IEEE PHOTONICS TECHNOLOGY LETTERS, VOL. 25, NO. 22, NOVEMBER 15, 2013 2185 16-Channel Fiber Laser Sensing System Based on Phase Generated Carrier Algorithm Gaosheng Fang, Tuanwei Xu, and Fang Li Abstract— A 16-channel fiber laser sensing system based on four wavelength and four space division multiplexing and phase generated carrier (PGC) algorithm is proposed. Through intro- ducing dc block and normalization technology, the influence of light intensity and fringe visibility is decreased. A point-by- point calculation is adopted during PGC processing and we realized 16 channels real-time synchronous demodulation with high performance and stability based on PC and LabView. The wavelength shift resolution of the system is 6 × 10−7pm/√Hz, the linearity is >99.9%, and the dynamic range is 115 dB at 10 Hz and 98 dB at 100 Hz. For multichannel demodulation, the amplitude consistency is >96% and crosstalk less than −60 dB has been obtained; the correlation coefficient between the real signal and demodulated signal is >98%. Index Terms— Fiber laser sensing, PGC, real-time demodulation. I. INTRODUCTION F IBER lasers have been developed for over 20 years [1], avariety of sensing applications have been discussed, such as fiber-optic hydrophones, accelerometers and so on [2]. Due to high sensitivity, immunity to electromagnetic interference, resistance to high temperature and high pressure, easy to multiplex, the fiber laser sensing technology and it’s networks are widely used in energy exploration, safety monitoring and seismic wave detection [3], [4]. It has increasingly become the primary technology in geophysical exploration technology [5]. In 2011, a 16-element fiber laser sensing system based on NI PXI platform was reported with the demodulation by phase generated carrier (PGC) [6]. Among the various demodulation schemes adopted in opti- cal fiber sensor system, the PGC technology is widely used for its advantages of high linearity, large dynamic range and high stability [7]. During traditional PGC processing, the system should carry out a great amount of calculations for each frame of data, in order to overcome the difficulties with large data processing, a point-by-point calculation was introduced in this letter during PGC processing. Also, a DC Manuscript received May 1, 2013; revised August 28, 2013; accepted September 12, 2013. Date of publication September 17, 2013; date of current version October 23, 2013. This work was supported in part by the National Natural Science Foundation of China under Grants 61107071, 61077059, and 41074128, in part by the 863 program of China under Grant 2013AA09A413, and in part by the key Instrument Developing Project of CAS under Grant CZDYZ2012-1-108. The authors are with the Optoelectronics System Laboratory, Institute of Semiconductors, Chinese Academy of Sciences, Beijing 100083, China (e-mail:
[email protected];
[email protected]; lifang@ semi.ac.cn). Color versions of one or more of the figures in this letter are available online at http://ieeexplore.ieee.org. Digital Object Identifier 10.1109/LPT.2013.2282400 block and normalization technology was adopted to decrease the influence of light intensity disturbance. In this letter, we demonstrate a real-time demodulation distributed feedback fiber laser (DFB FL) sensing system using four wavelengths and four space division multiplexing technology and PGC interrogation scheme [8]. The key point of the system to realize space division multiplexing is the design of the multichannel interferometer, which is the use of a piezoelectric transducer (PZT) wounding with multi-fiber to form a multichannel unbalanced Michelson interferometer and it is the first time to do this according to our best knowledge. Compared with traditional ways to realize real-time and high performance demodulation for sensing networks, who usually used high speed DSP, FPGA [9] or high cost platform (NI PXI) [6], we realized a 16-channels fiber sensing system with real- time synchronous demodulation with high stability and per- formance based on personal computer (PC) and NI LabView. The cost is reduced and R&D cycle is shortened. The system has obtained a wavelength shift resolution of 6*10−7 pm/ √ Hz, whose noise level for fiber sensing network(16 elements and above) is lower than reported [6]. The linearity is larger than 99.9%, the dynamic range is 115dB@10Hz and 98dB@100Hz with the sampling rate of 20KS/s and carrier frequency of 1 kHz. For multichannel demodulation, the amplitude con- sistency of the system is larger than 96% and the crosstalk is less than −60dB, the correlation coefficient between the real signal and demodulated signal is larger than 98%. The system performance has relationship with the specifications of the components in the system, tests indicate that our system has the ability to improve the multiplexing capability and demodulation performance with increasing the PGC carrier frequency, improving sampling rate, increasing wavelength multiplexing, etc. II. SYSTEM SETUP AND OPERATION PRINCIPLE The scheme diagram of the proposed fiber laser-based sensing array is shown in Fig. 1. It can be divided into five parts: pump lights module, DFB FL sensors array, unbalanced Michelson interferometer (MI) module, photoelectric detector module, and data sampling and processing module. Pump lights module consists of four 980nm pump sources and four 980nm/1550nm wavelength division multiplexing (WDM), the output power of four pump sources is 380mW, the fiber optic cable between the WDM and DFB FL sensors can reach several kilometers. Unbalanced MI module is made up of a PZT wounding with multi-fiber to form an interferometer 1041-1135 © 2013 IEEE 2186 IEEE PHOTONICS TECHNOLOGY LETTERS, VOL. 25, NO. 22, NOVEMBER 15, 2013 Fig. 1. The diagram of 16-channels DFB FL sensing system based on PGC algorithm and NI LabView on PC. Fig. 2. The structure of the multichannel interferometer. TABLE I SPECIFICATIONS OF DWDM IN THE SENSING SYSTEM with four inputs and four outputs, which is shown in Fig. 2. The isolators used in this part to reduce the reverse injection influence of the reflected light from the Faraday rotation mir- rors to the DFB FL. Photoelectric detector module consists of 4 DWDM and 16 photoelectric detectors with low-noise PIN photodiodes and amplified circuits, whose saturated output voltage is ±10V. The parameters of the DWDM are shown in Table I, which determine the central wavelengths of the DFB FLs in the DFB FL sensors array. Data sampling and processing module is made up of a PC and a 16-bits AD data acquisition card with 24 synchronous channels, whose maximum sampling rate is 240KS/s. The PGC algorithm is carried out on PC based on NI LabView, which is discussed in Part A in details. The DFB FL sensors array consists of 16 DFB FLs, whose central wavelengths are within the range of the DWDM channels and the DFB FL optical spectra is shown Fig. 3. The amplifying medium of the DFB FL is Er3+-doped fiber whose length is 4 cm. Fig. 3. The output optical spectrum of the DFB FLs with the pump power equals 100mW (The central wavelengths of the fiber lasers are 1530.37nm, 1535.06nm, 1539.77nm and 1544.48nm). Fig. 4. The diagram of PGC demodulation algorithm. A. Interferometric Wavelength Shift Modulation and Digital PGC Demodulation Technique Traditional wavelength demodulation method, such as scan- ning FP filters, match FNGs, and micro optical spectrum analyzer, usually has a resolution on the order of 1 pm. In order to detect ultra-weak signal, an unbalanced Michelson interferometer (MI) was adopted to transform the DFB fiber laser sensor wavelength shift into MI phase shift. Because a PGC carrier was introduced to modulate the PZT on one arm of MI, the output signal of the MI is I = I0(1 + κ cos(C cosωt + ϕs(t)+ ϕ0)) (1) Where I0 is the light intensity signal of MI output, κ is the fringe visibility, and ϕ0 is the initial phase, ϕs(t) is the wavelength shift converted phase shift, as ϕs(t) = 2πnl λ2B �λB (2) Where λB is the lasing wavelength, and �λB is the wave- length shift, l is the path difference (round trip) of MI, and n is the effective refractive index of the fiber core. Digital PGC scheme was adopted in the system to execute the phase demodulation. The diagram of PGC demodulation algorithm is shown in Fig. 4. The signal is mixed with the carrier of fundamental frequency and double frequency cosωt and cos 2ωt , then low pass filtered (LPF), a pair of signals was produced as follows −κ I0 J1(C) cosϕs(t) −κ I0 J2(C) sin ϕs(t) (3) FANG et al.: 16-CHANNEL FIBER LASER SENSING SYSTEM 2187 After differential cross-multiplication subtract, integral and high pass filtered (HPF), the signal output is κ2 I 20 J1(C)J2(C)ϕs(t) (4) Because the final demodulated result is related to κ2 I 20 , the recovered signal will change with the fluctuation of the light and external disturbance. By introducing peak-to-peak normalization technology in PGC algorithm [10], the effect of κ and I0 are eliminated. The DC block removes I0 in equation (1). The basic principle of the normalization is shown as signalnormalizat ion = signalorigin − meanvalue Maxvalue − Minvalue (5) Where signalnormalizat ion is the result of the normalization of signalorigin , Minvalue and Maxvalue is the minimum value and maximum value of the sampled data. Then the demodulation result can be expressed as J1(C)J2(C)ϕs(t) (6) Equation (6) contains Bessel term J1(C)J2(C), the carrier’s amplitude C = 2.37rad is chosen to minimize the shift of J1(C)J2(C) when C changes slightly. B. The Realization of PGC With Real-Time Demodulation Based on PC and NI LabView NI LabView is a graphical programming language with integrated libraries which can run on PC. In order to real- ize PGC technique with real-time demodulation on PC, two key techniques have been adopted in the program. Firstly, the program is developed according to data stream, the 16 channels’ sampling data is parallel executed to make sure that the 16 channels are demodulated synchronous. Secondly, a point-by-point calculation is introduced during differential cross multiplying (DCM) processing of PGC algorithm, which is conductive to real-time processing and display. III. EXPERIMENTS AND RESULTS The measurement range of the signal frequency of the demodulation system is 5∼400Hz. A large amplitude (C value) phase modulation carrier at a frequency (1 kHz in this letter) outside the signal band is introduced by stretching one arm of the MI using a PZT. The output signal of MI then is received by low-noise PIN photodiodes and amplified circuits, the signals in each channel are sampled using a 16-bit resolu- tion ADC with 24 synchronous channels, and then processed through PGC algorithm based on PC and NI LabView. In order to realize real-time demodulation and display, point-by- point calculation was adopted during DCM processing of PGC algorithm. Fig. 5 shows an example of demodulation signal in this system with the test signal at 100 Hz in channel 1. Fig. 6 shows the linearity of the demodulation system with the test signal at 10Hz. We have obtained R2 larger than 99.9%. Fig. 7 shows the noise level of the demodulation system and the noise level is equal to 6*10−7 pm/√Hz, which is tested under a quite environment. Along with the measurement of the Fig. 5. An example of demodulation signal with test signal at 100Hz in channel 1. Fig. 6. The linearity of the demodulation system. Fig. 7. The noise level of the demodulation system. maximum wavelength shift, the dynamic range was calculated and shown in Fig. 8. The dynamic range is 115dB@10Hz and 98dB@100Hz. For a multichannel demodulation system, crosstalk and amplitude consistency are key specifications. A 100 mV@ 100 Hz sinusoidal test signal was applied to the interferometer to modulate the PZT, the results of every channel are shown in 2188 IEEE PHOTONICS TECHNOLOGY LETTERS, VOL. 25, NO. 22, NOVEMBER 15, 2013 Fig. 8. The dynamic range of the demodulation system. Fig. 9. The amplitude consistency of the demodulation system. Fig. 9, and the amplitude consistency is larger than 96%. By introducing a stable signal at 100Hz to a DFB fiber laser sensor and other sensors are isolated from acoustic and vibration, we measured crosstalk between each channel, and the result shows that the crosstalk is less than −60dB. The crosstalk has relationship with the channel isolation of DWDM, the interference between fiber lasers, the crosstalk between each channel of the A/D, we can choose high quality devices to decrease the crosstalk. The correlation coefficient between the real signal and demodulated signal is also measured and it is larger than 98%. IV. CONCLUSION In this letter, a 16 channels DFB FL sensors system based on four wavelengths and four space division multiplexing technology is proposed. The system has realized synchronous real-time demodulation based on PC and NI LabView. We believe that this is the first time to realize fiber-optic sens- ing networks with real-time demodulation and high stability and performance based on PC. By this mean, the cost of the demodulation system can reduce a lot and R&D cycle can be shortened. The system has a high wavelength shift resolution of 6*10−7 pm/ √ Hz. The linearity is larger than 99.9%, the dynamic range is 115dB@10Hz and 98dB@100Hz with the sampling rate equals 20KS/s and carrier frequency equals 1 kHz. For multichannel demodulation, the amplitude consistency of the system is larger than 96% and the crosstalk is less than −60dB. The correlation coefficient between the real signal and demodulated signal is larger than 98%. ACKNOWLEDGMENT The authors would like to thank Y.H. Liu for the assistance on the providing of DFB fiber lasers, and thank G.Q. Wang for his helpful assistance during the system building. REFERENCES [1] G. A. Ball and W. H. Glenn, “Design of a single-mode linear-cavity erbium fiber laser utilizing Bragg reflectors,” J. Lightw. Technol., vol. 10, no. 10, pp. 1338–1343, Oct. 1992. [2] Y. Liu, W. Zhang, T. Xu, J. He, F. Zhang, and F. Li, “Fiber laser sensing system and its applications,” Photon. Sensors, vol. 1, no. 1, pp. 43–53, 2011. [3] K. T. V. Grattan and T. Sun, “Fiber optic sensor technology: An overview,” Sens. Actuators A, Phys., vol. 82, pp. 40–61, May 2000. [4] A. D. Kersey, “A review of recent developments in fiber optic sensor technology,” Opt. Fiber Technol., vol. 2, no. 3, pp. 291–317, 1996. [5] Z. Y. He, Q. W. Liu, and T. Tokunaga, “Development of nano-strain- resolution fiber optic quasi-static strain sensors for geophysical appli- cations,” Proc. SPIE, 22nd Int. Conf. Opt. Fiber Sensors, vol. 8421, no. 84210T, Oct. 4, 2012. [6] W. Lin, H. Jun, L. Fang, and L. Yu-Liang, “Realization of 16-channel digital PGC demodulator for fiber laser sensor array,” J. Phys., Conf. Ser., vol. 276, no. 1, pp. 012134-1–012134-6, 2011. [7] L. Yang, W. Liwei, T. Changdong, Z. Min, and L. Yanbiao, “Analysis and optimization of the PGC method in all digital demodulation system,” J. Lightw. Technol., vol. 26, no. 18, pp. 3225–3233, Sep. 15, 2008. [8] J. He, L. Wang, F. Li, and Y. L. Liu, “An ameliorated phase generated carrier demodulation algorithm with low harmonic distortion and high stability,” J. Lightw. Technol., vol. 28, no. 22, pp. 3258–3265, Nov. 15, 2010. [9] L. Feng, J. He, J. Y. Duan, F. Li, and Y. L. Liu, “Implementation of phase generated carrier technique for FBG laser sensor multiplexed system based on compact RIO,” in Proc. 1st Apos Conf., 2008, pp. 54–57. [10] J. He, F. Li, W. Zhang, L. Wang, T. Xu, and Y. Liu, “High performance wavelength demodulator for DFB fiber laser sensor using novel PGC algorithm and reference compensation method,” Proc. SPIE, 21st Int. Conf. Optical Fiber Sensors, vol. 7753, no. 775333, May 17, 2011. /ColorImageDict > /JPEG2000ColorACSImageDict > /JPEG2000ColorImageDict > /AntiAliasGrayImages false /CropGrayImages true /GrayImageMinResolution 150 /GrayImageMinResolutionPolicy /OK /DownsampleGrayImages true /GrayImageDownsampleType /Bicubic /GrayImageResolution 600 /GrayImageDepth -1 /GrayImageMinDownsampleDepth 2 /GrayImageDownsampleThreshold 1.50000 /EncodeGrayImages true /GrayImageFilter /DCTEncode /AutoFilterGrayImages false /GrayImageAutoFilterStrategy /JPEG /GrayACSImageDict > /GrayImageDict > /JPEG2000GrayACSImageDict > /JPEG2000GrayImageDict > /AntiAliasMonoImages false /CropMonoImages true /MonoImageMinResolution 400 /MonoImageMinResolutionPolicy /OK /DownsampleMonoImages true /MonoImageDownsampleType /Bicubic /MonoImageResolution 1200 /MonoImageDepth -1 /MonoImageDownsampleThreshold 1.50000 /EncodeMonoImages true /MonoImageFilter /CCITTFaxEncode /MonoImageDict > /AllowPSXObjects false /CheckCompliance [ /None ] /PDFX1aCheck false /PDFX3Check false /PDFXCompliantPDFOnly false /PDFXNoTrimBoxError true /PDFXTrimBoxToMediaBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXSetBleedBoxToMediaBox true /PDFXBleedBoxToTrimBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ] /PDFXOutputIntentProfile (None) /PDFXOutputConditionIdentifier () /PDFXOutputCondition () /PDFXRegistryName () /PDFXTrapped /False /Description > >> setdistillerparams > setpagedevice