arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2404.16232cs.CRcs.DC
收藏

SECO:跨多服务器层级的模型分割安全推理

SECO: Secure Inference With Model Splitting Across Multi-Server Hierarchy

Shuangyi Chen, Ashish Khisti

AI总结:

针对预测即服务中的隐私问题,提出安全推理协议SECO,通过多服务器层级分割神经网络模型,结合多方同态加密与混淆电路,在不泄露数据与模型隐私的前提下协作完成推理,并降低用户端的计算与通信开销。

摘要:

在预测即服务的语境下,关于数据和模型隐私的担忧已被提出,并通过安全推理协议加以解决。这些协议由在多种不同安全假设下设计的单一或多种密码学工具构建而成。\n在本文中,我们引入了SECO,一种安全推理协议,使得持有输入数据向量的用户能够与部署了分割神经网络模型的多个服务器节点协作计算预测结果,同时不损害任何一方的数据隐私。我们扩展了先前要求整个神经网络模型位于单一服务器节点上的安全推理工作,提出了一种多服务器层级结构,其中用户与网关服务器节点通信,网关服务器节点再与远程服务器节点通信。推理任务被分割到各个服务器节点上,并且必须在数据向量的加密副本上执行。\n我们采用多方同态加密和多方混淆电路方案,使系统能够抵御半诚实服务器的恶意多数攻击,同时保护用户免受部分模型结构泄露的影响。我们在多个模型上评估了SECO,实现了用户计算和通信成本的降低,使得该协议适用于资源有限的用户设备。

英文摘要:

In the context of prediction-as-a-service, concerns about the privacy of the data and the model have been brought up and tackled via secure inference protocols. These protocols are built up by using single or multiple cryptographic tools designed under a variety of different security assumptions. In this paper, we introduce SECO, a secure inference protocol that enables a user holding an input data vector and multiple server nodes deployed with a split neural network model to collaboratively compute the prediction, without compromising either party's data privacy. We extend prior work on secure inference that requires the entire neural network model to be located on a single server node, to a multi-server hierarchy, where the user communicates to a gateway server node, which in turn communicates to remote server nodes. The inference task is split across the server nodes and must be performed over an encrypted copy of the data vector. We adopt multiparty homomorphic encryption and multiparty garbled circuit schemes, making the system secure against dishonest majority of semi-honest servers as well as protecting the partial model structure from the user. We evaluate SECO on multiple models, achieving the reduction of computation and communication cost for the user, making the protocol applicable to user's devices with limited resources.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑