1

I’m exploring a data security concept where files do not exist as conventional static entities but instead “float” as encrypted fragments, retrievable only through specific contextual parameters.

The Idea Floating Storage: The file is split into multiple small fragments, all given uniform filenames (e.g., datafile.dfv).

Contextual Binding: Access to these fragments is only possible when verified against behavioral or environmental parameters (e.g., device signature, time-bound tokens).

Dynamic Mapping: A mapping system continuously changes the internal references between the fragments and the file structure, making stolen fragments useless without the correct context.

This idea was inspired by how map pins on Google Maps are meaningless until the system overlays them with contextual data. Similarly, stolen fragments would be useless without the AI or mapping layer.

My Questions Is there any known implementation in production systems where data exists only as context-bound fragments?

Would traditional key-chunk mapping (used in filesystems) break this concept by making fragments predictable or clonable?

Could HSM/TPM-backed keys help maintain the floating nature by constantly regenerating internal references?

Example Pseudocode:

# Simplified representation of fragment hashing logic
import hashlib, time

def generate_fragment_key(fragment_data, device_id, time_slot):
    raw = fragment_data + device_id + str(time_slot)
    return hashlib.sha256(raw.encode()).hexdigest()

# Time-slot ensures key changes dynamically
key = generate_fragment_key("fragment_1", "DEVICE_ABC123", int(time.time() // 3600))
print(key)

This pseudocode only shows the dynamic key regeneration idea for each fragment.

12
  • 1
    I'm not sure if I understand the concept: either one has to continuiously re-encrypt the fragments with new keys (which requires decryption with the old key first) or the encryption keys for each fragment needs to be static - in which case I don't see any "dynamic mapping". Still, the question stands what problem you are trying to solve which is not solved already by simpler methods (like simple TPM/HSM backed encryption). Commented Jul 23 at 18:24
  • @Steffen Ullrich Thanks for clarifying. I agree—re-encrypting fragments repeatedly would essentially act like static keys or periodic re-encryption. My “dynamic mapping” idea doesn’t rotate encryption keys but adds a separate mapping layer that shifts over time or per access session. The key stays stable, but fragment references change, so a stolen snapshot can’t easily reconstruct the file. I know TPM/HSM is the standard—I’m just exploring whether this extra “floating mapping layer” adds value against attackers with only periodic snapshot access.
    – Cero
    Commented Jul 23 at 18:29
  • What problem are you trying to solve?
    – vidarlo
    Commented Jul 23 at 18:32
  • @ vidarlo The goal is to make stolen or leaked snapshots harder to reconstruct into a full usable file. Even if fragments are copied, shifting mapping/references aim to add an extra layer of complexity, making partial periodic snapshots less valuable to attackers.
    – Cero
    Commented Jul 23 at 18:34
  • 1
    While you have a lot of ideas, this is too vague and metaphorical for a concrete answer. I'm also not sure how throwing in “AI” is supposed to help. You should write down specific security goals (what needs to be protected from whom) and learn about standard solutions like full-disk encryption and encrypted file containers. It's quite possible that if you remove all the fancy metaphors, you end up with a relatively simple problem that has a known solution.
    – Ja1024
    Commented Jul 24 at 3:14

0

You must log in to answer this question.

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.