Long-Range and High Spatial Resolution Brillouin Time Domain Sensor Using Oversampling Coding and Deconvolution Algorithm

Abstract

Brillouin optical time domain reflectometer is widely used due to the advantage of single-ended input, but facing the trade-off between sensing distance, spatial resolution, and measurement time in long range application. Harnessing the co-propagating Raman amplification, we introduce Golay code under oversampling conditions to enhance the performance of the sensor. The relationship between the coding gain and the oversampling rate is investigated to determine the optimal sampling rate. At the sampling rate of 250 MSa/s, long distance experiments are conducted over a 100 km sensing fiber with 64-bit return to zero Golay coding using 100 ns optical pulses, yielding a BFS uncertainty of about 1 MHz along the fiber and complete rising edge of 20 m. At the same time, the total variation regularization-based deconvolution algorithm is proposed to resolve the deterioration of spatial resolution due to direct decoding. The robustness of the algorithm over different signal-to-noise ratio, temperature gradients, and sampling rates is investigated through simulations and experiments. Finally, the spatial resolution of the experimental results is improved to be higher than 6 m, which is one third of the original result. These methods greatly facilitate optical pulse coding in long range and high-resolution sensing.