Organizer: Prof. Thomas Schneider
Modern web applications are complex entities amalgamating different languages, components, and The concrete efficiency of secure computation has been the focus of many recent works. In this work, we present concretely-efficient protocols for secure 3-party computation (3PC) over a ring of integers modulo 2^l tolerating one corruption, both with semi-honest and malicious security. Owing to the fact that computation over ring emulates computation over the real-world system architectures, secure computation over ring has gained momentum of late.
Cast in the offline-online paradigm, our constructions present the most efficient online phase in concrete terms. In the semi-honest setting, our protocol requires communication of 2 ring elements per multiplication gate during the online phase, attaining a per-party cost of less than one element. This is achieved for the first time in the regime of 3PC. In the malicious setting, our first result requires communication of 4 elements per multiplication gate during the online phase, beating the state-of-the-art protocol by 5 elements. In a follow-up work, we could reduce this cost to 3 ring elements. Realized with both the security notions of selective abort and fairness, the malicious protocol with fairness involves slightly more communication than its counterpart with abort security for the output gates alone.
We apply our techniques from 3PC in the regime of secure server-aided machine-learning (ML) inference for a range of prediction functions-- linear regression, linear SVM regression, logistic regression, and linear SVM classification. In the follow-up work, we could extend the application to training of functions -- linear regression, logistic regression, Neural Network (NN) and Convolutional Neural Network (CNN). Our setting considers a model-owner with trained model parameters and a client with a query, with the latter willing to learn the prediction of her query based on the model parameters of the former. The inputs and computation are outsourced to a set of three non-colluding servers. Our constructions catering to both semi-honest and the malicious world, invariably perform better than the existing constructions.
Ajith is a Ph.D. research scholar in the CrIS lab at Indian Institute of Science, Bangalore, India. His work orbits around Secure Multi-party Computation (MPC), more specifically in the area of MPC for small number of parties with applications to Privacy Preserving Machine Learning (PPML). He is a recipient of Google PhD fellowship for the year 2019.
Department of Computer Science and Automation, IISc