System and method for data compression and encryption using asymmetric codebooks

Patent metadata

Field	Value
Publication number	US-12407364-B2
Application number	US-202318491754-A
Country	US
Kind code	B2
Filing date	Oct 21, 2023
Priority date	Oct 30, 2017
Publication date	Sep 2, 2025
Grant date	Sep 2, 2025

Abstract

A system and methods for data compression with encryption using asymmetric codebooks, that utilizes machine learning to generate asymmetric encoder and decoder codebooks for data encoding and compression. The use of asymmetric codebooks enables a variety of data obfuscation techniques such that output data remains useful to the decoder and any downstream entity, while concealing the original sourceblock data. The asymmetric codebooks may be utilized in reverse order to restore the original data as needed, facilitating a two-way reversible operation.

First claim

What is claimed is:

1. A system for data compression with encryption using asymmetric codebooks, comprising:
a computing device comprising a processor and a memory;
a codebook generator comprising a first plurality of programming instructions stored in the memory which, when operating on the processor, causes the computing device to:
receive training data and additional input data;
preprocess the additional input data to remove unnecessary data values and normalize the input for machine learning model training;
train a machine learning model using the training data and the preprocessed input data;
generate a plurality of codebooks using the trained machine learning model, wherein at least one of the generated codebooks is an asymmetric codebook configured to transform encoded data rather than reconstruct the original sourceblock data; and
provide the asymmetric codebook to a data decoder; and

the data decoder comprising a second plurality of programming instructions stored in the memory which, when operating on the processor, causes the computing device to:
receive the asymmetric codebook;
receive a codeword from a data encoder, the codeword comprising encoded data based on a sourceblock that was encoded by the data encoder using a codebook differing from the asymmetric codebook; and
decode the codeword using the asymmetric codebook to produce a derived data chunk based on the sourceblock, wherein the derived data chunk contains data that is transformed from the original sourceblock according to the asymmetric codebook.

2. The system of claim 1 , wherein the derived data comprises transformed data based on the sourceblock and the asymmetric codebook.

3. The system of claim 1 , wherein the derived data comprises encrypted output based on the sourceblock and the asymmetric codebook.

4. The system of claim 3 , wherein the data decoder is further configured to:
apply the asymmetric codebook in reverse to the encrypted output, to reproduce the received codeword; and
send the reproduced codeword to the data encoder.

5. A method for data compression with encryption using asymmetric codebooks, comprising the steps of:
receiving, at a codebook generator, training data and additional input data;
preprocessing the additional input data to remove unnecessary data values and normalize the input for machine learning model training;
training a machine learning model using the training data and the preprocessed input data;
generating a plurality of codebooks using the trained machine learning model, wherein at least one of the generated codebooks is an asymmetric configured to transform encoded data rather than reconstruct the original sourceblock data;
providing the asymmetric codebook to a data decoder;
receiving, at the data decoder, the asymmetric codebook;
receiving a codeword from a data encoder, the codeword comprising encoded data based on a sourceblock that was encoded by the data encoder using a codebook differing from the asymmetric codebook; and
decoding the codeword using the asymmetric codebook to produce a derived data chunk based on the sourceblock, wherein the derived data chunk contains data that is transformed from the original sourceblock according to the asymmetric codebook.

6. The method of claim 5 , wherein the derived data comprises transformed data based on the sourceblock and the asymmetric codebook.

7. The method of claim 5 , wherein the derived data comprises encrypted output based on the sourceblock and the asymmetric codebook.

8. The method of claim 7 , further comprising the additional steps of:
applying the asymmetric codebook in reverse to the encrypted output, to reproduce the received codeword; and
sending the reproduced codeword to the data encoder.

Assignees

• Atombeam Technologies Inc

Inventors

• Cooper Joshua
• Riahi Aliasghar

Classifications

• H03M7/6005

  Decoder aspects · CPC title

• G06N20/00

  Machine learning · CPC title

• G06N3/126

  Evolutionary algorithms, e.g. genetic algorithms or genetic programming · CPC title

• H03M7/6052

  Synchronisation of encoder and decoder · CPC title

• H03M7/4062

  Coding table adaptation · CPC title