Device-Independent Communication

Device-independent communication uses observed input–output correlations to certify a cryptographic resource with limited assumptions about the internal quantum devices. The research question is how to obtain a positive, composably secure output at useful rates under realistic losses and finite data. A Bell violation is necessary for many protocols, but it is not by itself a complete key-distribution result.

Required background. Device-independent QKD develops the trust boundary, trial definition and finite-key ledger.

Device independence is a precise trust model

Security still requires appropriately private input choices, authenticated classical communication, protected laboratories, valid causal separation or isolation assumptions, and trusted classical processing. No-click events cannot be discarded through an unjustified fair-sampling assumption. Heralding must be arranged consistently with the security proof and trial definition.

A schematic finite-key ledger is

$$\ell\leq H_{\min}^{\varepsilon}(A^n|E)-\mathrm{leak}_{\mathrm{EC}}-\Delta_{\mathrm{security}}.$$

The smooth min-entropy bound depends on the observed correlations and proof assumptions; the correction term depends on the chosen security parameters. This expression is an accounting structure, not a plug-in key formula.

What the experiments establish

Nadlinger and colleagues demonstrated key distribution certified using Bell’s theorem in an ion-based experiment. Liu and colleagues’ 2026 work reports a finite-size device-independent demonstration at one distance and an asymptotic positive-rate result at a longer distance. Preserve that distinction: asymptotic rate positivity does not imply a finite secret key from the collected data. 2022 experiment, 2026 experiment.

Established: bounded demonstrations under explicit security models. Active: higher accepted rates, tighter finite-size analysis, robustness to drift and memory, and practical interoperability. Conjectural: security claims that quietly omit a trust assumption or substitute a large Bell statistic for a full proof.

Open questions

How much test data should be sacrificed to estimate nonlocality? Which event-ready architecture best balances detection efficiency and trial rate? Can a proof exploit realistic device information without losing the intended independence guarantee? How should protocol aborts and temporal drift be handled without post hoc selection?

Graduate project: finite-size sensitivity analysis

Reproduce a published finite-key calculation using its exact security proof, observed parameters and failure-budget allocation. Sweep block length, Bell violation, key-basis error and test fraction. Deliver a region of positive key length and identify the dominant uncertainty. A laptop suffices for many published bounds; some optimization steps need specialist convex solvers.

The baseline is the authors’ stated operating point. Failure tests count no-clicks consistently, perturb calibration-derived inputs and replace an asymptotic formula with the finite-size one. If the selected theorem assumes independent trials, do not claim security against arbitrary memory. Research begins with a proved tighter bound, an experimentally feasible protocol change or a drift-robust analysis that retains composable security.

Entry points and dates

Start from the stable security-contract page, then reconstruct one experiment’s proof assumptions. Sources checked on 11 October 2026; selected evidence spans 2022–2026, not every current implementation or security proof.

References

• W.-Z. Liu et al., “Long-lived remote ion–ion entanglement for scalable quantum repeaters,” Nature 652, 51–57 (2026). Article.
• D. P. Nadlinger et al., “Experimental quantum key distribution certified by Bell’s theorem,” Nature 607, 682–686 (2022). Article.