ito-org / strict Goto Github PK

It's probably impossible to prevent replay attacks but it should be mentioned in the doc.

A replay attack would mean that you can collect beacons and re-broadcast them at different places. If you have collected enough tokens at a crowded place and manage to re-broadcast them somewhere else as long as the tokens are still valid, you might be able to create a virtual epidemic. With people colluding that might even be easier. I guess you might preclude this scenario under "Clients are assumed to be individually malicious, but not colluding at scale.". The question is how many damage can be done without big scale but e.g. with just two persons colluding, one collecting tokens and one broadcasting them at a place with people you want to harm.

A related attack is to create, boradcast, and report tokens without being infected at all. This is somewhat alluded to in the doc as "false positives" but the threat model (and counter-measures) is not explicitly described.