50' at absolute wor
Re: (Score:3)
I'm really familiar with business and rural broadband systems design. You'll have to trust me on that. Fiber is the best solution for the long run; it's better to terminate it and route indoors to a home cable plant of some kind.... call it a rural IDF.
Somewhere there's a communication breakdown; if on my part, I apologize. I've seen utilities try IP over utility wires in various configurations, modulated twisted pairs, point-to-point hybrid systems involving elaborate towers and pringles cans, and much oth
Re: (Score:2)
To defuse the tensions, I'll agree to a communications breakdown and we can ignore the blame game. Unless you want to blame Microsoft, I'll go with that.
Fibre to each house, although technologically the best solution, is problematic. Fibre-capable routers cost money. The best compromise is to have fibre to as few houses as you can absolutely get away with (which will be all farmhouses, some McMansions and isolates) and then have communal taps wherever two or more end-points are going to be within range of a
Re: (Score:3)
The fiber routers really aren't that expensive, and if you're sensitive, converting fiber to copper isn't that expensive, either. If you're going to go thru the expense, you might as well do it right unless geography and nearby drops are in your favor. Fiber terminations aren't that much more expensive than copper, and deliver long term viability that's hard to beat. If budget's a problem, then your up-thread spanning-tree suggestion is viable.
Ask the people in Loma Linda CA about what it's like to have FTT
Re: (Score:2)
I'd worry about anyone living in a place with such a weird name, but I think we're basically on the same page and I think we're in agreement that there is no practical reason (cost included) as to why any community, including rural ones, should have poor connectivity.
Re: (Score:1)
I'm setting up a smart proxy to reduce my bandwidth usage. hopefully by 10-20% and have more clients.
Re: (Score:3)
Cat6 has a maximum length of 100m (330 feet or so) for 1Gb Ethernet and below; faster speeds get less. Are you sure wiring rural areas with that is going to be cheap? You'll need more than just better cables; IIRC the distance comes from the travel time for a signal, and after 100m the latency is such that two stations will transmit at the same time, not knowing the other one is transmitting. So you'll need a switch or certain hubs (Class 2?) roughly every 100m.
There are plenty of areas with more than 100m
Re: (Score:3)
There is, as other posters have noted, plenty of dark fibre. Communities, ESPECIALLY rural ones (and my mother's side comes from North Dakota, so I know how rural things get), have a central point. They're not randomly dotted around.
Option 1: For a close community, you find a central point and run a tap from the nearest dark fibre to that point. That gives you a rural hub to run from.
Option 2: You run a ring of fibre round the community, again running a tap from the nearest pre-existing dark, with branches
Re: (Score:2)
It's being done already! (Score:3)
It's being done already, in other countries
My house is connected to 3 Internet providers. Each provider has a fiber-optic trunk line, mediaconverter and a switch with Cat5 wiring connecting individual flats to it.
Such providers are historically called "local local networks" (heh, almost like "ATM machines") here because they started appearing in late 90-s and early 2000-s when Internet connectivity was EXPENSIVE here, like 20 cents per _megabyte_ expensive. So operators of these networks provided 'local' re
Re: (Score:2)
Communism is Soviet power plus the electrification^Wnetworkification of the whole country. - V.I. Lenin
As for the GP, no one said you had to run ethernet over the CAT5. DSL will give you a lot more range than ethernet over the same copper. But it would be stupid to lay copper at this point anyway.
Re:Isn't this contradictory? (Score:4, Insightful)
I can agree that it can cause quite a racket if the FCC ever gets a report of abused radio spectrum.
And what looks like unused may not be unused at all but can actually be used for measurements, alarm systems or even remote detonations so you can't tell that it's unused by sniffing it.
Something like the parking spot right outside your window that's empty when you are at home - that actually is used when you are at work by the maintenance company that happens to have an office in the building you live in.
Re: (Score:3)
Re: (Score:1)
Yea, this isn't a technical problem. It's a legislative one. Meaning: forget thinking about it unless you have a few spare millions to bribe politicians.
Re: (Score:3)
Imagine for a minute you could build a generically unlicenced device, which shouldn't interfere with licenced ones. So it works on unlicenced space (or on licenced space but it can't find a free channel), now you have two choices, either the device can fail to operate, since there are too many devices. Or it can automatically go hunting for new channel space.
Ideally a device should be able to hunt around for free wireless spectrum, and then resolve if it can stay there when something else shows up. I can
Re: (Score:3)
This connotes that both sender and receiver know when channels switch. In Bluetooth, there's frequency hopping that allows this, albeit at really low power. But if your device is at one channel, and must suddenly shift away, then sender and receiver must know what they are, otherwise you're a broadcaster. Low power broadcast is ok, within certain bounds in the US, given certain spectra.
When the low power device interferes with something in a licensed band, it could be critical equipment, public safety, FAA,
Re: (Score:2)
Well I may not be understanding this right but that would basically just be a issue of firmware. I know I unlocked more channels for my wireless router by installing FOSS firmware on it and if using unlicensed channels became legal all you would have to do is update firmware for existing devices.
Re: (Score:1)
In traditional /. fashion, I didn't RTFA, but I did see a MS researcher give an eerily similar sounding presentation a couple months ago. He discussed an algorithm to detect users on the 700 MHz spectrum. This is a "licensed" spectrum, but has been opened up by the FCC for unlicensed use on unoccupied channels. A major issue was determining which channels were not being used, and how to adaptively change when a licensed user starting using the channel. They discussed proposing a protocol to the FCC specific
Re: (Score:2)
How do you propose we handle several devices transmitting on the same frequency in close proximity? I'm pretty sure Disaster Relief would be happy with your idea, especially when their CBs are drowned out by a local, still-transmitting radio relay playing Rebecca Black.
Or how would a device know where to look for a broadcast intended for it? The way it stands, WiFi for example, know to look for WiFi broadcasts around 2.4GHz, and on specific frequency bands. If we spread that out even further, how would the
Re: (Score:1)
Re: (Score:2)
Okay, I get it where my reasoning derailed. But still, it looks more complicated than single-frequency communications, especially frequency-hopping (which would make the most sense if we're talking about unused bands): synching up devices, and making a standardized format of checksums to identify each type to prevent one device using data from another type altogether...
Re: (Score:2)
Re: (Score:2)
Re:Wouldn't be necessary with spread spectrum (Score:4, Informative)
You're kidding, right? Instead of simply allocating the spectrum based on frequency, we have to allocate by frequency and time, synchronize all relevant devices, and build every device with proper radio hardware to switch frequencies rapidly. Maybe it's a small price to pay for the increased security of such a transmission... until the CIA/NSA/Illuminati/tinfoil-hat-enemy-of-the-week gets their hands on a single receiver, and they know the psuedorandom sequences involved anyway.
Most spread spectrum algorithms improve resistance to accidental interference, because they simply provide a "moving target". If two spread-spectrum devices are transmitting simultaneously, they will seldom interfere with each other during normal operation. If the interference is intentional, no amount of hopping or alteration will stop it for long, because the interfering transmitter can be designed to follow the same pattern, or simply broadcast on all frequencies the device will use.
Re: (Score:2)
That's completely impossible. Do you have any idea how much it would cost to make transceivers capable of using the entire spectrum? Do you know how long it would take for your wireless mouse to bind if it had to sweep through the entire spectrum to find the host? Want to change the radio station in your car? Better be prepared to wait a few minutes while it sweeps back and forth across the spectrum trying to find the channel (which itself is hopping around).
And even if it were possible, it would be stu
Snore (Score:4, Informative)
Microsoft has been in this space for years. They, for example, contributed to the original FCC TV white space trials in 2008 [fcc.gov] (see the February and March entries).
Re: (Score:2)
Re: (Score:2)
the military has been looking at this for a LOT longer than Microsoft - I heard some concepts for this back in the 90's
GPS signals are weak (Score:1)
Reminds me of the static IP address days (Score:3)
Re: (Score:2)
Ah, I remember that as a tech back at my high school. Only problem is, we had more than 254 computers, so we would be continuously knocking computers off the network. But it was OK, because it was almost never the same people. Then Netware would break, and things got interesting.
Eventually we figured out that we could assign the lab computers to a subnet. Implementing this took some fighting becase "it worked well enough".
Glad to be done with that...
Re: (Score:2)
If you had a single class-c, wouldn't subnetting make it worse?
A /24 has 254 usable IPs. A /25 only has 252 usable. For every subnet bit you set, you lose more and more usable IP space.
With that being said, subnetting (especially CIDR) is an invaluable tool. It breaks my heart that a lot of our newer network engineers just can't do it. They learned it to pass the CCNA and then went back to using CIDR calculators.
Re: (Score:2)
Can't say I'm a network engineer, so subnet may be the wrong term. We were all behind a NAT, so we just made a new /24
Re: (Score:2)
No problem. Yeah, NAT is the way to go for most networks.
If you have a large network, you can do a 10.0.0.0/8 and then subnet/VLAN that down.
Re: (Score:1)
Re: (Score:2)
You are correct that a /25 has 126 usable IPs. I was referring to the fact that the engineer has a single class-c to work with. If he applies a /25 to a class-c, he will have two subnets with 126 usable IPs in each; a total of 252 IPs for the class-C.
Re: (Score:2)
"They learned it to pass the CCNA and then went back to using CIDR calculators."
Just like I had to learn the capitals of the 50 states, then went back to looking it up when I needed to. No, I'm not a network engineer nor a CCNA; did have to do some network subnetting once in a (rare) while. There is a reason we human-like beings build tools- to make make stuff easier and quicker. But if you're happy memorizing stuff a program can do just as well, more power to you (seriously- I have respect for anyone th
Re: (Score:1)
Re: (Score:2)
Well, it was a high school and a middle school, separated by a block or two. They actually had a pretty decent 10 megabit (or so) microwave link set up between the schools, because they could only afford a T1 at the one and centralizing the network made sense (I guess?)
In any case, the middle school had 5 computers or so per classroom (to facilitate group projects), the library had about two dozen, and all the offices and other rooms had a few each. Probably about the same in the high school, except each te
Re: (Score:1)
Re: (Score:2)
Uh, this was about 6 years ago. But the computers might as well have been Apple ][s. To their credit, they weren't wasting money - the computer labs and library got the new computers on about a 3-year cycle (30 new computers a year or so), and the rest got 'trickled down' through the district on a need-based system - labs in the high school and the HS library got them first, those newish machines went to the teachers in the high school, and their only-slightly older machines went to the middle school and el
Re: (Score:2)
Only now, with RFC 1924, you don't even need to pick a random number. Just smash your face into the keyboard a couple times, and voila: 4)+k&C#VzJ4br>0wv%Yp! A perfectly usable address!
Re: (Score:2)
For most link-local addresses, yeah, it kinda works like that. Of course, the host is going to pull additional info from the default gateway and then *politely* ask if anyone else is using this IP.
If I read the documentation properly, there should never be a duplicate link-local. If a duplicate is detected, the newer host will just modify its address.
Re: (Score:2)
This is how IPv6 address assignment works today.
And how DDP worked c. 1985. The secret is a decent naming service, which AppleTalk had, so nobody ever worried about addresses. It worked very well for local networks. Lord knows why the rest of the world hasn't figured out mDNS's place by now.
Re: (Score:2)
Remember assigning static IP addresses by the seat of your pants? Pull a number, X, between 2 and 253 out of your ass, ping 192.168.254.X, if nobody answered, go ahead and assign your new network printer to it. Hey, what could go wrong?
Actually, the practice is getting standardized: ZeroConf [zeroconf.org]
What Net? (Score:1)
ANd after saying it couldn't be done (Score:1)
Re: (Score:2, Insightful)
It's just like when a crook starts to give to charity. You automatically start to look for the reason because you're just not used to him doing something nice.
And what about Ham Radio? (Score:1)
Re: (Score:1)
How about a global RF-based internet, controlled only by the UN and free to all people on the planet? High-speed high capacity links on microwave and UHF, slower-speed longer-range links on VHF and HF. Free internet, controlled by no government and no corporations, for and by the people. The technology exists. The frequencies exist. We just have to get rid of the hams and get the work done. The broadband monopolies would die, and their regimes of censorship and thought controls with them.
This would be a great leap forward for all mankind, and all we have to do is clear out the hams and then make it happen.
This is the kind of stuff the hams are SUPPOSED to be doing themselves, but they're too lazy and/or invested in the existing regimes to do it.
Why would the UN get involved with this? What makes you think that UN control would mean no effective governmental control? A UN-controlled internet would likely turn into a recording/movie industry-controlled internet. It would have every spy agency getting whatever access they wanted. Countries would force each other into it, and blame the treaties they signed so no one has to answer to their people.
Free? Yeah, right. Who would pay for it? Are you talking about a zero-infrastructure network? If so, it's e
Re: (Score:2)
Re: (Score:2)
Okay, I'll bite. It's been a while since I've wasted time arguing on Slashdot.
Ham Radio is just a big scam so that a bunch of Republican retirees can sit around with ancient equipment and chat with their buddies without having to pay for phone time or internet access.
'Cause ham radio equipment is so much cheaper than the phones that they already have, and that they still use? Of the few hams I know, they all have phones, most of them including cell phones.
They do nothing that cannot be done better by actual professional crews
Some hams are also radio professionals. One ham I met was also a (cell phone?) network tech, who became a ham operator because he wanted to learn more about the fundamentals of radio technology. Being a ham makes him a better member of his "p
What about the RF cochlea? (Score:1)
Re: (Score:2)
That's a really interesting design, thanks for posting it. However, from their paper [ieee.org] (paywall) it seems like the RBW (to use the spectrum analyzer term) is dependent on how many of these filters you put in. If you want a fine resolution over a wide band, you're going to need hundreds of these things per decade, and thousands overall. Certainly possible, and I wouldn't be surprised to start seeing them before long, but maybe not that much cheaper than conventional spec ans -- remember, a lot of that $10k
Not even new then (Score:2)
See US patent 6,768,398 [google.com], filed 12 December 2001. The RF cochlea is a relatively old idea.
This is BS (Score:4, Informative)
The Microsoft thing is BS.. not the idea in general.
The whole FCC idea of "Whitespace" is that we have a huge chunk of the best overall spectrum put aside for OTA television. But in most areas, most of that spectrum isn't used.. even given the losses due to original cellular (channels up to 83) and the more recent 700MHz auction for 4G (channels in the 60's on UHF).
So the idea of whitespace radio is simple: treat it as ISM radio (like 900MHz and 2.4GHz in the USA) once you acertain that the channel (in 6MHz chunks, just like TV, in the USA) is not used.
The problem is, just using sensing, you can't know if the channel you pick is clear. Your receiver can go into spectrum analyzer mode and not see a thing, but it's still very possible your transmitter is going to interfere with the guy down the street. who for whatever reason (rooftop antenna with 40dB LNA) can actually get that OTA channel.
Thus, the current plan for whitespace radiio... radios need to be location aware, and only use channels legal for that specific location. This is trivial to do, and it pretty much just works. Nothing MS is doing here improves this, far as I can tell. You can't be correct about the usability of a channel from a single monitoring point, whether you spend $100 or $100,000 on that spectrum analyzer. And so, given the need for one node in the network to have a separate internet connection, nothing MS does online is an improvement over the basic idea -- we absolutely know where the licensed radio is, because it's LICENSED! That license is for a certain areas, and no army of MS spectrum analyzers can be certain that your neighbor can't receive that channel, within the licensed area. Beyond that area, it just doesn't matter -- you get to use that channel anyway.
It's all fun and game (Score:1)
Until the device goes "Oh look, nobody's using this, I'll grab it" and it happens to be an emergency frequency that HAS to be kept clean for, well, emergencies.
I dunno about your country, but in mine you better have a DAMN GOOD reason to use certain frequencies that are "unused"...
Re: (Score:1)
I'm pretty sure that if such a device was approved by the FCC it would be programmed to avoid using emergency frequencies.
Wasn't Google doing this a few years ago? (Score:2)
Wasn't Google doing this a few years ago and own a patent or two on it?
If I recal correctly, the FCC disallowed this technology to be used in the upcoming Google phones.