[servowire]

With a DVB/Radio link to the network one could listen to transactions only to their address. You could think of bitcoin operated gas stations, bitcoin network-hubs in rural areas, micropayments to car-charging stations (all you need is power and a cheap DVB receiver to listen to the address "beacon" for a payment.

Now sending back transactions to the network is not covered, but can be done by using SMS or a small radio link to a local node. A bitcoin transaction is only about 64 bytes (!) so even a DTMF call or SMS should work fine.

A transaction send would look like this (normal TXid, in hex)

  a9d4599e15b53f3eb531608ddb31f48c695c3d0b3538a6bda871e8b34f2f430c

That would be

  76816061505687581068522220805344294890167166251289478281934616246172205973504 

in DTMF. Would take about 15 seconds to complete (http://www.audiocheck.net/audiocheck_dtmf.php)

[Buge]

    a9d4599e15b53f3eb531608ddb31f48c695c3d0b3538a6bda871e8b34f2f430c

is just a transaction ID, not a transaction. A full transaction is longer. For example [1], a typical transaction with 1 input and 2 outputs, is 226 bytes.

[1] https://blockchain.info/tx/2e395e9567bd43d3eecd7a6bdf990387d...

[asdfaoeu]

Why not just send the confirmation back by SMS?

[zouppen]

The idea is that the payment receiver doesn't need to do it. The payee needs to have some kind of uplink (like a mobile phone, amateur radio, or anything).

There are mobile payment applications already, but they are doing it using centralized service and they need make deals with payment agents. With Kryptoradio it is possible to receive payments without any extra middlemen.

[servowire]

That would require trust - in the form of the SMS provider. No way to verify that SMS is valid, it could have been forged.

The proof-of-work blockchain transmission means it's the real and proper blockchain and is valid.

[wyager]

>Would take about 15 seconds to complete

That's unfortunately not sufficient at modern transaction rates. Is there any reason you chose DTMF in particular?