1134 JOURNAL OF LIGHTWAVE TECHNOLOGY, VOL. 30, NO. 8, APRIL 15, 2012 Reduction in the Number of Averages Required in BOTDA Sensors Using Wavelet Denoising Techniques Mohsen Amiri Farahani, Student Member, IEEE, Michael T. V.Wylie, Student Member, IEEE, Eduardo Castillo-Guerra, Senior Member, IEEE, and Bruce G. Colpitts, Senior Member, IEEE (Invited Paper) Abstract—This paper reports on a new mechanism to decrease the number of averages and, consequently, the measurement time of Brillouin optical time-domain analysis (BOTDA) sensors using wavelet shrinkage techniques. Two different wavelet shrinkage techniques, VisuShrink and SureShrink, are applied to denoise signals acquired from measurements in BOTDA sensors. The conventional method to denoise signals in BOTDA sensors is ensemble averaging. Ensemble averaging is a time consuming technique, as it requires many acquisitions of signals to provide an acceptable SNR. To reduce the number of acquisitions, the setup of the BOTDA sensor is modified to denoise acquired signals using VisuShrink or SureShrink before applying ensemble averaging. Experimental results show a significant reduction in the number of averages required to provide an accurate measurement, and consequently, a substantial saving in the measurement time of the sensor. It has been shown that the combination of ensemble av- eraging with VisuShrink or SureShrink reduces the measurement time of the sensor up to 90%. This reduction in the measurement time enables the implementation of dynamic and fast measure- ments with BODTA sensors and opens opportunities to target a new range of applications. Index Terms—Brillouin optical time-domain analysis (BOTDA), ensemble averaging, measurement time, number of averages, SureShrink, VisuShrink, wavelet shrinkage. I. INTRODUCTION I N the last few decades, distributed optical fiber sensors, particularly those based on stimulated Brillouin scattering (SBS), have attracted a significant amount of research due to their competitive advantage of enabling continuous measure- ment of strain and temperature over long distances and their ca- pacity of working in hazardous environments [1]–[3]. There are several configurations for SBS-based optical fiber sensors such as Brillouin optical time-domain analysis Manuscript received May 31, 2011; revised August 15, 2011; accepted September 09, 2011. Date of publication September 19, 2011; date of current version March 16, 2012. This work was supported in part by the National Science and Engineering Research Council of Canada, the New Brunswick Innovation Foundation, and Springboard Atlantic Inc. The authors are with the Department of Electrical and Computer Engineering, University of New Brunswick, Fredericton, NB E3B 5A3, Canada(e-mail: z88b5@unb.ca; michael.wylie@unb.ca; ecastill@unb.ca; colpitts@unb.ca). Color versions of one or more of the figures in this paper are available online at http://ieeexplore.ieee.org. Digital Object Identifier 10.1109/JLT.2011.2168599 (BOTDA) and Brillouin optical frequency-domain analysis (BOFDA) [2], [3]. In the BOTDA approach, a pulsed beam and a counterpropagating continuous wave (CW) beam interact at different frequencies through the intercession of an acoustic wave in the fiber. In essence, the power from the higher fre- quency beam is transferred to the lower frequency one when the frequency difference between them is within the local Brillouin gain spectrum of the fiber. The frequency showing the maximum interaction in the spectrum is called the Brillouin frequency shift (BFS). The BFS depends linearly on the local strain and temperature conditions of the fiber, which allows the distributed sensing of these two parameters. The connection among the BFS, strain, and temperature is given by (1) where is the temperature coefficient in MHz/ C, is the changes in the temperature in C, is the reference Brillouin frequency in MHz, is the strain coefficient in MHz/ , and is the strain in [4]. The BFS at all points along a fiber is found yielding the dis- tributed temperature and strain along the fiber. Based on the theory, the BFS is located at the central frequency of an ideal spectrum that can be modeled by a Lorentzian curve in the fre- quency domain as (2) Three parameters are required to describe the spectrum: the BFS , the bandwidth , and the peak gain [4]. The central frequency is calculated by finding the frequency of the maximum in the ideal spectrum, but this calculation is not straightforward in real measurements. In effect, spectra are noisy and do not have a perfect Lorentzian distribution; therefore, the noise and distortion should be eliminated before fitting an ideal Lorentzian curve to the spectra to estimate the BFS. The accuracy of the results achieved by the curve fitting is directly related to the level of noise in the spectra. Studies on BOTDA sensors show that there are multiple sources of noise that are difficult to eliminate, especially noise in the light source, optical transmission, and electrical signal acquisition systems [5]. The noise in optical fiber sensors 0733-8724/$26.00 © 2011 IEEE