Secure Inference for Vertically Partitioned Data Using Multiparty Homomorphic Encryption

TL;DR

This paper introduces a secure inference protocol for vertically partitioned data using multiparty homomorphic encryption, enabling encrypted data aggregation at a server without transmitting model weights.

Contribution

It presents a novel packing scheme for MPHE that facilitates secure inference on vertically partitioned data without model weight transmission.

Findings

Efficient encrypted data aggregation via the proposed packing scheme

Secure inference achieved without sharing model weights

Applicable to distributed deep learning scenarios

Abstract

We propose a secure inference protocol for a distributed setting involving a single server node and multiple client nodes. We assume that the observed data vector is partitioned across multiple client nodes while the deep learning model is located at the server node. Each client node is required to encrypt its portion of the data vector and transmit the resulting ciphertext to the server node. The server node is required to collect the ciphertexts and perform inference in the encrypted domain. We demonstrate an application of multi-party homomorphic encryption (MPHE) to satisfy these requirements. We propose a packing scheme, that enables the server to form the ciphertext of the complete data by aggregating the ciphertext of data subsets encrypted using MPHE. While our proposed protocol builds upon prior horizontal federated training protocol~\cite{sav2020poseidon}, we focus on the inference for vertically partitioned data and avoid the transmission of (encrypted) model weights from the server node to the client nodes.