Oblivious Transfer (OT) enables a sender to transmit multiple messages such that a receiver can retrieve selected messages without revealing their choices to the sender. However, existing OT extension (OTE) protocols primarily focus on 1-out-of-2 OTE, 1-out-of-n OTE and 1-out-of- \(\infty \) OTE, which transmit only one selected message from amount messages. The number of selected messages is limited by these protocols and cannot be revised once the protocols are determined. These protocols are inefficient for applications requiring batch processing or dynamic selection of multiple messages. In this paper, a flexible adaptive selection number OTE protocol is proposed, inspired by Kolesnikov et al. in CCS 2016. Unlike traditional OTE protocols, our protocol enables the receiver to flexibly select k messages from the n messages transmitted by the sender within a single OT instance. The value of k can be dynamically adjusted at runtime, making the protocol well-suited for applications with frequently changing selection requirements. This flexibility significantly improves the efficiency and versatility of oblivious transfer. In contrast to the fastest state-of-the-art OTE algorithm, where implementing m k-out-of-n OTE instances incurs the cost of implementing km standard 1-out-of-2 OTs, our approach only requires the cost of m standard 1-out-of-2 OTs. For computational security parameter k, the proposed k-out-of-n OTE offers O(k) factor performance improvement in communication, compared to KKRT16. We further explore the application of our protocol to semi-honest secure OTE. Notably, our OTE can overcome the restriction of single selection in the existing OTE protocols. We implemented our OTE protocol and found that it is 6.4–7.2 times faster than KKRT16 for 8-bit string OTEs and small-scale OTE instances. Specifically, our protocol takes only 13.7 s to securely compute \(2^{14}\) OTs. For larger-scale OTE computations, our protocol is also 4.1 times faster than KKRT16.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Adaptive Batched K-out-of-N Oblivious Transfers Extension

  • Zhaoyi Liu,
  • Huijie Yang,
  • Jian Shen,
  • Jianfei Sun

摘要

Oblivious Transfer (OT) enables a sender to transmit multiple messages such that a receiver can retrieve selected messages without revealing their choices to the sender. However, existing OT extension (OTE) protocols primarily focus on 1-out-of-2 OTE, 1-out-of-n OTE and 1-out-of- \(\infty \) OTE, which transmit only one selected message from amount messages. The number of selected messages is limited by these protocols and cannot be revised once the protocols are determined. These protocols are inefficient for applications requiring batch processing or dynamic selection of multiple messages. In this paper, a flexible adaptive selection number OTE protocol is proposed, inspired by Kolesnikov et al. in CCS 2016. Unlike traditional OTE protocols, our protocol enables the receiver to flexibly select k messages from the n messages transmitted by the sender within a single OT instance. The value of k can be dynamically adjusted at runtime, making the protocol well-suited for applications with frequently changing selection requirements. This flexibility significantly improves the efficiency and versatility of oblivious transfer. In contrast to the fastest state-of-the-art OTE algorithm, where implementing m k-out-of-n OTE instances incurs the cost of implementing km standard 1-out-of-2 OTs, our approach only requires the cost of m standard 1-out-of-2 OTs. For computational security parameter k, the proposed k-out-of-n OTE offers O(k) factor performance improvement in communication, compared to KKRT16. We further explore the application of our protocol to semi-honest secure OTE. Notably, our OTE can overcome the restriction of single selection in the existing OTE protocols. We implemented our OTE protocol and found that it is 6.4–7.2 times faster than KKRT16 for 8-bit string OTEs and small-scale OTE instances. Specifically, our protocol takes only 13.7 s to securely compute \(2^{14}\) OTs. For larger-scale OTE computations, our protocol is also 4.1 times faster than KKRT16.