graphs for the fair one

27 Nov 11 Leave a comment

Categories: Uncategorized

for mog

07 Nov 11 Leave a comment

Firstly, sorry, I’m so lagged out on the actual site I can hardly post anything at the moment.

Secondly, imo, encryption is not the issue. It’s denial of context that’s the issue. Suppose you’re looking for a pattern in movements. You don’t really want random person to know precisely where you’re interested. This can be done through shifting the coordinates of each pathing point by some amount (on the x, y, and t coordinates). So long as you’re consistent within the current datablock being parsed you can translate this back to meaningful data on your end. If this is different for each block that’s sent out and you’ve a system in place to keep the data from accumulating, making sense of all of it becomes very difficult.

You can further increase this difficulty by ignoring coordinates in the data altogether and just focusing on having the distributed system analyze vectors (possibly even splitting the vector, into it’s x and y components sending the x and y in different data blocks for reassembly later).

Then, you should also add in both dummy blocks and dummy data. True enough they’re taking up computing time, but there’s definitely an optimum ratio to pump up the noise.

I admit to not being a cryptanalyst so this might not be as robust as I believe, but I think that through encoding, alteration, and padding you can make the mass of data you’d use a distributed system for nigh on unintelligible to anyone who doesn’t know how to put it all back together.

Categories: Uncategorized