Cryptography Reference
In-Depth Information
may pass through a number of different people before finally being
fetched from the right server and passed back through the chain.
This process offers a high degree of anonymity, but not one that
can begin to fool an omniscient attacker watching all sites. The sim-
plicty offers a strong amount of confusion. Your machine may re-
ceive a request from Alice, but there's no way to know if Alice is ac-
tually interested in the information itself. Her machine may just be
passing along the request from someone else who might be passing
it along from someone else etc. Each individual in the chain can
only know that someone out there is interested in the information,
but they can't be certain who that person is.
10.4.4 Freenet
One of the most ambitious and successful anonymous publication
systems is Freenet, a peer-to-peer network originally designed by Ian
Clarke. The project itself is highly evolved and open source distri-
butions of the code are available from freenet.sourceforge.net .
[CSWH00, Cla99]
The systemdistributes information across a random collection of
servers donating their spare diskspace to people seeking to publish
documents. The network has no central server that might be com-
promised so all searches for information fan out across the network.
Each machine remembers a certain amount about previous searches
so it can answer requests for popular documents.
Each document is known within the network by three different
keys which are really 160-bit numbers created by applying the SHA
hash function. If you want to retrieve a document, you ask for it with
oneofthethreekeynumbers.Thesearchprocessissomewhatran-
dom and unorganized, but also resistant to damage to the network.
Here are the steps:
1.Youstartasearchbyspecifyingthekeyvalueanda“timeto
live” number which limits the depth of the nodes you want to
search.
2. You choose one node in the network to begin the search.
3. This node checks to see if the key matches any files stored lo-
cally. If there's a match, the node returns the file.
4. If there's no local match, the node checks a cache of recent
searches. If the key is found there, the node retrieves the docu-
ment. In some cases, this document is already stored locally. In
Search WWH ::




Custom Search