(Doc. ID 65774) The influence of the slow-light effect on the performance of distributed Brillouin sensors is studied. We show that, while in most situations it can be neglected, it may greatly affect the results obtained for certain experimental configurations. More specifically, for one of the experimental arrangements described in the literature (a strong continuous-wave pump and a weak pulsed probe) we show that this effect induces a large time biasing of the traces that depends not only on the fiber length but also on the frequency separation between pump and probe. This biasing reduces the available resolution in this experimental arrangement. Distributed fiber-optic sensors offer unique capabilities for the monitoring of quantities such as temperature and strain over long distances. In many applications, these sensors avoid the need for thousands of pinpoint sensors and complicated multiplexing schemes. Among these, distributed Brillouin sensors have attracted much research interest in the past years 1-7 and are now widely used for the monitoring of strain and temperature distribution within large structures in civil engineering.2 Several techniques have been proposed for performing distributed Brillouin sensing, and the principal configurations have been summarized in Fig. 1, together with the typical parameters used in experimental setups selected from representative publications. In the Brillouin optical time-domain reflectometer (BOTDR) configuration, a pump pulse is launched into the fiber (ϳ200 ns duration, ϳ150 mW of peak power), and the spontaneous Brillouin backscattered light is synchronously analyzed as a function of the time (distance along the fiber) by using heterodyne detection.
3For each position the pump-Stokes frequency shift is determined, which is then translated into strain or temperature values.In the Brillouin optical time-domain analysis 4 (BOTDA-1) configuration, however, the Brillouin interaction is performed in the stimulated regime, thus requiring two counterpropagating waves, a powerful pulsed pump wave (ϳ50 ns with ϳ70 mW of peak power) and a weak probe wave ͑ϳ100 W͒. If particular phase-matching conditions are met (namely, f pump Ϸ f probe + B , B being the Brillouin shift), an acoustic wave is generated that scatters photons from the pump to the probe wave. This leads to a local amplification of the probe wave, which yields a time-dependent variation of the detected probe signal in the pump end. An earlier version of this technique 5 (labeled BOTDA-2 in Fig. 1) uses a powerful, continuous-wave (cw) pump wave ͑ϳ6 mW͒ and a weak pulsed probe wave (ϳ50 ns pulse length with ϳ1 mW of peak power) tuned at f probe Ϸ f pump + B . The position-dependent Brillouin attenuation of the pump induced by the probe is recorded in terms of a variation in the cw pump power detected from the probe end. Schemes similar to BOTDA-1 and BOTDA-2 can be developed in the frequency domain. 6 The conclusions of the analysis of the timedomain configuration can be exactly translated into the frequency...