Advanced detection of information in optical pulses with negative group velocity

In this Rapid Communication, we experimentally demonstrate that the signal velocity, defined as the earliest time when a signal is detected above the realistic noise floor, may be altered by a region of anomalous dispersion. We encode information in the spatial degree of freedom of an optical pulse so that the imprinted information is not limited by the frequency bandwidth of the region of anomalous dispersion. We then show that the combination of superluminal pulse propagation and realistic detectors with nonideal quantum efficiency leads to a speed-up of the earliest experimentally obtainable arrival time of the transmitted signal even with the overall pulse experiencing unity gain. This speed-up is reliant upon nonideal detectors and losses, as perfect detection efficiency would result in the speed of information being equal to the speed of light in vacuum, regardless of the group velocity of the optical pulses.

Figure 1

Experimental scheme for the arrival-time measurement of spatial information in optical pulses. Depicted are the two situations for acquiring the reference pulse, and when the fast-light medium is switched on. Probe pulses are shaped with a mask and injected into the rubidium cell at an angle of $\approx$$1^{\circ }$ relative to the pump beam. A fast gated camera is used to detect and temporally resolve the pulses. The gating time of the camera intensifier can be lowered to 2.44 ns and the gate delay is then swept temporally to acquire images during the arrival of the pulse. Images of the reference pulse are taken without the pump beam present.

Figure 2

(a) A single dark stripe on a bright background is used as a test mask. Integration over the full pulse cross section shows a pulse peak advancement of 90 ns. (b) Arrival time for the spatially integrated intensity of the pulses for the reference pulse without 4WM (dashed blue line) and the superluminal pulse (solid red line). (c) The stripe visibility $M=\frac{I_{\text{max}}-I_{\text{min}}}{I_{\text{max}}+I_{\text{min}}}$ of the transmitted pulses. A cross section is taken through the image along the vertical axis to analyze the double-peak structure in the spatial mode. (d) The SNR for the stripe visibility defined in the text for the advanced and the reference pulse as a function of time. In the inset, we show the leading edge of the pulses where the SNR crosses the threshold of 1.

Figure 3

Integrated stripe visibility signal. (a) The integration $\int \text{SNR}\left(t\right)dt$ performed for the experimental data shown in Fig. 3d. (b) A model based on a phase-insensitive amplifier with time-dependent gain $G\left(t\right)$, as discussed in the text, is shown.