Open Access
Add to BibSonomy
Abstract
Enabling communication networks with sensing functionality has attracted significant interest lately. The digital subcarrier multiplexing (DSCM) technology is widely promoted in short-reach scenarios for its inherent flexibility of fine-tuning the spectrum. Its compatibility with large-scale as-deployed coherent architectures makes it particularly suited for cost-sensitive integrated sensing and communication applications. In this paper, we propose a scheme of spectrally integrating the digital linear frequency modulated sensing signal into DSCM signals to achieve simultaneous sensing and communication through shared transmitter. Consequently, this cost-effective scheme has been demonstrated to achieve 100-Gb/s dual-polarization quadrature phase-shift keying (DP-QPSK) and 200-Gb/s dual-polarization 16-ary quadrature amplitude modulation (DP-16QAM) transmission with a distributed acoustic sensing sensitivity of 69 ${{\textrm{p}\mathrm{\varepsilon}} / {\sqrt {\textrm{Hz}} }}$ and 88 ${{\textrm{p}\mathrm{\varepsilon}} / {\sqrt {\textrm{Hz}} }}$ respectively, at a spatial resolution of 4 m.
© 2023 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
Endowing intelligent functionalities to ubiquitous optical fiber networks has attracted significant interests lately, as it provides the potential add-value of the fiber assets [1–4]. In addition to data transmission, the intelligent networks equipped with high-performance sensing possess the capability to monitor the external environments, which facilitates the advances in public safety and the development of smarter cities [5,6]. To enable cost-effective data transmission and sensing over the same optical fiber, the integrated network architecture should be intra-compatible and simple. Furthermore, it is necessary to minimize any mutual potential impact between the communication and sensing signals, thereby enhancing the overall efficiency of the system.
On the early stage of ongoing research, it has been demonstrated that state of polarization (SOP) and optical phase, extracted from telecom transponders, can be applied for vibration sensing [7,8]. However, these techniques cannot provide the sufficient localization accuracy required in distributed sensing applications, especially for intricate environments. In contrast, distributed optical fiber sensing (DOFS) based on backscattered light enables the measurement of a wide range of physical parameters with a significantly high spatial resolution. In the realm of DOFS, phase-sensitive optical time-domain reflectometry (Φ-OTDR) with high spatial resolution and sensitivity has emerged as a particularly intriguing methodology. The remarkable similarity between Φ-OTDR and a conventional optical communication system, in the aspect of the system architecture, signal modulation and signal detection, makes them ideal for integrated optical transmission and monitoring [9]. In direct detection systems, a scheme of integrated Φ-OTDR and communication in an optical fiber (ISAC-OF) has been demonstrated using the same wavelength channel [9]. The ISAC-OF scheme extends the intelligent functionality for optical fiber communication system with enhanced transmission performance. However, it also introduces a notable complex network architecture. Additionally, distributed acoustic sensing (DAS) based on Φ-OTDR that co-exists with optical communication is proposed using mode division multiplexing over a two-mode fiber [10]. However, it is far from feasible considering the as-deployed single-mode-fiber networks. In recent years, Φ-OTDR has also been integrated into coherent optical communication networks by wavelength-division multiplexing (WDM) and frequency division multiplexing (FDM) to enable simultaneous data transmission and distributed vibration sensing [11–13]. However, these schemes for combined Φ-OTDR and communication simply share the fiber with two distinct transceivers, resulting in a significant increase in deployment cost. Recently, a novel scheme with the shared transceiver has been proposed in self-homodyne coherent (SHC) detection system [14]. However, the bidirectional SHC structures increase the system complexity and may be incompatible with the as-deployed traditional coherent architecture.
Regarding traditional coherent optical network architecture, the digital subcarrier multiplexing (DSCM) technology has gained significant attention for its potential in short-reach scenarios, deemed as a key driver for the development of next-generation software-configurable optical networks [15]. The DSCM technology operates on the principle of digitally multiplexing several low-baud-rate subcarriers, instead of a high-baud-rate single carrier. This approach offers a convenient way for spectrally multiplexed sensing and transmission through a shared transmitter. Consequently, the DSCM system, possessing the flexibility of fine-tuning the spectrum and compatibility with large-scale as-deployed coherent architectures, is particularly suited for cost-sensitive short-reach sensing and communication applications.
In this paper, we propose a scheme of integrating Φ-OTDR into DSCM system to achieve simultaneous sensing and communication. To realize high-performance sensing, continuous linear frequency modulated (LFM) probe signal is derived from p-order fractional Fourier transform (FrFT) of a direct current signal (hereinafter referred to as FrFT-DC signal), and then compressed by 1-p order FrFT to obtain a high peak to lobe ratio and narrow main lobe [16]. Compared to LFM pulses, the continuous LFM signal relieves the demand for high power sensing probe and alleviates the impact of nonlinearity on both sensing and communication [17]. Combined with the adjustable digital spectrum of DSCM system, we localize the digital LFM sensing data in the central portion of the DSCM spectrum, thus achieving the spectrally multiplexed sensing and communication in a single channel. As a result, a cost-effective system architecture is realized with the sole traditional coherent transmitter. In the experiments, a comprehensive study is conducted to investigate the impact of the communication-to-sensing power ratio (CSPR) on the system's performance. At an optimal CSPR of 13 dB, the system successfully demonstrates simultaneous sensing and communication over a 10 km fiber. It achieves a 100-Gb/s dual-polarization quadrature phase-shift keying (DP-QPSK) and 200-Gb/s 16-ary quadrature amplitude modulation (16QAM) transmission with a high DAS sensitivity of 69 ${{\textrm{p}\mathrm{\varepsilon}} / {\sqrt {\textrm{Hz}}}}$ and 88 ${{\textrm{p}\mathrm{\varepsilon}} / {\sqrt {\textrm{Hz}}}}$ respectively, at a spatial resolution of 4 m.
2. Principle
In a typical coherent Φ-OTDR system, a sensing probe signal p(t) is modulated on a ultranarrow linewidth laser and then transmitted into the sensing fiber via an optical circular. The backscattered light from the sensing fiber is detected using a coherent receiver with the local oscillator originating from the same laser in the transmitter. The coherent receiver consists of 90° optical hybrid and polarization diversity detectors. Then the baseband electrical signal is sampled and digitized by the analog-to digital converter (ADC).
The received Rayleigh backscattered signal is the superimposition of the backscattered light from different scatter centers within the fiber length corresponding to the input probe length. The scatter centers are considered as spatially discrete reflectors with various specifications due to the inhomogeneous material density and fluctuations induced by the manufacturing process. The backscattered signal for each segment of fiber length $\Delta z$ centered at position z can be expressed as:
where ${\alpha _s}$ is the attenuation coefficient of the fiber, $k = 2\pi vn/c$ is the propagation constant, with v being the optical frequency of the laser and n the effective refractive index of fiber, $\tau = 2nz/c$ is the round-trip propagation delay from the output of the circulator to the scattering center. The total backscattered signal is the integral of Eq. (1) over the entire fiber length L, which can be regarded as the impulse response of the sensing fiber:The detected sensing signal by the coherent receiver is the convolution between the sensing probe signal p(t) and the impulse response h(t):
where n(t) denotes the additive white Gaussian noise in the system. To interrogate the sensing fiber without ambiguity, the period of the sensing probe signal must be greater than the round-trip delay ${\tau _L}$, defined by the length of fiber ${\tau _L} = 2nL/c$. Generally, for the system using a LFM sensing probe signal, an estimation result of the impulse response can be obtained by a matched filter. However, the imperfect correlation between the transmitted and received LFM signal may limit the system performance. To realize high peak side-lobe ratio (PSLR) and narrow main lobe, we apply the FrFT to LFM signal generation and compression [16]. Continuous LFM signal is derived from p-order fractional FrFT of a DC signal.Considering a typical LFM signal in the following form,
where A is the amplitude, ${\varphi _0}$, k denote phase and modulated rate respectively. In practical, the LFM signal has a finite duration, and a rectangle function can be introduced asHence, the FrFT of $c(t)g(t)$ with a rotation angle $\alpha ={-} \textrm{arccot} (k)$ is expressed as [18]
Equation (7) shows that the amplitude envelope of a finite duration LFM signal in $\alpha ={-} \textrm{arccot} (k)$ order FrFT domain is a sinc signal. The full width of the main lobe is
Thus, a LFM signal can be compressed by FrFT in the similar way with the match filter. The main lobe size is smaller in the case of FrFT compression compared to the matched filter method. This reduction is attributed to the presence of a product factor $\cos \alpha$, being less than 1, from FrFT compression. Moreover, utilizing coordinates rotation by FrFT, we generate LFM signal by p order FrFT of a direct current signal, and then compress it by taking 1-p order FrFT. In theory, the operation of two-step FrFTs is equivalent to the direct FFT of a DC signal [18]. Hence a fully compressed LFM pulse with high PSLR and narrow main lobe can be promised.
For a backscattered light reflected at position z with a round-trip delay $\tau = 2nz/c$, the received signal can be expressed by
The corresponding FrFT result of the signal is given by
From Eq. (10), it can be seen that the impulse response can be retrieved and the round-trip delay is also obtained from the peak of the sinc signal. The power of the estimated impulse response is scaled by the signal duration T, thus the compression gain of the FrFT is the same as the match filter method. To recover the response of the whole sensing fiber, T should be longer than the round-trip delay of the fiber end. The long-period continuous FrFT-DC signal relieves the demand for high power sensing probe, and thus further alleviates the impact of nonlinearity. To eliminate the signal fading, frequency diversity can be applied to the sensing channel. However, it also reduces the spectral efficiency. Hence, the moving rotated-vector-average (MRVA) method is adopted [19]. To alleviate the amplitude jitter of the signal caused by interference fading, the FrFT result of the signal is processed with moving averaging along the “distance” axis. It should be noted that the MRVA method can mitigate the amplitude fluctuation of the detected signal and reduce the noise power at the expense of sacrificing the spatial resolution.
3. Experimental setup
We carry out experiments to verify the effectiveness of the proposed method for simultaneous sensing and communication. The experimental setup is shown in Fig. 1(a). The transponder of the integrated system comprises the transmitter, the communication receiver, and the sensing receiver.
Fig. 1. (a) Experimental setup; (b) block diagram of Tx Integrated DSP; (c) block diagram of Rx Communication DSP.
The continuous wave (CW) generated by the fiber laser (FL) is used as the transmitter-side (Tx) light source. The operation wavelength and the linewidth of the FL (NKT E15) are 1550.12 nm and ∼100 Hz, respectively. The light is split into two tributaries by a 50:50 polarization maintaining coupler. One tributary serves as the local oscillator (LO) of the sensing receiver. The other is modulated by a single polarization in-phase and quadrature (IQ) modulator driven by arbitrary waveform generator (AWG). As shown in Fig. 1(b), in the Tx digital signal processing (DSP), DSCM signals with 2 subcarriers of 25.6 Gbaud QPSK and 16QAM are offline generated respectively. The electrical DSCM signal is filtered by root-raised cosine (RRC) filters with a roll-off factor of 0.05. There is a guard band of 500 MHz between adjacent two subcarriers. The 500 Mbaud FrFT-DC signal is generated using 0.4 order FrFT and inserted in the center of the DSCM spectrum, as shown by the inset in Fig. 1(a). The theoretical value of PSLR of the sensing probe generated using 0.4 order FrFT is about 140 dB, with a maximum side lobe about -100 dB, as shown in Fig. 2(a). Also, the experimentally measured result is calculated as shown in Fig. 2(b). The measured probe signal is highly compressed as well, and the PSLR is higher than 60 dB. The degradation of the PSLR can be ascribed to the nonlinearity of the modulation and additional noise in the experiment. The period of the FrFT-DC signal is about 125 µs to ensure a coverage of the whole 10 km fiber. Then the digital signal is loaded into a two-channel 8-bit AWG (Keysight M8195A) with a sampling rate of 32 GSa/s to drive the single-polarization IQ modulator. It should be noted that the memory capacity of the AWG is inadequate to store dual-polarization data. Therefore, the polarization-division-multiplexing (PDM) is emulated by a pair of polarization maintaining (PM) polarization-beam splitter (PBS) and polarization-beam combiner (PBC), with a 15-ns optical delay line (DL, 3 m PM fiber) to decorrelate transmitted signals in orthogonal polarizations. The modulated signals, aimed at the 100 Gbps DP-QPSK and 200 Gbps DP-16QAM transmissions, are pre-amplified by an Erbium doped fiber amplifier (EDFA, Amonics AEDFA-PA-35) before propagating over the transmission link through an optical circulator. The noise figure (NF) of the EDFA is 4.2 dB. An optical bandpass filter (OBPF) with a bandwidth of 0.8 nm is used to filter the amplified spontaneous emission. After 10-km fiber transmission, a variable optical attenuator (VOA) is used to adjust the received optical power (ROP).
Fig. 2. (a) The theoretical PSLR of the FrFT-DC signal; (b) the PSLR of the experimentally measured FrFT-DC signal.
At the coherent communication receiver, the integrated signal is detected by an integrated coherent receiver (ICR, NeoPhotonics Class 40). An external cavity laser (ECL, Coherent solutions MTP1000) with a linewidth of 100 kHz is used as the local oscillator. After coherent detection, the received analog signal is digitized by a digital storage oscilloscope (DSO, Teledyne LeCroy 10-36Zi-A) with a sampling rate of 80 GSa/s per channel. The DSP flow for communication is conducted after down-sampling the signal to twice the baud rate, as shown in Fig. 1(c). The Gram-Schmidt orthogonalization procedure (GSOP) and frequency offset compensation (FOC) is then conducted to compensate for receiver-side (Rx) IQ mismatch and frequency offset, respectively. The subsequent DSP module of de-multiplexing to two independent subcarriers is implemented with matched RRC filters. After matched filtering, constant modulus algorithm (CMA) for pre-convergence is implemented for blind equalization. The blind phase search (BPS) is conducted for carrier phase recovery (CPR). Then a blind decision-directed least minimum square (DD-LMS) is applied to correct residual inter-symbol interference (ISI). The bit error rate (BER) is finally averaged over all subcarriers.
At the sensing terminal, the Rayleigh backscattered (RBS) light is mixed with the LO and detected by another ICR (Fujitsu FIM24706/301). It should be noted that polarization diversity detection of ICR eliminates polarization fading in Φ-OTDR and possesses the capacity of coping with PDM signal. Then, the electrical signal is sampled by a four-channel DSO (Teledyne LeCroy, WaveRunner 9404) and processed offline. The sampling rate of the DSO is set to 500 MSa/s and the analogue bandwidth is limited to 200 MHz to extract the sensing signal. The received complex signal corresponding to X and Y polarizations are processed using FrFT, respectively. The length of FrFT window is the same as that of the transmitted FrFT-DC probe signal. The length of the moving rectangular window for MRVA is set as 10 sampling points sliding along the “distance” axis for better signal fading elimination in the experiments. Finally, the complex signals are combined using the rotated-vector-sum (RVS) method to suppress signal fading [20].
4. Experimental results
The integration of communication and sensing signal through one transmitter shares the same digital-to-analog converter (DAC) channels. On one hand, the involvement of the sensing signal inevitably reduces the effective number of bits (ENOB) of DAC and ADC for communication signals, causing a degradation in back-to-back signal-to-noise ratio (SNR). On the other hand, inadequate power allocation to the sensing signal may induce a degradation of SNR for sensing signals. Considering the associated tradeoff, it is imperative to optimize the CSPR generated by the shared transmitter. The CSPR refers to the ratio between the power of communication signal and sensing probe at the transmitter. The power is calculated using the square of the mean absolute amplitude in dB unit. The CSPR can be defined as follows:
where ${A_c}$ and ${A_s}$ denote the amplitude of the communication signal and sensing probe, respectively. The CSPR can be swept by adjusting the scale factor $\kappa $.In the experiments, the optical power launched into the fiber is 5 dBm. The CSPR is swept from 19 dB to -2 dB and the corresponding system performance is evaluated.We start with the investigation of the relationship between the bit error rate (BER) of the communication signal and CSPR at ROP of -36 dBm for QPSK and -28 dBm for 16QAM. As shown in Fig. 3, for both scenarios, the performance of the communication signal is improved at first and then degraded gradually. This can be ascribed to the co-influence of DAC and ADC quantization noise, and nonlinear (NL) distortion. The former is inversely proportional to the ENOB of communication signal, and the latter is proportional to the communication signal power. When the CSPR is relatively high at first, the power of the communication signal is relatively high as well. Hence, the impact of NL penalty dominates. As CSPR decreases slightly, the power of the communication signal decreases accordingly, leading to a mitigation of NL penalty and an enhancement of communication performance. However, further decrease of CSPR severely reduces the ENOB. The quantization noise starts overriding the NL distortion, resulting in a degradation of communication performance.
Correspondingly, the relationship between the sensing performance and CSPR is also investigated. Before sampled by the DSO, the RBS signal is filtered by a 200 MHz analogue bandwidth to avoid interference from backscattered communication signal. Subsequently, the effective RBS signal is demodulated by operating the received RBS signal with FrFT. The SNR and demodulated phase standard deviation (STD) are chosen as the key indicators of sensing performance. To calculate the SNR, we use the effective RBS signals with interference fading elimination. The power of signal and noise is derived from the average amplitude and amplitude variance of the RBS signal, close to the end of the sensing fiber. The phase is retrieved using the effective RBS signal with interference fading elimination. To calculate the effective phase STD, we retrieve the phase of the effective RBS signal with interference fading elimination. And then by averaging the phase STD around the fiber end, we obtain the phase noise floor and sensing accuracy. Note that, the sensing gauge length used to demodulate the phase in averaging is about 4 m. And the phase STD along the sensing fiber is calculated using 40 traces of probing periods. To mitigate the variation of the demodulated phase along the fiber, the phase STD near the fiber end is averaged over a fiber length of about 10 m. Then the obtained average phase STD is taken as the effective phase STD. As shown in Fig. 4(a), for both modulation formats of communication, the SNR of the sensing signal increases with the decrease of CSPR. The noise floor of the sensing system is approximately -50 dB, as depicted in the inset of Fig. 4(a). At the beginning, the power of communication signal dominates in the integrated signal. Then as the CSPR decreases, the resultant power and ENOB for the sensing signal improves. Consequently, the SNR for sensing is boosted at first eventually reaching a state of gradual saturation where the sensing signal becomes the dominant factor. Compared with the 16-QAM communication, the superior overall performance can be observed in the integration of QPSK communication and sensing signal. It can be attributed to the lower peak-average power ratio (PAPR), which relieves the impact of quantization noise of DAC. As shown in Fig. 4(b), the average phase STD generally shows a downtrend with the decrease of CSPR. For relatively higher stage of CSPR, the sensing system is limited by SNR, thus the phase STD decreased remarkably with the increase of SNR. However, at sufficiently high SNR levels, the demodulated phase is significantly impacted by the laser phase noise, which restricts further improvement of phase STD. Based on a thorough assessment of the communication and sensing performance above, an optimal CSPR of 13 dB is selected to use in the following experimental demonstrations.
Next, we investigate the BER evolution with the increment of the received power on the optimal CSPR of 13 dB. The relationship between BER and ROP is shown in Fig. 5. The results indicate that, to achieve line rates of 100 Gb/s with QPSK and 200 Gb/s with 16QAM, the required ROPs are -36.7 dBm and -33.1 dBm according to the staircase hard-decision (HD) forward error correction (FEC) threshold at BER = 4.5E-3, respectively. As the output power is amplified to 5 dBm, which leads to power budgets of 41.7 dB and 38.1 dB respectively. Likewise, the required ROPs are -38.1 dBm for 100 Gb/s and -34.8 dBm for 200 Gb/s corresponding to concatenated soft decision (SD) FEC threshold at BER = 1.2E-2, and corresponding power budgets are 43.1 dB and 39.8 dB, respectively.
Subsequently, to furtherly verify the performance of vibration sensing, a 15-m fiber is wrapped around a piezoelectric transducer (PZT) to imitate external vibrations at the end of the sensing fiber. A sinusoidal waveform with a 1-kHz frequency and a 0.2-V amplitude is used to actuate the PZT. The period of the sensing probe is 125 µs, and about 40 trace periods are acquired. After RVS method applied to eliminate signal fading, the vibration phase is retrieved by differentiating the phase traces with a sensing gauge length of 4 m. The demodulated differential phase of different periods is presented in the time-distance domain, whose top view is shown in Fig. 6(a). There is no phase error except the vibration-induced phase anomaly at the end of the fiber, indicating the signal fading is well eliminated. And the vibration can be clearly discerned as shown in the inset of Fig. 6(a). The STD of the demodulated phase traces is calculated to determine the distributed phase fluctuation and the effective spatial resolution, as shown in Fig. 6(b). The phase STD remains relatively small along the fiber, except a notable peak induced by the vibration from PZT. The conventional definition of spatial resolution is the length between 10% and 90% rising or falling edge. A spatial resolution of 4 m is clearly achieved as shown in the inset of Fig. 6(b). Generally, the original spatial resolution of Φ-OTDR system based on LFM probe is determined by the width of the main lobe after pulse compression. As for the FrFT-DC signal used in the system, the width of the main lobe is about $\cos (0.6\ast \pi /2)/(500MHz) \approx 1\textrm{ }ns$, according to the previous work [16]. Thus, the limitation of the original spatial resolution is about 0.1 m. After adopting MRVR method for signal fading mitigation, the spatial resolution would be about 1 m. Besides, the final spatial resolution is also constrained by the sensing gauge length. Thus, for the same sensing signal, the spatial resolution can be improved by reduction of sampling points used in RVS method as well as the sensing gauge length.
Fig. 6. (a) 3D view of the demodulated vibration phase in time-distance plane; (b) phase PSD along the sensing fiber.
The impact of different communication modulation formats on the sensing performance is then further evaluated. The vibration signal’s amplitude is fixed at 0.2 V. The vibration frequency of the PZT is set to 0.5 kHz, 1 kHz, and 2 kHz, respectively. The period of the sensing probe is maintained as 125 µs, and about 40 trace periods are acquired for each vibration event. The time-domain traces of the demodulated vibration phase are shown in Figs. 7(a) and 7(b) for 16QAM and QPSK, respectively. For better illustration, the phase traces for 0.5 kHz and 2 kHz are shifted up and down respectively. It is clear that the amplitude of the vibration can be accurately restored. Due to the insufficient sampling rate, there exist distortions in 2 kHz vibration signal. The inaccuracy of the reconstructed waveform in Fig. 7(a) is mainly due to the higher phase noise for 16 QAM. As a comparison, the result in Fig. 7(b) shows better accuracy because of lower phase noise for QPSK. Limited by the storage capacity of the DSO, only 40 trace periods are acquired, corresponding to a record length of about 5 ms. It leads to inadequate spectral resolution for spectral analysis of the waveform at 500 Hz. The data with a longer duration can be acquired to further improve the accuracy, on the condition that the DSO with large storage capacity is available. The power spectral density (PSD) corresponding to different vibrations are calculated and shown in Figs. 7(c) and 7(d) for 16QAM and QPSK respectively. All the vibration frequencies are correctly recovered, and no harmonics are observed. The average SNR is about 20 dB. The noise floor in the PSD is about -50 dB rad2/Hz and -52 dB rad2/Hz for 16QAM and QPSK, respectively. The strain resolution can be calculated to be 88 ${{\textrm{p}\mathrm{\varepsilon}} / {\sqrt {\textrm{Hz}}}}$ and 69 ${{\textrm{p}\mathrm{\varepsilon}} / {\sqrt {\textrm{Hz}}}}$ respectively [21]. It can be concluded that higher-order modulation formats have a more significant impact on the sensing performance of the integrated system. This can be attributed to the fact that higher-order modulation formats increase the PAPR of the integrated signal and amplify the impact of quantization noise.
Fig. 7. (a) The demodulated vibration phase of different frequency for 16QAM; (b) the demodulated vibration phase of different frequency for QPSK; (c) PSD of the vibration phases for 16QAM; (d) PSD of the vibration phases for QPSK.
Finally, the linearity of the system response to strain is also verified. In the experiments, the driving voltage of the PZT is varied from 50 mV to 250 mV, and the frequency is maintained as 1 kHz. The linearity of the strain response for both 16QAM and QPSK is shown in Fig. 8. The measured results show a strong linear relationship between the phase and the applied driving voltage. The difference between 16QAM and QPSK may be caused by the different phase noise level. After linear fitting, the linear coefficient (R2) is 0.9991 for 16QAM and 0.9996 for QPSK, which demonstrates a good strain response linearity.
In summary, the study above has demonstrated the achievement of simultaneous vibration sensing and communication in DSCM system over 10 km fiber at 100-Gb/s DP-QPSK and 200-Gb/s DP-16QAM transmission. The performance and cost factors in the system have been taken into comprehensive consideration. In terms of performance, the noise floor in the PSD is approximately -50 dB rad2/Hz and -52 dB rad2/Hz for 16QAM and QPSK, respectively, which closely resembles the results (about -55 dB rad2/Hz at the end of 10 km fiber) in a typical Φ-OTDR system using LFM pulse [20]. We firmly believe that these results could meet the performance requirements of intelligent networks in large-scale short-reach systems. Moreover, the sensing distance can be further extended at the cost of higher phase noise and larger spatial resolution, as mentioned in Ref. [22]. In terms of cost, the transmitter is shared by the sensing and communication channels. As a result, the complexity and cost of the integrated system is significantly reduced, compared to the system simply sharing the fiber with two distinct transceivers. Furthermore, the continuous FrFT-DC sensing probe reduces the demand for high power amplifier. And the baseband sensing signal reduces the demand of bandwidth for the receiver. All these setups facilitate the cost-effectiveness of the integrated system. Here, the guard band between adjacent two subcarriers preserved for sensing is 500 MHz, which is mainly limited by the build-in analogue filter in the DSO. The spectral efficiency can be further improved by custom-designed narrower filters. However, it should be noted that narrower sensing bandwidth must result in poorer sensing performance regarding sensitivity and spatial resolution. Moreover, since the sensing probe is deployed in the baseband, high-pass filters can be easily utilized before ADCs to avoid the impacts from sensing signals on the ENOB at the communication receiver, to promise better communication performance. Although the simultaneous sensing and communication scheme is demonstrated in DSCM using two subcarriers, we believe it is also suitable for multiple subcarriers by taking advantage of the flexibility in spectrum fine-tuning of DSCM system, and it will be further investigated in the future.
5. Conclusions
In this paper, we propose a spectrally multiplexing enabled distributed sensing in DSCM communication system. By utilizing the spectrally flexibility of DSCM system, the continuous FrFT-DC sensing signal is inserted in the central spectrum of DSCM communication signal, thus achieving a cost-effective integration of both sensing and communication signals using the same transmitter. With optimization of the power ratios between sensing and communication signals, we realize 100-Gb/s DP-QPSK and 200-Gb/s DP-16QAM transmission with DAS sensitivity of 69 ${{\textrm{p}\mathrm{\varepsilon}} / {\sqrt {\textrm{Hz}}}}$ and 88 ${{\textrm{p}\mathrm{\varepsilon}} / {\sqrt {\textrm{Hz}}}}$ at a spatial resolution of 4 m. According to the experimental results, the proposed scheme brings a high-performance vibration sensing to the DSCM system, which offers a compelling solution for flexible and intelligent short-reach optical communication systems.
Funding
National Natural Science Foundation of China (61931010, 62205111, 62225110).
Disclosures
The authors declare no conflicts of interest.
Data availability
Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.
References
1. Z. Zhan, M. Cantono, V. Kamalov, A. Mecozzi, R. Müller, S. Yin, and J. C. Castellanos, “Optical polarization–based seismic and water wave sensing on transoceanic cables,” Science 371(6532), 931–936 (2021). [CrossRef]
2. A. Mecozzi, M. Cantono, J. C. Castellanos, V. Kamalov, R. Muller, and Z. Zhan, “Polarization sensing using submarine optical cables,” Optica 8(6), 788–795 (2021). [CrossRef]
3. N. J. Lindsey, T. C. Dawe, and J. B. Ajo-Franklin, “Illuminating seafloor faults and ocean dynamics with dark fiber distributed acoustic sensing,” Science 366(6469), 1103–1107 (2019). [CrossRef]
4. E. Ip, F. Ravet, H. Martins, M. F. Huang, T. Okamoto, S. Han, C. Narisetty, J. Fang, Y. K. Huang, M. Salemi, E. Rochat, F. Briffod, A. Goy, M. del Rosario Fernandez-Ruiz, and M. G. Herraez, “Using global existing fiber networks for environmental sensing,” Proc. IEEE 110(11), 1853–1888 (2022). [CrossRef]
5. F. Walter, D. Gräff, F. Lindner, P. Paitz, M. Köpfli, M. Chmiel, and A. Fichtner, “Distributed acoustic sensing of microseismic sources and wave propagation in glaciated terrain,” Nat. Commun. 11(1), 2436 (2020). [CrossRef]
6. G. Wang, Z. Pang, F. Wang, Y. Chen, H. Dai, and B. Wang, “Urban fiber based laser interferometry for traffic monitoring and analysis,” J. Lightwave Technol. 41(1), 347–354 (2023). [CrossRef]
7. Y. Zeng, X. Wang, W. Li, L. Wang, and M. Tang, “Integrated communication and polarization sensing in self-homodyne coherent systems,” Opt. Lett. 47(18), 4684–4687 (2022). [CrossRef]
8. E. Ip, Y. K. Huang, G. Wellbrock, T. Xia, M. F. Huang, T. Wang, and Y. Aono, “Vibration detection and localization using modified digital coherent telecom transponders,” J. Lightwave Technol. 40(5), 1472–1482 (2022). [CrossRef]
9. H. He, L. Jiang, Y. Pan, A. Yi, X. Zou, W. Pan, A. E. Willner, X. Fan, Z. He, and L. Yan, “Integrated sensing and communication in an optical fibre,” Light: Sci. Appl. 12(1), 25 (2023). [CrossRef]
10. J. M. Marin, I. Ashry, O. Alkhazragi, A. Trichili, T. Khee Ng, and B. S. Ooi, “Simultaneous distributed acoustic sensing and communication over a two-mode fiber,” Opt. Lett. 47(24), 6321–6324 (2022). [CrossRef]
11. E. Ip, Y. K. Huang, M.-F. Huang, F. Yaman, G. Wellbrock, T. Xia, T. Wang, K. Asahi, and Y. Aono, “DAS over 1,007-km hybrid link with 10-Tb/s DP-16QAM co-propagation using frequency-diverse chirped pulses,” J. Lightwave Technol. 41(4), 1077–1086 (2023). [CrossRef]
12. M. F. Huang, P. Ji, T. Wang, Y. Aono, M. Salemi, Y. Chen, J. Zhao, T. J. Xia, G. A. Wellbrock, Y.-K. Huang, G. Milione, and E. Ip, “First field trial of distributed fiber optical sensing and high-speed communication over an operational telecom network,” J. Lightwave Technol. 38(1), 75–81 (2020). [CrossRef]
13. S. Guerrier, K. Benyahya, C. Dorize, E. Awwad, H. Mardoyan, and J. Renaudier, “Vibration detection and localization in buried fiber cable after 80 km of SSMF using digital coherent sensing system with co-propagating 600Gb/s WDM channels,” in Optical Fiber Communication Conference (OFC) (Optica Publishing Group, 2022), paper M2F.3.
14. E. Ip, Y. K. Huang, T. Wang, Y. Aono, and K. Asahi, “Distributed acoustic sensing for datacenter optical interconnects using self-homodyne coherent detection,” in Optical Fiber Communication Conference (OFC) (Optica Publishing Group, 2022), paper W1G.4.
15. D. Welch, A. Napoli, J. Back, et al., “Digital subcarrier multiplexing: enabling software-configurable optical network,” J. Lightwave Technol. 41(4), 1175–1191 (2022). [CrossRef]
16. C. Zhao, L. Wang, H. Wu, and M. Tang, “Optical phase-sensitive reflectometry based on orthogonal frequency division multiplexed LFM signal in fractional Fourier domain,” Laser Photonics Rev. 17(5), 2200838 (2023). [CrossRef]
17. J. Jiang, Z. Wang, Z. Wang, Z. Qiu, C. Liu, and Y. Rao, “Continuous chirped-wave phase-sensitive optical time domain reflectometry,” Opt. Lett. 46(3), 685–688 (2021). [CrossRef]
18. H. M. Ozaktas, Z. Zalevsky, and M. A. Kutay, The Fractional Fourier Transform: with Applications in Optics and Signal Processing. (Wiley, 2001).
19. H. Qian, B. Luo, H. He, Y. Zhou, X. Zou, W. Pan, and L. Yan, “Fading-free Φ-OTDR evaluation based on the statistical analysis of phase hopping,” Appl. Opt. 61(23), 6729–6735 (2022). [CrossRef]
20. D. Chen, Q. Liu, and Z. He, “High-fidelity distributed fiber-optic acoustic sensor with fading noise suppressed and sub-meter spatial resolution,” Opt. Express 26(13), 16138–16146 (2018). [CrossRef]
21. Y. Dong, X. Chen, E. Liu, C. Fu, H. Zhang, and Z. Lu, “Quantitative measurement of dynamic nanostrain based on a phase-sensitive optical time domain reflectometer,” Appl. Opt. 55(28), 7810–7815 (2016). [CrossRef]
22. A. Loayssa, M. Sagues, and A. Eyal, “Phase noise effects on phase-sensitive OTDR sensors using optical pulse compression,” J. Lightwave Technol. 40(8), 2561–2569 (2022). [CrossRef]
