In onion routing, messages are repeatedly encrypted and then sent through several network nodes called onion routers. Each onion router removes a layer of encryption to uncover routing instructions, and sends the message to the next router where this is repeated.

Are there any (theoretical or practical) vulnerabilities in onion routing? Are there any known ways to break the anonymity that it offers?

To be specific, I am asking if the process of onion routing can be hacked.

3 Answers 3


A significant threat to onion routing is a correlation attack performed by an adversary that is able to observe the start and end nodes in your connection.

If somebody is watching you or your entry node, and the website you are visiting or your exit node - they can correlate the traffic. They see the website sending a large javascript file, and a few seconds later you receive a similarly sized piece of encrypted data. They see the website serving a large image, and then you receive a large image. The website isn't sending anything, and you aren't receiving anything.

This is even easier when they are simply trying to confirm that it is you visiting a website (known as traffic confirmation), and not looking to see which of every website in the world you are visiting (looking for a needle in a haystack, which is possible, but more difficult). And in both cases — the more traffic, the greater confidence they have that they got it right.

Another way to perform the traffic confirmation attack is when you are able to receive arbitrary data from an attacker. This is the case if you run a Hidden Service, or if you are signed in on a chat program — the attacker sends you a large (several megabyte message) and see if you receive a large encrypted message shortly thereafter.


The simplest (relative term) method of breaking onion routing is a correlation attack performed by a global adversary that controls a majority of nodes in the network.

What that means is that somebody watching your entry and exit node can correlate requests coming in from your computer to their entry node at time n and leaving their exit node at time n1 for repeated communications over time to say with a fair degree of confidence that the traffic coming form your IP is going to that server.


Doesn't the first server see who I am?

Possibly. A bad first of three servers can see encrypted Tor traffic coming from your computer. It still doesn't know who you are and what you are doing over Tor. It merely sees "This IP address is using Tor". Tor is not illegal anywhere in the world, so using Tor by itself is fine. You are still protected from this node figuring out who you are and where you are going on the Internet

From FAQ.

In the light of the last events: Exclusive: Secret contract tied NSA and security industry pioneer SAN FRANCISCO Fri Dec 20, 2013 5:07pm EST

Backdoored Dual_EC_DRBG. crypto.stackexchange.com: Explaining weakness of Dual EC DRBG to wider audience?

However, this backdoor was detected in November 15, 2007 Bruce Schneier's blog: The Strange Story of Dual_EC_DRBG

Imagine, that NSA pay to Tor's dev team 10,000,000$ + 1$ to inject kind of backdoor into their asymmetric crypto.

Thereafter, they have no needs to stay as a Tor-node, even. Simply, record every connection through internet on every IX - An Internet exchange point. Pub-keys they can get from any Tor-Dir, and every names and IPs of Tor-nodes too.

Your onion, thrice encrypted traffic, has not yet reached a first server, it will be cracked on the IX, your connection in that case is a plain-text connection. NSA could see your surfing as Your <-> Server. Tor doesn't matter.

However, FAQ of Tor say the same that RSA team, Ron Rivest, Adi Shamir and Leonard Adleman told us about their crypto, before.

  • if all three servers on the way would be NSA it would of course also break TOR, right? Commented Mar 29, 2015 at 17:40

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .