: We experimentally demonstrate that all-optical signal calculus can be realized based on dynamic Brillouin gratings in fibers. Temporal integration and first-order differentiation were performed for optical pulse with various waveforms. Development of all-optical temporal integrator and differentiator in photonic signal processing circuit has received a great attention in the optical community, since they are fundamental functions required for all-optical computing systems and networks. In this paper, we experimentally demonstrate a new type of all-optical differentiator and integrator based on dynamic Brillouin gratings (DBGs) in fibers. The proposed operators are conceptually similar to previously developed systems using uniform long period gratings and weak fiber Bragg gratings [1,2], since DBGs can considered as a weak fiber Bragg grating [3]. However, this type of function has a crucial advantage: the grating length can be flexibly tuned so as to properly adopt a wide-range signal bandwidth. Moreover, arbitrarily long gratings are possible to integrate a long sequence of signal pulses.DBG-based optical signal integration Optical signal integration in our system is actually based on two different physical processes. First, an acoustic wave was uniformly created in a 42 cm-long polarization maintaining fibers (PMFs), resulting from the stimulated Brillouin scattering (SBS) interaction between two continuous wave pumps counter-propagating through the PMF along one principle polarization axis. Then the acoustic wave acts as a grating reflector with weak reflectivity. This grating can reflect a light at distinct optical frequency to the pump frequencies, but orthogonally polarized with respect to the pumps. This optically generated grating is denominated DBG. Second, a signal pulse spectrally centered at DBG was launched into the PMF, and then the distributed pulse reflection corresponds to the temporal integration of the input signal. Time integrations of two different signal waveformsa 800 ps single pulse and 300 ps double pulses, spaced by 900 ps -were measured, as shown in Figure 1(a) and 1(b). The experimental results show a good agreement with numerically calculated integration of the waveforms.DBG-based optical signal differentiation Time differentiation is realized using a slightly modified configuration of the coherent light storage technique [4]. A 7 ns optical pulse was launched into a 20 m-long PMF and its waveform was stored along the fiber using a 800 ps writing pulse, in the form of an acoustic wave as a result of the SBS interaction. Double 800 ps pulses with a time interval of 800 ps are used as a reading pulse, but the two pulses are phase shifted in amplitude. Each pulse is then reflected by the DBG, resulting in two identical signal waveforms with a time delay of 800 ps, but with opposite sign. The delayed destructive interference between the two waveforms eventually represents the first-order derivative of the signal waveform, as shown in Figure 1(c). References[1] R.

We experimentally demonstrate that structural slow light can enhance Beer-Lambert absorption. A 4-fold reduction of the group velocity induced by mere cavity effects has caused an increase of molecular absorption by 130%.

We propose a simple but powerful scheme for the complete analysis of the frequency chirp of a gainswitched optical pulse using a fringe-resolved interferometric two-photon absorption autocorrelator. A frequency chirp imposed on the gain-switched pulse from a laser diode was retrieved from both the intensity autocorrelation trace and the envelope of the second-harmonic interference fringe pattern. To verify the accuracy of the proposed phase retrieval method, we have performed an optical pulse compression experiment by using dispersion-compensating fibers with different lengths. We have obtained close agreement by less than a 1% error between the compressed pulse widths and numerically calculated pulse widths.

A couple of experiments are here presented to clarify the impact of slow light on light-matter interaction. The experiments are designed, so that the process generating slow light and the probed light-matter interaction only present a marginal cross-effect. The impact of slow light on simple molecular absorption could be separately evaluated under either material or structural slow light propagation in the same medium and led to an entirely different response.

We experimentally demonstrate that temporal differentiation of optical pulses can be realized in a slow & fast light system based on a resonance. The waveform of a 13 ns Gaussian pulse was experimentally first-order differentiated.OCIS codes: (060.4370) Nonlinear optics, fibers; (290.5900) Scattering, stimulated Brillouin; (070.6020) Signal processing Development of basic mathematical operations in all-optical signal processing circuits -such as differentiators, integrators and logic gates -is attracting an increasing interest in the optical community, since they are key functions required for all-optical computing and networks. All-optical temporal differentiators and integrators have been experimentally realized in optical fibers, based on a single uniform long-period fiber grating (LPG) and a phaseshifted fiber Bragg grating. In this paper, we propose a new technique that can provide a first-order derivative of an arbitrary optical waveform, using a resonance-based slow & fast light system. Sharp spectral resonances can generate a strong dispersion leading to a substantial change of the effective group velocity at the center of the resonance. Neglecting the higher order dispersion terms, the spectral distribution of the pulse exiting from such a system E out (ω) is well approximated by [1]:where e G is the net signal power gain, ω o is the central frequency of the resonance, Γ B is the resonance FWHM width and E in (ω) is the pulse input spectrum. According to the Fourier transform properties, it can be simply anticipated that the output pulse will be temporal shifted by an amount of G/Γ B and the effective delay can be tuned by varying G, which is essentially the principle of resonance-based slow & fast light. For small frequency detunings ω−ω o , the phase shifts satisfies the condition G/Γ B (ω−ω o ) ‹‹ 1 and the transfer function can be further simplified as:It is interesting to point out that the second term in the transfer function is the required spectral response for a firstorder differentiator and the output signal in the time domain can be simply expressed by this sum.The output signal is formed by the sum of the input signal plus its derivative weighted by the gain factor. For large gains the derivative term can be strongly dominant and leads to a simple visual interpretation of the delay/advancement effect and of the signal temporal shapes, as illustrated by the experimental traces shown in Fig. 1 for a fast light system. The sign of the derivative is simply reversed for a slow light system. + + + Fig. 1. Time traces of an optical pulse exiting from a Brillouin fast light system for different pump powers (black line represents the input pulse). The insert depicts the sum given by Equ.3 to interpret the experimental traces squared (thus rectified) by the power-dependent photodetection.