Yet another question about some confusing wording and how the technology of the setting is canonically "supposed" to work.
On page 252 of the Core Book, under Sniffing, the rules say, "To capture the information you must succeed in an Infosec Test. If successful, you capture data traffic from any targeted devices in range. Note that sniffing does not work on encrypted traffic (including VPNs and anything else using public key cryptography) as the results are gibberish. Quantum encrypted communications cannot be sniffed."
I'm a bit confused by the statement "sniffing does not work on encrypted traffic". Does that mean that the sniffer program literally fails to function at all? or that it does capture something, but what it captures is gibberish?
Let's say you use a sniffer program to eavesdrop on a (non quantum) encrypted signal. You get gibberish. You record 15 minutes of that gibberish. Then, later, you get a copy of the key used to encrypt that transmission. Could you then go back and decode the gibberish? What if you fed the gibberish into a quantum computer?
I guess my main question is, is the "gibberish" something that can be saved and decrypted later when you have the means, or is it forever useless?
Welcome! These forums will be deactivated by the end of this year. The conversation continues in a new morph over on Discord! Please join us there for a more active conversation and the occasional opportunity to ask developers questions directly! Go to the PS+ Discord Server.
Intercepting Encrypted Signals
Thu, 2013-10-24 19:03
#1
mikegentry
Intercepting Encrypted Signals
Fri, 2013-10-25 14:34
#2
Smokeskin
You can record it and decrypt
You can record it and decrypt it later.
Mon, 2013-10-28 05:34
#3
Armoured
Quantum cryptography lets you
Quantum cryptography lets you do some very odd things to encryption, and currently has several technical issues which make it hard to implement. In Eclipse Phase, quantum encrypted messaging is presumed to be unsniffable without alerting the people using it to the fact that it is being sniffed, thus they can instantly cut the connection, or start to backtrace you.
Smokeskin is right about non-quantum encrypted traffic though, it can be stolen and later decoded. However, its also presumed that EP encryption is arbitrarily hard to break; you either have to steal the key (though physical theft, hacking, or lead-pipe interrogation of someone) or have a quantum computer brute-force it, which is expensive and long.
Mon, 2013-10-28 09:02
#4
LatwPIAT
Armoured wrote:Quantum
What the EP rules describe is a very simplified model of actual quantum cryptography; the unsniffability is a quality of quantum [i]key exchange[/i], where Alice sends Bob a One-Time-Pad using a communications channel where information that has a quantum superposition can be exchanged (in the real world, this is mostly lasers through fiberoptic cables; in EP, non-photon particles can be quantum-entangled near-indefinitely.)
However, the One-Time-Pad is entirely random, and cannot itself be used to exchange information; instead, it's used to encrypt the message, and the encrypted message is then transmitted normally from Alice to Bob. It should be possible for Eve to sniff the encrypted message, but EP doesn't model this. EP assumes that since Eve can't sniff the key exchange, Eve can't sniff the message at all. Realistically, Eve should be able to sniff the message and then steal Bob's laptop to get access to the key.
It's mostly an unanswered question, but quantum computers aren't necessarily the end-all of cryptography. RSA is known to be weak to quantum cryptography because RSA is broken with prime factorization, for which there are known, efficient quantum algorithms. There currently exists cryptographic protocols, such as lattice-based cryptography, which has no currently known quantum weaknesses.
Either way, the discovery of a quantum algorithm that can break a previously quantum-proof cryptosystem, or a cryptosystem that is impervious to all known quantum algorithms, would both be excellent MacGuffins or plot-seeds for almost any kind of campaign; Oversight, the Lunar Banks, the Triads, Firewall, and Project OZMA would all be interested in getting hold of such algorithms, and possibly, making sure no-one else gets hold of them.
—
@-rep +2
C-rep +1
Mon, 2013-10-28 09:42
#5
Armoured
LatwPIAT wrote:It should be
I did know about quantum key exchange theory, and how that is limited to exchanging a key. To make the EP canon make sense, I thought perhaps the keys are constantly updated, meaning its encryption changes and makes sniffing harder. While you could sniff the data stream, its only useful if the communicators don't securely wipe their keys once they are finished- maybe this is part of the protocol? :)
Which is why I said it is "arbitrarily hard"; its speed of plot. Yeah, cryptography is always an arms race, presumably its going to get hotter if/when quantum computing becomes available. It makes for excellent plot devices though... How much will you pay to be able to crack Oversight's servers, or get encryption a TITAN can't break?
Tue, 2013-10-29 08:17
#6
Jaberwo
What changes for the user
What changes for the user when they switch from the usual asymmetric procedure with primes and public/private keys to this lattice based thing? Is the implementation on the surface level at all similar?
Tue, 2013-10-29 10:40
#7
LatwPIAT
I don't know too much about
I don't know too much about the implementation, but as far as I can tell, the end-user won't really experience any differences; perhaps comparative performance, speed and storage issues compared to RSA, but with the excessive computation power in EP, this likely won't be an issue.
The real problem is more likely to be lack of implementation. Encryption is a two-way channel, and if lattice-based cryptography isn't implemented as a common protocol, you won't have anyone to talk to over your super-secure channel:
"My data is safe behind six layers of symmetric and public-key algorithm."
"What data is it?"
"Mostly me emailing with people about cryptography."
http://xkcd.com/1269/
—
@-rep +2
C-rep +1
Tue, 2013-10-29 14:48
#8
Smokeskin
Jaberwo wrote:What changes
From what I can tell, it will be pretty much the same. Lattice crypto however has not been proven to not be breakable by quantum computers. Someone might one day discover a quantum algorithm like Shor did for prime factorization crypto.