Secure Delegated Computation is a two or multi-party cryptographic functionality where a client (or multiple clients) securely delegates the computation of a function to one or more external servers or computing parties. These servers are often more powerful but untrusted. The functionality ensures privacy and correctness, and often verifiability (check [verifiable page]), while allowing the client(s) to compute less than what is needed to evaluate the function locally.
This abstract functionality can occur in different scenarios, including:
The main goal of this functionality is to enable the computation of highly complex functions through a cloud platform, using powerful servers, while maintaining the privacy of the data and/or the algorithms of the user from the server.
This functionality is closely related to other important multi-party cryptographic functionality, such as homomorphic encryption and multi-party computation.
The protocols that implement this functionality are:
Secure delegated computation is a functionality that exists both classically and quantumly. The most well-known classical counterparts are:
The main properties include:
Privacy: The client’s input (and possibly the function) remains hidden from the server(s). This is achieved via encryption (FHE), secret sharing (MPC), or blindness (quantum).
Correctness: Having confidence that the server performed the computation correctly.
Blindness (Quantum Setting)[5]: In quantum protocols, blindness ensures that the delegated computation, including quantum gates and inputs, remains hidden from the quantum server.
There also exist additional properties such as:
Universality: If the protocol satisfies the above properties for any computation, or only a specific subset of functions. If the earlier, the protocol is called universal.
Efficiency or Computational cost: Client offloads work without needing to simulate or recompute. Efficiency is often quantified by the cost of the computation on the server as well as the overhead cost that is required for maintaining privacy. The efficiency also highly depends on the assumption, and can often be reduced if certain trust assumptions have not been considered.
Communication cost: One of the relevant resources for delegated protocols, especially in the quantum setting, is the amount of communication, particularly since some of the quantum protocols require quantum communication.
Verifiability: If a delegated computation protocol allows the client to check whether the computation has been performed correctly, then the protocol has verifiability. However, not all delegated and private multiparty computations satisfy verifiability, specifically in the quantum setting.
No content has been added to this section, yet!
Secure Delegated Computation is a two or multi-party cryptographic functionality where a client (or multiple clients) securely delegates the computation of a function to one or more external servers or computing parties. These servers are often more powerful but untrusted. The functionality ensures privacy and correctness, and often verifiability (check [verifiable page]), while allowing the client(s) to compute less than what is needed to evaluate the function locally.
This abstract functionality can occur in different scenarios, including:
The main goal of this functionality is to enable the computation of highly complex functions through a cloud platform, using powerful servers, while maintaining the privacy of the data and/or the algorithms of the user from the server.
This functionality is closely related to other important multi-party cryptographic functionality, such as homomorphic encryption and multi-party computation.
No protocols implement this functionality yet.
Secure delegated computation is a functionality that exists both classically and quantumly. The most well-known classical counterparts are:
The main properties include:
Privacy: The client’s input (and possibly the function) remains hidden from the server(s). This is achieved via encryption (FHE), secret sharing (MPC), or blindness (quantum).
Correctness: Having confidence that the server performed the computation correctly.
Blindness (Quantum Setting)[5]: In quantum protocols, blindness ensures that the delegated computation, including quantum gates and inputs, remains hidden from the quantum server.
There also exist additional properties such as:
Universality: If the protocol satisfies the above properties for any computation, or only a specific subset of functions. If the earlier, the protocol is called universal.
Efficiency or Computational cost: Client offloads work without needing to simulate or recompute. Efficiency is often quantified by the cost of the computation on the server as well as the overhead cost that is required for maintaining privacy. The efficiency also highly depends on the assumption, and can often be reduced if certain trust assumptions have not been considered.
Communication cost: One of the relevant resources for delegated protocols, especially in the quantum setting, is the amount of communication, particularly since some of the quantum protocols require quantum communication.
Verifiability: If a delegated computation protocol allows the client to check whether the computation has been performed correctly, then the protocol has verifiability. However, not all delegated and private multiparty computations satisfy verifiability, specifically in the quantum setting.
No content has been added to this section, yet!