[forwarded message originally from Shae Matijs Erisson on 2006-10-11.] Kragen Javier Sitaker <[EMAIL PROTECTED]> writes:
>> I completely agree with you, though I missed the original post. > It's archived on FoRK and kragen-tol --- are you subscribed to either? I think I was... I keep meaning to get around to subscribing to the other mailing lists, but I like your earlier suggestion of a single signon to for all the lists. >> I wish for a peer to peer wifi that has many miles of range. >> Then you can use an ad-hoc network with voip. > > Yup. I think there are lots of problems to solve there first, though. > >> > Not at the low end, although you can run Pippy or some Schemes on the >> > Treos, >> > and Symbian Python on the Symbian phones. So far I don't know of any >> > phones >> > that can compile their own operating systems. >> >> The Nokia 770 (not a cellphone) runs maemo, a derivative of Debian. > > Yes, it looks quite nice; my friend Jesse has one. Does it have > enough space to compile its kernel? With a 2gb RS-MMC card I'm sure there's enough space, but with the 225MHz cpu patience would be required. >> I've thought about this a lot lately, and the only solution I see is >> to sell a software radio handset with as wide a range as possible >> and let people implement their own protocols, whether they be wifi, >> gsm, or whatever. >> >> Basically, forget the wireless protocol standards, don't worry about >> trampling your neighbor's signal, make it easy to develop and deploy >> new encodings, and see what protocols emerge from the chaos. > > Yes, I think that's an important part of the solution. There are > several things we have to solve to get there: > > 1. Regulation. Is there a way to end-run this at first? Sell in a > country with a large informal sector, where such regulations are > hard to enforce? Restrict the hardware to some frequencies or > power levels that can legally be used without a license, in some > country? I don't have any good answers for this. The FCC is paranoid about wifi cards, much less software defined radios. > 2. Cheap hardware. Someone has to design, build, and test it before > anyone can use it. How are your RF hardware engineering skills? > Mine suck. Steal, borrow, etc ideas and design work from the GNU USRP: http://www.comsec.com/wiki?UniversalSoftwareRadioPeripheral Or maybe start a bounty to pay Matt Estes to design a smaller paperback book sized version? Maybe just ruggedize the USRP for backpack use? > 3. Killer apps. It needs some application (software and any needed > infrastructure) that runs on it that will justify its purchase > price. Maybe this is a cellphone network? Store-and-forward p2p > voice pagers? Multiplayer games? Software defined radio would have a surprisingly large number of uses. Things that come to mind are distributed detectors for nuclear explosions, time/gravity distortions, radioactivity, etc As for games, what about a "crack this encoding" that works like one of http://www.pythonchallenge.com/ http://www.matessa.org/~mike/dutil-dumas.html Voice chat would be the biggest sell, I discovered that 90% of my voice traffic has endpoints within 30 miles, and I'm sure that either an encoding or a mesh network would get that sort of range easily. Teenagers would love that sort of device, their social network is usually limited to their house and their school. Free calls to everyone in that community would save their parents lots of money. Maybe 'social distance' could be a way to discover a good voice mesh encoding? That is, 90% of humans live within distance X of their job and family. 90% of humans live in density X people per square kilometer or below. Then find an encoding that can handle the top end of that? Cheap $50 mesh phones would be popular in places where cellphone networks will never be setup due to economic or terrain problems. I've also been thinking about using ambient energy to power these devices. http://en.wikipedia.org/wiki/Stirling_engine http://en.wikipedia.org/wiki/Alkali-metal_thermal_to_electric_converter For example a mesh network node the size and shape of a pringles can (or longer?) could be stuffed halfway into a sand-dune or whatever in order to get enough temperature difference to power a small internal stirling engine. It could also include the self winding watch kind of power, off balance metal disks (spheres?) that swing around when the device is moved, carried, etc. I've been thinking about a way to do that same thing with sloshing liquids but haven't figured it out yet. Any ideas? There's a lot more radio information in the environment than anyone uses. How many crystals when stressed emit EMF that's SDR detectable ? If several common elements fit in there, you could detect earthquakes, passing vehicles, possibly even humans/animals in a distributed fashion. Or you could use this principle to detect underground temperature changes. I'm not sure how that would be useful, maybe find volcanoes? Maybe stick a spike into the ground and throw out various frequencies to do area effect detection of metals, caves, water table, fault lines... I wonder if you could use SDR to *write* onto existing crystals? Geocaching would be a whole new game if you could leave hidden messages... I think software defined radio is its own infrastructure, but it will need a heavy duty collection of signal well documented easy to use libraries. One missing piece of infrastructure is that you'd need either two SDRs, one to send and one to receive, or some way to multiplex existing SDRs. I discovered when reading about 802.11 frame injection attacks, I don't think any wifi cards can listen while injecting custom frames. Another important piece of infrastructure would be a library of antenna designs and ways to build your own, extending on the cantenna idea. For example, GPS resolution is ... a meter or so, I think ... if you built your own dual GPS antenna that was longer than that, you'd be able to get heading information as well as location. FPGAs are good for reconfigurable hardware, but does that exist for antennas? I could go on, but I know I'd have had enough if I were reading this email... -- I've tried to teach people autodidactism, | ScannedInAvian.com but it seems they always have to learn it for themselves.| Shae Matijs Erisson
