Recent advances in quantum computers challenge the security of key agreement protocol that are designed with the intractability assumptions of discrete logarithmic problem and integer factorization problem. Hence, motivated to develop the key agreement protocol that is secure in post-quantum era, recently Islam proposed a provably secure two-party authenticated key agreement (2PAKA) protocol. The protocol is based on the intractability assumption of the famous Ring Learning With Error (RLWE) problem. In this work, we show that Islam’s two-party key agreement is vulnerable to the modified version of the signal leakage attack (SLA) which is termed as improved-signal leakage attack (i-SLA) in this article. Using i-SLA, the attacker can successfully recover the long term private key of the honest user by instantiating utmost $q$ number of key exchange sessions with the honest user using $q$ number of malformed public keys. To overcome the attack, we provide a countermeasure without changing the original design of the protocol.

量子计算机的最新进展挑战了密钥协商协议的安全性，这些协议是根据离散对数问题和整数分解问题的难处理性假设设计的。因此，为了开发后量子时代安全的密钥协商协议，最近伊斯兰提出了一种可证明安全的两方认证密钥协商 (2PAKA) 协议。该协议基于著名的环形错误学习 (RLWE) 问题的难处理性假设。在这项工作中，我们表明伊斯兰的两方密钥协议容易受到信号泄漏攻击 (SLA) 的修改版本的影响，本文将其称为改进信号泄漏攻击 (i-SLA)。使用i-SLA，攻击者可以通过实例化最大程度成功恢复诚实用户的长期私钥$q$ 与诚实用户使用的密钥交换会话数 $q$格式错误的公钥的数量。为了克服这种攻击，我们在不改变协议原始设计的情况下提供了一种对策。