Verifiable and Privacy-Preserving Outsourcing of Matrix Multiplications

Matrix operation is a type of basic operations in many application scenarios. Especially, in the server-assisted applications, the user may need to delegate remote server to perform matrix operations to accomplish the target task. This paper studies the privacy, verifiability and efficiency issues in outsourcing matrix multiplications to an untrusted server. A lightweight outsourcing scheme is constructed, where the user is allowed to request matrix multiplications with privacy protection from the server and verify the correctness of the result. Performance analysis and comparison indicate that our scheme is more efficient than existing solution in each phase, which confirms the practicality of our proposal in supporting applications by resource-limited devices.