Post
Topic
Board Announcements (Altcoins)
Re: [ANN][DRK] DarkCoin | First Anonymous Coin | First X11 | First DGW | DarkSend+ Is Live!
by
AlexGR
on 17/08/2014, 19:15:47 UTC

Perhaps I am still misstating the math or misunderstanding the spread-sheet.
Let's say I have 1 different coin sources. 60/600 compromised nodes I choose 2 hops and make 1 transaction. Then according to the spreadsheet, that one transaction has 1% chance of being compromised.
Now lets say I have 100 different coin sources, one for each of 100 transactions. 60/600 compromized nodes at a 2-hop rate. Now each transaction itself has a 1% chance of compromise. thus in total, 63% that at least one of them is compromized.
An example: let's say I have 100 dice with each having 100 sides. So each dice has 1/100 chance of rolling a "7". Now if I roll all 100, then there is 63% chance that at least one out of the 100 gets a "7", regardless of the fact that they are different dice. It's just acumulating the independent probabilities of the unrelated events
What is being traced? You have 1% chance of traced back to the one address. But that original address has no context. The coins for that address has itself been mixed.

My original question was why are there different numbers of masternodes being chosen, in other words, 1, 2, etc.. rounds of mixing. the purpose of this question was I was wondering the speed of transaction involved. The answer was that "The  several rounds are just to eliminate the possibility of a large number of masternodes  colluding with each other to deanon transactions." and a spreadsheet was given with the probability of de-anoning happening. So as I understand it, the spreadsheet is telling you the chance of a de-anon'd transaction. Is this incorrect? My main point here is why do I have to do many hops if it takes a long time? If I only have to do 1 hops, why is there an option for 2 hops? why not just always do 1 hops?

It's not the speed of transaction you are concerned about, but the speed of the pre-mixing. Up to RC3, transaction and mixing was in the same stage. Therefore if mixing was delayed, the transaction was delayed.

In RC4 you premix the coins at some point of time and this process might take minutes or hours - it depends on the number of other people who are also mixing. Essentially you are readying the coins for when you want to make transactions.

Now... When you transact, the transaction speed is like Litecoin (2.5m confirmation block), whether you transact non-anonymized funds, or anonymized funds.

Regarding the rounds of mixing / hops.

In a 60 sockpuppet nodes / 600 total nodes scenario, this is how I see it:

one round of mixing means you have 10% risk of deanonymization.
two rounds means 10% x 10% = 1%
three rounds mean 0.1%
four means 0.01%
five means 0.001%
six means 0.0001%
seven means 0.00001%
eight means 0.000001%

So if you do 5000 transactions after 8 round mixing, the possibility that just one of these transactions might be deanonymized by the holder of the 60 nodes is 0.000001% x 5000 = 0.005%

If you are more paranoid about your security, you can probably do 8 rounds x 8 rounds of mixing for a total of 16 rounds Grin