Entanglement’s Benefit Survives an Entanglement-Breaking Channel

Entanglement is essential to many quantum information applications, but it is easily destroyed by quantum decoherence arising from interaction with the environment. We report the first experimental demonstration of an entanglement-based protocol that is resilient to loss and noise which destroy entanglement. Specifically, despite channel noise 8.3 dB beyond the threshold for entanglement breaking, eavesdropping-immune communication is achieved between Alice and Bob when an entangled source is used, but no such immunity is obtainable when their source is classical. The results prove that entanglement can be utilized beneficially in lossy and noisy situations, i.e., in practical scenarios.

Figure 1

Experimental setup. SPDC: spontaneous parametric down-converter; DM: dichroic mirror; C: collimator; CWDM: coarse wavelength-division multiplexer; BS: beam splitter; Attn: attenuator; EDFA: erbium-doped fiber amplifier; DL: delay line; PC: polarization controller; PM: phase modulator; AAG: adjustable air gap; Pol: polarizer; DCF: dispersion-compensating fiber; DSF: dispersion-shifted fiber; TEC: thermoelectric cooler; OPA: optical parametric amplifier; D: detector.

Figure 2

${\mathrm{BER}}_A$ and ${\mathrm{BER}}_E$ versus source brightness $N_S$ for $500\mathrm{kbit}/\mathrm{s}$ communication. Inset: 25 bits of OPA-receiver detector output (blue) and Bob’s corresponding modulation waveform (red). See text for more information.

Figure 3

Alice’s Shannon information, an upper bound on Eve’s Holevo information, and a lower bound on Alice’s information advantage, versus Alice’s source brightness $N_S$. The top panel computes Alice’s Shannon information using the error probabilities from the solid blue curve in Fig. 2. The bottom panel computes Alice’s Shannon information when Eve takes 90% of the Bob-to-Alice light instead of the 10% she receives in the top panel.