Coffee Space


Listen:

Next Generation Encryption

Just a quick idea this time. It seems as if encryption is reaching an issue where people with massive super computer capabilities are now able to hack what more modest computers might consider to be difficult to do. Nobody wants to spend the time and resource to do this but everybody wants to live off the benefits.

My proposal is a de-centralised protocol system that uses numbers generated and tested by the end user of the protocol to better itself. With many devices using the system, no sup[er-computer would be able to ever touch the protocol as long as it remained popular. After all, the majority always wins.

I'm still figuring out the details, the hardest part I think is verification and authentication - issues that have plagued security systems since security began. Once a secure connection is established and identity can be confirmed, I think it is secure. Making sure that no un-wanted users become part of the protocol in the first place is in my opinion the difficult part. A person or group wanting to sabotage the protocol would merely have to generate and authenticate weak numbers and you suddenly have a weakness.

Perhaps the end user self-authenticates the number regardless. Usually this process is much easier than initial generation. There are already a few nice equations for this, specially concerning primes. No one-user could perform all the checks and the idea is that the load is shared, so there would need to be some way of determining how to generate new primes amongst many users without the prime generation being sabotaged. Perhaps these are generated in trusted networks and then shared? Or the problem reduced statistically?

This certainly requires more thought...