Can a 1930-40's cryptography analyst (or similar profession) from the Allies, Axis or any country at that time, break or decrypt a digital signal that had been encrypted with AES?

by YukkuriOniisan

I don't know if I should ask this to a Historian or to a Computer savvy people, however, I think a military Historian may be more familiar with historical Signal Intelligence than normal people. Please excuse my curiosity.

After reading about Allies and Axis signal intelligence work in World War 2, I am intrigued whether or not any of the signal analysts at that time could break through our modern encryption method. Since as far as my understanding, any encrypted message at that time was transmitted as an analog signal, I am also wondering whether or not the analysts could interpret a digital signal. I had tried to read and search for the answer myself, but alas, I found none.

So, to frame my question, assume that during the 1930's or 1940's a continuous AES-encrypted digital signal from the future leaked through the past. Could the contemporary signal analysts at that time decrypt the data using contemporary technology or science?

If this question is not appropriate for this subreddit, may I ask for a help to point which subreddit will be the better?

Thank you for spending the time to accompany my curiosity. Cheers.

Platypuskeeper

Signals were not always analogue. Wire signals were sent with teletype using digital Baudot coding (a 5-bit character set), including in encrypted form using for instance the German T52 Geheimschreiber. Although the radio transmissions used were analogue, the Enigma cipher itself is digital; one discrete letter becomes another; it's not one continuously-variable signal being transformed into another. In present day terms Enigma the T52 and similar are what's known as stream ciphers (as opposed to AES which is a block cipher)

The question is odd; There's no known attack on AES in the present day, and no reason to assume WWII cryptographers know anything today's don't. (rather the opposite; I think it's safe to say that in terms of published volume; more cryptography research is done in a single year these days than had been done in human history prior to WWII)

As for decrypting it with a known key (and knowing the method), that would certainly be doable. You could even do it with a paper and pencil if you liked, tedious as it'd be. There are no fundamental mathematical/logical operations involved that weren't known at the time. Ciphers like AES are not really that complicated; it's a design goal that they should be simple and fast to implement in computer hardware. Simplicity is also important to aid the cryptanalysis. In short, the more complicated a cipher is, the bigger the risk there's some hidden vulnerability, and a common maxim of modern cryptography (also voiced by some WWII cryptographers like Arne Beurling) is that one should not rely on obscurity and complexity for security.