Use public key crypto for authenticated message transport

[TKussel]

At the moment the message integrity is evaluated using an MAC scheme. We could switch to an asymmetric crypto based approach and have the sender sign the message. This would improve the security, as impersonations and MITM would become impossible (see issue #88), but it would also increase the complexity. We would need a mechanism to generate/import public/private keys as well as fingerprint validations of all other parties.
@prasser @fnwirth What do you think?

[fnwirth]

That would mean we need a new type of interface to exchange the public keys?