MIT boffins identify Tor hidden services with 88 per cent accuracy

Boffins from the Massachusetts Institute of Technology (MIT) have demonstrated a vulnerability in Tor which, if exploited, could lead to hidden services being identified with up to 88 per cent accuracy. Infosec bods from MIT and the Qatar Computing Research Institute (QCRI) pwned the anonymity network for a paper to be …

[phil dude]

surprised...

I would have thought making traffic of uniform entropy would be key to defeating density analysis attacks.

At least, that is what I got from the article...?

P.

[Anonymous Coward]

Re: surprised...

Yes, but occasionally the padding bytes compromise the encryption algorithms. You must be careful when doi

ng that sort of thing. . . Null bytes look weird!

--

[phil dude]

Re: surprised...

That was my point. Normally(!) data is padded internally so that encrypted packets are indistinguishable based on content.

Anyone else knows if this is the weakness?

P.

[CommanderGalaxian]

Re: surprised...

One of the "design constraints" for TOR is that it can be used for real-time web browsing. Once you start adding in padding, it makes things less responsive.

[sisk]

Time for something new. This is the third time in the last few years I've heard of a way to compromise TOR's security. Then again I don't use TOR so what do I know?

[Valeyard]

A burglar can get through my front door given enough time, Doesn't mean I'm going to be defeatist and just leave it all unlocked though

[Lars]

"This is the third time in the last few years I've heard of a way to compromise TOR's security." One could, of course, also read that as an attempt, by the powers who hate TOR, to discredit TOR as much as they can. I don't use TOR either, perhaps one should even if for no other reason than increasing the traffic assuming it will make it harder for spooks.

[Binnacle]

the BIG IF -- headlines, lies and statistics

"That means that an adversary who lucked into the position of guard for a computer hosting a hidden service"

Right, so you have to become a HS's guard node and you can correlate it's traffic.

This is not news--correlation attacks are the always-known weakness of all low-latency anonymity routing systems. So it's 88% multiplied by what? 1%? 5%? 10% or maybe 0%? Good HS operators are aware of this issue and take steps to establish/protect/rotate safe entry guards.

But is is news and is interesting that they have found a way to improve the system.

[Anonymous Coward]

No surprise at all

Anyone who thinks they can hide online is a fool.

[Lars]

Re: No surprise at all

So should we give up or try harder.

[Anonymous Coward]

Re: No surprise at all

Try harder.

[Anonymous Coward]

Re: No surprise at all

You can only do so much. Trying to keep oneself anonymous necessarily requires inefficiency because trying to be as efficient as possible sculpts your habits into unique, identifiable shapes as cruft is cut away. You see this with Freenet. To help disguise your tracks, your activity is padded and mixed in with other people's activity and bounced around a few times. The end result is that it takes a long time to receive anything and the effective data rate's the pits.