One of the new and exciting cryptographic tools is homomorphic encryption (HE) which enables arbitrary computation over ciphertext, generating an encrypted result that, when decrypted, matches the result of operations performed on the plaintext. It is a game changing technology that could soon unlock the ability to perform secure data processing in environments such as the cloud or other sensitive systems.
So, What is Homomorphic Encryption?
With conventional encryption, we encrypt information before we found it or transmit it. In order to do any calculations on the data, it has to be decrypted first. However, this process brings along serious security risks, as the sensitive data is exposed for possible attacks during the decryption process.
In contrast, homomorphic encryption enables computation on encrypted data. This means that the data can still be processed and analyzed without ever revealing the actual data, thus maintaining its confidentiality.
Different types of homomorphic encryption
Homomorphic encryption is mainly of two types:
Intensive Homomorphic encryption (PHE):
Can only support a small number of operations: addition, multiplication, etc.
Homomorphic encryption is less efficient and practical for many applications than partially homomorphic encryption.
Fully Homomorphic Encryption (FHE):
This allows an unlimited number of operations to be carried out on the encrypted data, and in theory, any computation on the data can be done.
But FHE is slow and very computationally expensive compared to normal encryption.
Uses of Homomorphic Encryption
There are many potential applications of homomorphic encryption, these include:
Cloud Data Processing without compromising your data security
Privacy-Preserving Data Analytics: Analyzing large datasets without compromising privacy of data subjects.
Privacy & Security For Electronic Voting: Protecting the integrity & privacy of electronic voting
Medical Data Analysis With Security: Evaluating medical data while maintaining patient anonymity.
Financial transactions can be processed without exposing sensitive financial information.
Obstacles and Next Steps
However, there are several challenges with homomorphic encryption that needs to be addressed [2], such as:
The cost in terms of computation: The cost that homomorphic encryption takes has limited granularity in regards to its practicality.
Noise Management: The process of encryption is not trivial, as computations over encrypted data accumulate noise and the results may be less accurate.
Work is ongoing to overcome these hurdles and to create homomorphic encryption schemes that are more efficient and make more practical sense. The remarkable thing is as the technology works on, there will be more tendency of homomorphic encryption in several enterprises and apply.
