Train-accident-like post-quantum IBE

Train-accident-like Grandfather is me, don’t understand deep mathematical details, believe certain things by reference to so-called sir-math and construct new applications assuming that sir-math has already proved the assumptions within the appropriate limitations, which can now be explored by them for my work.

We start in lattice post-quantum cryptography. It is assumed that right non-IBE single user field has been set.

KEM: Critical first part

Given two encryption systems, one S2 an IBE, another S1 not, both exponentially strong, one can build an IBE s3 similar to S2. In S3

Encry3 (s) = Encry2 (Encrp1 (s))

Decry3 (s) = Decry2 (Decrp1 (s))

Exponential-ness is important because trivial enumeration attacks are exponential.

Why would one build IBE this way? Certain weaknesses, e.g. INDCCA’s of the IBE can be eliminated this way, provided IBE1 is free! For example, my/cocks AAIBE is based on discrete square-roots of huge integers, well known to have INDCCA problems. One effective simple fix is s2 = square-root-IBE, while s1 = OAEP+ RSA. That in turn can be viewed as this structure, with a public symmetric encryption prior to RSA! The joint encryption is free of defects.

post-quantum non-IBE

Modulo fancy math limitations, let us define operations of +, - between vectors elementwise, modulo a global number. * is defined by polynomial multiplications, restoring the n-size by subtraction the higher powers.

Now let us call vector BIG if all elements are big (capitalized), small if all are small (lowercase). Basic rule is R = As + e is easy if A,s,e are given but post-quantum-crypt-hard to recover s if A, R are given. We call s the secret and e the error. Without telling anyone, pick own s and e and publish some A and R as this equation. Post-quantum-crypt-hard to get s. Depends on no one!

Encryption and decryption

To send Message, encrypt normal by any nice method which can be inverted by receiver knowing Key. Let the Key be encoded somehow in K. Now compute and publish X = Aa+b and Y = Ra+c +K for any small vector a, b, c.

Decryption is easy – Compute K(receiver) = Y – Xs. One can show that this K = K + junk.

Here is a simple way to recover K.  Sender only considers top w bits for encoding K. Receiver only keeps the top w bits. Just 1 bit over several is needed. Typical vector sizes are power of 2 like 1024.

There is the question of small. Let us assume a 1024 bit key and byte polynomials with hiding in the highest bit. One can remove smallness error  with reconcialation, or slightly larger (6%) key with no reconciliation in newhope-simple.

Every one publishes polynomial coefficients for the two public variables This is in add to the family semiprime and full name. My hardware extracts these from user computer, verifies them, encrypts and decrypts. The encryption to be send and decryption are published to the user computer. My hardware has a simple usb interface.

Every thing can be files and streams. Several Gb/sec can be handled allowing quantum computer proof sportcasts, telephonic & videophonic conversations, newscasts etc.

post-quantum IBE

Here S2 = post-quantum non-IBE while S1 = AA/cocks/OAEP+ IBE. Doing it so will import the self-break-able but foreign-unbreakable properties of AAIBE with corruptible police! Apart from other things my crypt-libertarianism survives quantum computers!

Crypt-libertarianism

Very few can point to something that survives 1000 years, let alone million or billion. But crypt-libertarian ism is different, it will survive the new civilizations next million years. That is the ultimate snobbery that finally trumps the USA Boston-Harver shrink-pope

Q Do you think yourself like Jesus Christ?
A (me) 1. Now why would I compare self to uncivilized shepherd so far back in middle east!
2. You MUST ascribe me more ego than that!

What is it anyway?

Libertarian ism without a head, built from soul of bitcoin, only assumption is that a numerical majority is never evil.