FHE Fully Homomorphic Encryption

  1. What is FHE

In simple terms, while keeping the data encrypted, you can perform arbitrary calculations on the data.

F Fully: Compared with HE (homomorphic encryption, which is limited by the calculation method), FHE supports almost any calculation logic

H Homomorphic: When we say that an encryption scheme is homomorphic, it means that the ciphertext calculation and the plaintext calculation have the same result.

E stands for Encryption. Encryption is easy to understand and is one of the core concepts of cryptography.

Problem: FHE is usually computationally inefficient, since processing encrypted data is certainly not as fast as direct computation.
Solution: Hardware acceleration, solution design, and cross-use with other encryption technologies

Simple understanding

A classic example of explanation is: processing gold

  • I need to hire a worker to help me process gold, but I don't want him to steal the gold

  • So I designed a transparent closed box, added a lock, and then left two gloves free to fasten the box together.

  • The operator can process the gold with gloves, but because the box is locked and sealed, no scraps will be taken away

    Similar to this

    correspond:


    Box: Encryption Algorithm
    Lock: Key
    Golden embryo: raw data
    Processed Gold: Results of FHE Calculations
    Processing: Direct calculation and processing of encryption results


    (Reduced efficiency: Processing with gloves is definitely not as efficient as processing with bare hands)




    1. Applications of FHE


    As we know, the core of FHE is to directly process encrypted data.


    So you, the smart one, must have thought of [cloud computing]


    Users need to process data - payment - with the help of cloud servers and cloud computing - FHE, no privacy leakage


    In Web3, several directions of FHE application include:
    Restaking, data storage, AI, Depin, Layer3, games, voting, privacy transactions, MEV, quantum-resistant computing, etc.


    Some projects:

    @zama_fhe Open Source + FHE Solution


    @mindnetwork_xyz FHE encryption privacy middle layer


    @Privasea_ai FHE and machine learning, AI+Depin


    @FhenixIO FHE Layer 2


    @inconetwork Modular FHE encryption layer


    Some perspectives

    • FHE is very compatible with AI and Depin. As long as AI and Depin are still there, FHE will be able to shine.

    • Compared with ZK and MPC, FHE can be combined and more simply applied on a large scale.


    1. FHE and ZK, MPC


    First, let’s take a look at the introduction:

    • FHE: Allows computations to be performed on encrypted data without decrypting it.

    • ZK: Allows one party to prove to another that a statement is true without revealing any specific information about the statement.

    • MPC: enables multiple parties to jointly perform computations on their private data without revealing the input data to other participants.

    *There is no essential difference between the three technologies. They are used in different scenarios and can be used in combination.