Homomorphic Encryption for Private Information Retrieval
Homomorphic encryption is a type of cryptography that allows computations to be performed on encrypted data. For instance, such a scheme can be used in a cloud environment to query encrypted data without decrypting that data, avoiding privacy issues that plague online setting.
A fully homomorphic encryption (FHE) supports arbitrary computation on encrypted data. This offers a very powerful capability, however at present, it is computationally too intense for practical use. Partially homomorphic encryption only allows some operations to be preserved in ciphertexts, such as additions, multiplications, or quadratic functions. These schemes are becoming practical with today’s computing resources.
Homomorphic encryption is powerful tool that can be leveraged in many application. One such application is Private Information Retrieval (PIR). Private Information retrieval is a protocol whereby a user can retrieve an item from a database without revealing which item is retrieved. Under the guise of homomorphic encryption, one can search encrypted data without ever decrypting. PIR has been in the making for many year and now it is on the horizon of practicality!
Computers govern our world. Whether browsing at a clothing store, putting a deposit into a bank account, or even ordering food, computers are instrumental in each and every single activity that we do. However, the limits of traditional computing have been reached. Moore’s law is predicted to be dead by the year 2021, while algorithms and processes continue to become even more complex by each passing day. A replacement, however, is already in the works. Quantum computing is the future of computers.
Traditional computers rely on transistors that represent the smallest piece of information, a bit, which can be a zero or a one. In contrast, quantum computers represent information in qubits which can be multiple states simultaneously due to the rules of quantum physics. In quantum physics, subatomic particles can act like waves. However, they can also take on being a particle or a wave, or a particle and a wave in a process known as superposition. As these states scale, exponential computing power is unleashed.
The development of actual quantum computers is still in its infancy. However, experiments have been carried out in which fully working quantum computational operations were executed on a very small number of qubits. This allows quantum computers with a modest number of qubits to rapidly solve problems that would take today’s most powerful computers with millions of bits years to solve like integer factorization using Shor’s algorithm or the simulation of quantum many-body systems. Large-scale functional quantum computers may exist in little as 10 years. The advent of quantum computing would be able to solve certain types of complex problems in machine learning, medicine, and material science that are simply intractable for classical computers today. At the same time, quantum computing will pose a devastating threat to today’s security.