A Lockdown Technique to Prevent Machine Learning on PUFs for Lightweight Authentication
A Lockdown Technique to Prevent Machine Learning on PUFs for Lightweight Authentication
复制标题
DOI：
10.1109/tmscs.2016.2553027
复制
发表时间：
2016-07-01
期刊：
影响因子：
--
通讯作者：
Verbauwhede, Ingrid
中科院分区：
文献类型：
--
作者：
Yu, Meng-Day (Mandel);Hiller, Matthias;Verbauwhede, Ingrid
我们提出了一种基于PUF的轻量级身份验证方法，该方法在服务器对设备进行身份验证的设置中以及在设备生命周期内身份验证次数有限的用例中非常实用。我们的方案使用服务器管理的挑战/响应对（CRP）锁定协议：与以前的方法不同，具有机器学习能力的自适应选择挑战对手在没有服务器隐式许可的情况下无法获得新的CRP。对手面临的问题是，用有限的机器学习训练数据导出PUF模型。我们的系统级方法允许将所谓的强PUF用于轻量级身份验证，其方式是通过最坏情况下的CRP暴露算法验证来对抗当今最好的机器学习方法。我们还提出了一个退化的实例，使用一个弱的PUF，是安全的计算不受限制的对手，其中包括任何学习对手，为实际的设备寿命和读出率。我们使用硅PUF数据验证了我们的方法，并证明了支持10、1，000和1 M身份验证的可行性，包括无法通过多项式资源学习的实际配置，例如，CRP的数量和攻击运行时间，使用最近的结果的基础上的可能近似正确（PAC）的复杂性理论框架。
We present a lightweight PUF-based authentication approach that is practical in settings where a server authenticates a device, and for use cases where the number of authentications is limited over a device's lifetime. Our scheme uses a server-managed challenge/ response pair (CRP) lockdown protocol: unlike prior approaches, an adaptive chosen-challenge adversary with machine learning capabilities cannot obtain new CRPs without the server's implicit permission. The adversary is faced with the problem of deriving a PUF model with a limited amount of machine learning training data. Our system-level approach allows a so-called strong PUF to be used for lightweight authentication in a manner that is heuristically secure against today's best machine learning methods through a worst-case CRP exposure algorithmic validation. We also present a degenerate instantiation using a weak PUF that is secure against computationally unrestricted adversaries, which includes any learning adversary, for practical device lifetimes and read-out rates. We validate our approach using silicon PUF data, and demonstrate the feasibility of supporting 10, 1,000, and 1M authentications, including practical configurations that are not learnable with polynomial resources, e.g., the number of CRPs and the attack runtime, using recent results based on the probably-approximately-correct (PAC) complexity-theoretic framework.