AI 中文总结
该研究通过扩展算子框架刻画符合相对论因果性的关联，证明量子关联的非平凡干扰会导致隐藏超光速信号，并构造RC攻击破坏设备无关密码原语的安全性，指出安全性需结合测量事件的时空位置。
AI 中文摘要
在某些时空构型中，相对论因果性允许干扰型超光速因果影响，然而这种影响并不能实现超光速信号传递。我们通过扩展PRL 104, 140404的算子框架，完整刻画了符合相对论因果性（RC）的关联。当要求 underlying 态为量子态时，我们证明任何非平凡的干扰——无论是与态无关还是与态相关的——必然会导致隐藏的超光速信号。因此，唯一一致的可能是无干扰的标准量子关联，或是完整的相对论因果关联集合，而量子关联的中间干扰扩展会导致信号产生。作为该刻画的一个应用，我们研究了设备无关（DI）密码原语在仅受相对论因果性约束的对手面前的安全性。我们构造了明确的RC攻击，这些攻击在干扰几何中会破坏DI位承诺和DI秘密共享，即便这些协议对无信号对手仍然安全。我们的结果表明，针对相对论对手的安全性要求，需同时指定行为及其测量事件的时空位置；仅输入输出统计是不够的。
英文摘要
Relativistic causality in certain spacetime configurations permits jamming - superluminal causal influences that nevertheless do not enable superluminal signaling. We fully characterize the correlations compatible with relativistic causality (RC) by extending the operator framework of PRL 104, 140404. When the underlying state is required to be quantum, we prove that any nontrivial jamming - whether state-independent or state-dependent - necessarily leads to hidden superluminal signaling. Thus, the only consistent possibilities are standard quantum correlations without jamming or the full relativistically causal correlation set, with intermediate jamming extensions of quantum correlations leading to signaling. As an application of our characterization, we investigate the security of device-independent (DI) cryptographic primitives against adversaries constrained only by relativistic causality. We construct explicit RC attacks that break DI bit commitment and DI secret sharing in jamming geometries, even when these protocols remain secure against no-signaling adversaries. Our results show that security against relativistic adversaries requires behaviors to be specified together with the spacetime locations of their measurement events; input-output statistics alone are insufficient.
Comments20 pages, 1 figure
