Optical signal modulation systems have gained impetus to augment the functionality of all-optical systems for on-field real-time applications. Such modulation or conditioning systems facilitate distinguishing signals in spectral and/or temporal domains, signal amplification, noise cancellation, and improving signal-to-noise ratio (SNR). Such systems have become an integral part of ultrafast communication systems [1–3], system non-linearity compensation [4], fast switching applications [5], and optical detection and monitoring platforms [6–8]. To materialize versatile optical systems, enhancement and tuning of observed data using passive signal processing techniques have gained importance [9], wherein the recorded information is improved by algorithm-based post-processing. However, in real-time monitoring applications, active signal conditioning is essential for augmenting the functionality of integrated systems.

Dynamic field monitoring optical systems have become significantly viable and are prominently being developed with specialty waveguides and processed waveguides [10–15]. Although the functionalities of waveguide-based optical systems are worthwhile, their performance is influenced by their sensitivity curve, measurand field non-uniformity, and external disturbances. For instance, optical detection systems used for distributed sensing of seismic activities with distinct signal processing techniques [16,17] show low SNR due to external disturbances. Consequently, such systems require active optical signal conditioning to improve their on-field performance. The functionality of most active optical signal conditioning systems is centralized on the spatial processing of sensor responses in multiple optoelectronic probing systems [18] and on signal amplitude modulation. However, along with amplitude modulation, active cancellation of disturbance components and signal enhancement by noise filtering are of paramount importance for comprehensively conditioning the performance of distributed and remote dynamic field monitoring systems. Signal enhancement using metamaterial devices [19], dimensional image restoration techniques [20], additional interferometers [4], fiber laser amplifiers [21], and plasmonic layers [6,22] have proven to be competent methods. However, these optical systems utilize intricate instrumentation and hardly provide any scope for signal attenuation, which is essential for suppressing unwanted components and lowering signal power to comply with threshold limits of integrated devices. Besides, tuning high-frequency dynamic signals in real time with imaging techniques would require complex systems. Additionally, phase demodulation techniques have been realized for signal improvement using multiple fiber interferometer combinations [23,24]. These involve lowering the phase noise in the resultant signal by controlling the signal about reference interferometers. However, these techniques require high coherence laser sources with opto-electronic modulators and are limited to signal filtering only, without any scope for attenuation applications. Hence, realizing real-time active modulation of optical signals that involves signal amplification, attenuation, and filtering over a broad range while the system is in operation has remained elusive.

In this work, we propose a parallel combined identical modal interferometry system, developed using solid core photonic crystal fiber (SCPCF) configuration, for real-time signal modulation over a broad frequency range. SCPCF-based identical interferometers are considered in this work for demonstrating the concept. The similarity in the combined interferometers’ parameters enables the overlap of their characteristic spectra for real-time optical signal conditioning. One arm of the interferometers’ system is used to monitor real-time dynamic fields. The other interferometer arm is subjected to a controllable equivalent dynamic field to achieve signal conditioning of the former arm. We demonstrate active controllable amplification, attenuation, and filtering of dynamic optical response. As an exemplary case, the technique is utilized to condition the response of an in-house optical flowmeter that is based on SCPCF configured modal interferometer. The proposed system successfully lowers the total noise power and improves the resultant signal when combined with the flowmeter at different fluid velocities.

In this work, the interferometers are fabricated by splicing a small section of solid-core photonic crystal fiber (SCPCF) along single-mode fiber (SMF) channels and combining the interferometers in parallel configuration [Fig. 1(a)]. One interferometer acts as the sensing arm (SA) while the other acts as the reference arm (RA). In the case of each interferometer, when SMF is fusion spliced with SCPCF, air voids in the SCPCF clad region collapse over a short length under arc heat [25]. These short regions have uniform refractive index and are called the collapse regions. When a light beam with a broad spectral composition is coupled to the interferometer section, light diffracts at the first collapse region and leads to excitation of higher-order SCPCF modes along with the fundamental mode. The numerical simulation (FEM) of the light beam propagation through the modal interferometer is shown in Fig. 1(b), which depicts the excitation and recombination of specialty fiber modes about the collapse regions. These modes have different propagation constants (β), as the effective indices of the higher-order mode (ne) and the fundamental mode (n1) are different. If the length of the collapsed region is ∼200μm, the excited modes are likely to possess azimuthal symmetry such that the HE11 core mode and the quasi-HE22 type modes (cladding mode) are excited [26]. These modes superpose about the interferometer length (L) and generate an interference pattern over the spectral range as [27] I(λ)=Icr(λ)+Icl(λ)+2Icr(λ)Icl(λ)cos(ϕ),(1)where Icr(λ) and Icl(λ) represent the intensities of core and cladding modes respectively, while ϕ represents the relative phase between them. This phase factor is given by ϕ=κL, where κ represents the mode coupling coefficient value. This coefficient is dependent on the effective modal index difference (Δn=n1−ne), and spectral composition (λ) as κ is given as [28] κ(Δn,λ)=2πηΔnλ,(2)where η is the overlap integral of the core and cladding modes over the fiber cross-sectional plane [29]. In Eq. (1), the spectral components for which κ(λ) values satisfy the mode coupling condition κL=mπ (constructive) form an interference peak while the components for which κ(λ) values satisfy the condition κL=mπ/2 (destructive) result in interference dip. Here, m takes integer values such that m=1,2,3,… and Δn is considered to be constant as it depends on the collapsed region length. Thus, a continuous sinusoidal interference spectrum is generated at the transmission port of each interferometer, and the interferometer is referred to as modal interferometer. The interferometers’ spectra are characterized by their distinct features, namely free spectral range (FSR) and fringe contrast (FC). FSR corresponds to the spectral difference between interference peaks while FC depicts the intensity variation of peak-to-dip of the spectrum (details of FSR and FC are covered in Appendix A). These features of the spectrum can be controlled by configuring the length of the interferometer and collapse region. The SCPCF used for fabricating the modal interferometers has a solid core of diameter 8 μm surrounded by a holey cladding with a 2D photonic structure. The length of each interferometer is chosen to be 12±0.1mm to obtain a moderate FSR value [30] and generate a resultant spectrum having an interference peak at about 1550 nm to minimize propagation loss along SMF. With optimized splicing parameters, the length of the collapse region fabricated is ∼211μm, leading to excitation of the mentioned waveguide modes (see Appendix A for a more detailed description of interferometer fabrication). A schematic representation of the modal interferometer is provided in Figure 4(a) of Appendix A. The interferometers in parallel configuration are powered by a light source with broad spectrum (1400–1600 nm). The recorded spectra for each interferometer [Fig. 1(c)] show that the FSR and FC values of interferometers are equivalent, quantifying the similarity of the interferometers’ spectra (see Appendix B for a more detailed description of data acquisition). Due to the diffraction of light at the collapse regions and insertion losses [31], there is a finite decrease in transmitted power of 6.8 dB. This loss can be reduced by lowering the length of the collapse region while fabricating the interferometers. Besides, as the system is operated by wavelength demodulation instead of intensity interrogation, the technique would function with lowered output intensity. However, the transmitted spectrum is stable and temperature tolerant [32], and acts as a guiding spectrum for sensing applications [33–36]. The spatial frequencies of the interferometers overlap with negligible intermodulation frequencies [inset in Fig. 1(c)]. This implies that each interferometer’s modal composition and effective transmitted spectrum are equivalent. Thus, the phase mismatch value between the spectra of the interferometers is considered to be zero. On combining the outputs of the modal interferometers, the transmitted spectra of the individual interferometers overlap about the interference peaks and dips.

In the presence of an external perturbing field about the interferometers, which alters its effective L (by bending) or effective Δn (by refractive index change) values, its κL values for the spectral components get modulated, resulting in the shift of the peak and dip wavelengths of the spectrum. Under the simultaneous operation of parallel combined interferometers, the resultant shift in the spectrum is a linear combination of each interferometer’s response toward the external field. In the event of SA being subjected to a dynamic field, such that its response is weak and/or noisy, the resultant signal can be conditioned by applying an equivalent dynamic field about RA. Thus, by applying a controlled dynamic field about RA, the resultant signal is amplified, attenuated, or filtered due to modulation of the interferometers’ responses. The proposed system has the potential to act as an optical signal conditioner for systems that implement optical fiber-based interferometers for monitoring real-time dynamic fields such as acoustic field, mechanical vibrations, fluid motion, and environmental surveillance wherein the nature of the external field adversely affects the response of the probe.

The optical signal modulation system is characterized for conditioning the SA response toward the vibration field of varied amplitudes and frequencies. The results for real-time signal amplification, attenuation, and filtering are discussed below. Further, the results for applying the technique to modulate the response of an in-house optical flowmeter are included. The optical flowmeter is developed using SCPCF-based interferometer configuration and the observed results demonstrate the conditioning ability of the proposed technique.

To demonstrate the optical signal modulation ability of the system, the interferometers are subjected to external mechanical vibrations and their composite optical response is recorded. Light from a fiber-coupled CW broadband source is split and coupled to each interferometer arm by the 2×1, 3 dB fiber coupler. The transmitted spectra are combined by a similar 2×1 coupler and fed to a wavelength interrogator that enables tracking of a particular interference peak wavelength of the resultant spectrum. The output signal of the interrogator is observed on a display unit via its user interface. The modal interferometers are mounted on piezoelectric transducers powered by a function generator through a multi-channel piezocontroller (refer to Appendix B for a detailed description of the experimental process). External mechanical vibrations of varied parameters are applied to the interferometers, and each interferometer responds independently toward the external field. When the SA is subjected to mechanical vibrations of specific frequency and amplitude, it undergoes bending along its length that modulates its spectrum. Consequently, the position of the tracked spectrum peak wavelength oscillates periodically with equivalent frequency (f) and amplitude (Δλ) as the applied vibrations. This response of the interferometer towards the external field is considered as the optical signal. This signal of SA is modulated by applying a tunable equivalent field about RA that conditions the resultant spectrum.

In order to demonstrate signal amplification, both interferometers are operated simultaneously. The frequency, amplitude, and initial phase of applied vibrations are fixed about SA such that the optical signal is low with an amplitude (peak shift) of 9 pm. By observing the optical signal of SA, the RA is then subjected to vibrations concurrently of the same frequency, initial phase, and an initial value of amplitude. The resultant signal is a sinusoidal signal of the corresponding frequency with an amplitude equal to the sum of the optical signal amplitudes of interferometers. By varying the vibration amplitude about RA, which is achieved by controlling the voltage supply to the piezoelectric transducer, the amplitude of the resultant signal changes. The increment in the resultant signal amplitude in terms of Δλ (peak-to-dip) is shown in Fig. 2(a). The inset shows a real-time resultant transmission signal of the parallel interferometers for operating only SA (red curve) at 100 Hz and constant amplitude and then operating SA and RA (green curve) simultaneously at 100 Hz with field amplitude of 7 V about RA. If the vibration amplitude about RA increases or decreases, the resultant amplitude increases or decreases linearly over the voltage range. The rate of increase in signal amplitude with applied voltage about RA is observed to be 7.8 pm/V at 100 Hz. A similar type of signal amplification is also observed at different frequencies. The minimum value of amplitude for the resultant signal is equal to the amplitude of the signal in SA when no vibrations are applied to RA. Such amplification ability can be utilized for enhancing the SA response towards low-intensity dynamic fields such as seismic activities, fluid motion, and underwater acoustic fields to meet the threshold signal limits of subsequent integrated instruments.

Signal attenuation deals with suppressing frequency components in SA optical signals by controlling the parameters of a dynamic field about RA. Experimentally, the frequency, amplitude, and initial phase of vibrations about SA are fixed. In real-field applications, at least the frequency and amplitude of the measurement field about SA are easily read out via the interface display. Subsequently, mechanical vibrations of similar frequency and amplitude are applied to RA. In order to attenuate the resultant signal, the relative phase difference between the signals about the interferometers is varied by changing the initial phase of vibrations about RA. As the relative phase difference between the vibration fields increases, the amplitude of the resultant optical signal decreases, reaching a minimum value for a relative phase difference of 180°. The characteristic attenuation plot that has a linear nature is represented in Fig. 2(b). The inset shows a real-time resultant transmission signal of the parallel interferometers for operating only SA (red curve) at 100 Hz, constant amplitude and phase, and then operating SA and RA (green curve) simultaneously at 100 Hz, equivalent amplitudes with an initial phase of field about RA as 180°. A decrement in the signal amplitude of 22 pm/deg phase difference is observed at 100 Hz. In cases where the signal amplitude of SA due to the external dynamic field is higher than the certain threshold level of the detection system, it can be attenuated by controlling the phase of the field about RA. Such utility can be used to avoid physical damage to subcomponents due to high amplitude signals, noise suppression, and active harmonic filtering.

Dynamic field perturbations tend to be noisy due to environmental disturbances, non-linearity in probe responsivity, and the irregular nature of the measurand field. In the presence of such external fields, the response of the SA gets noisy with a very low SNR value that can adversely affect subsequent computational processing of signals. The proposed system enables signal improvement by enhancing the SNR of the observed signal. Here, a noisy signal of SA is superposed with a controlled signal of RA to generate a resultant signal of higher SNR value. Such modulation lowers the jitter in the SA optical signal and improves the signal power. In the experimental setup, random noise is introduced in the piezo vibration signal about SA. As a result, the signal of the interferometer is noisy with low SNR value. Concurrently, a dynamic field of equivalent frequency but low amplitude is applied to RA such that the noisy signal of SA gets modulated to generate a resultant optical signal with a higher SNR value. The enhancement in SNR value with a change in field voltage about RA is computed from the recorded data and shown in Fig. 2(c) with a real-time signal in the inset. By increasing the amplitude of the RA field, the resultant signal becomes more regular. SNR enhancement of about 10.4 dB is achieved by applying only a 1 V field about RA. The SNR values increase and get saturated at a certain value of RA field amplitude. The saturation of SNR value is attributed to interferometer sensitivity that depends on interferometer parameters and its positioning over the piezo transducer that determines the length of fiber being vibrated (l). The maximum deflection of the fiber (w) towards a particular field amplitude depends primarily on the length of the fiber being vibrated (w∝l3). In this case, the maximum deflection value is estimated to be ∼10μm perpendicular to fiber length at an applied field voltage of 3 V. Thus, the resultant signal SNR increases and saturates thereafter due to the limit on the maximum deflection of the vibrating length. However, this limit can be adjusted by altering the interferometer positioning with respect to the RA field.

Additionally, the variation in SNR values with applied field frequency is investigated over the operational range. The SNR values for different field frequencies about SA are maintained at a nearly uniform level, and then its response is modulated with an equivalent low RA field of fixed amplitude. The filtering ability of the system is analyzed over the frequency range and is observed to be nearly uniform over the operational range [Fig. 2(d)]. The mean SNR of only the SA signal is computed to be 9.08 dB. Subsequently, by applying a field of fixed amplitude 0.5 V at RA, the SNR over the frequency range is improved with mean value of 13.94 dB and standard deviation of 0.63 dB. Such operational utility of the proposed system has the potential to improve the system’s response towards low unsteady perturbing dynamic fields encountered in optoelectrical, mechanical devices, which are employed in communications, environmental monitoring systems [37], and industrial surveillance.

To exhibit the optical signal modulation utility of the proposed system, the technique is applied on an in-house optical fluid flowmeter that acts as SA and enhances performance. The optical flowmeter is an SCPCF modal interferometer-based sensing probe that is encapsulated in metal casing and placed in a pipe perpendicular to fluid flow. The probe is fixed near a bluff body that acts as an obstacle and generates fluid vortices downstream of the body. The SA functions as an optical flowmeter and detects the frequency of vortex shedding due to controlled fluid flow in the pipe [38]. The detected frequency of fluid vortices is read out from the interface of the interrogator and is used to determine the fluid velocity. The response of the flowmeter is observed to have a low SNR value at different frequencies and is influenced by external mechanical disturbances. To enhance the resultant signal, an identical modal interferometer, mounted on a controllable piezo transducer, is combined parallel with the flowmeter as RA. The field about RA is tuned to match the parameters of the SA response and modulate the resultant optical signal. The schematic diagram of the experimental setup is shown in Fig. 3(a) and a more detailed version is provided in Figure 6 along with the experimental procedure in Appendix C.

Figure 3(b) depicts the real-time optical signal of the flowmeter at fixed fluid velocities that are actively conditioned by the concurrent operation of RA. For instance, at a fluid velocity of 1.47 m/s, the detected vortex shedding frequency is 74.13 Hz and the observed signal has low SNR. As the RA is operated under an equivalent field (74.13 Hz vibrational field), the signal SNR improves by 4.1 dB while lowering the jitter and total noise level by 2.02 dB. Here, the RA field amplitude is kept low at 0.8 V so that it does not supersede the SA signal amplitude. This is evident from the real-time signal amplitude that increases after applying the field about RA, but the increment remains lower than the initial amplitude. Similar observations are made at other fluid velocities. For SA optical signal amplification at a particular fluid velocity, an equivalent field of similar frequency is applied to RA and its amplitude is varied. The RA field voltage is varied over 0.8–1.4 V and the corresponding peak shift is evaluated, as shown in Fig. 3(c). The amplification rate (peak shift) with respect to RA field voltage is observed to be 12.74 pm/V which is significant for the flow sensing utility. The SNR for the enhanced signal is computed for different RA field voltages and is shown in the inset of Fig. 3(c). The SNR value increases almost linearly with the field voltage at a rate of 7.9 dB/V. For RA field voltage below 0.8 V, the SNR is less than 3 dB, with negligible effect of RA response on the resultant signal. However, this RA field range of operation can be altered by modifying the interferometer parameters and positioning. Additionally, the signal attenuating ability of the system is investigated for the same flowmeter. In this regard, external mechanical vibrations are introduced to the pipe. Experimentally, this is achieved by removing the vibration isolator from the setup that isolates the SA from the vibrations of the pump. The external vibrations are picked up by the flowmeter or SA. As a result, the optical response of SA includes the vortex shedding frequency (signal) and an additional noise component at a frequency of 100 Hz due to pump vibrations. A field of frequency 100 Hz and fixed amplitude (∼20mV) is applied to RA to attenuate the noise signal, and its phase is varied gradually. At the particular initial phase value of the RA field (∼178°), the noise component gets attenuated significantly. A fast Fourier transform of the recorded signals before and after applying field about RA is shown in Fig. 3(d). The resultant signal is then devoid of noise components without any variation in the SA signal. In this way, the unwanted noise components in the resultant signal can be removed without affecting the measurement of the actual field. Similar testing is carried out at different fluid velocities that show equivalent results. The performance of the technique for a definite application depends on the SA response and can be tuned by controlling the RA field. Hence, the proposed system successfully conditions the flowmeter’s response and enhances its performance.

We report a system of modal interferometers in parallel combination to act as a comprehensive system for optical signal modulation over a broad range of frequencies. With the increase in demand for optical sensing systems for real-time dynamic field monitoring, our system offers a technique to modulate the response of optical sensing probes. The identical nature of the modal interferometers facilitates the overlapping of transmission spectra to allow interactive modulation of the interferometer signals. The interferometers’ features, such as lengths of interferometers and collapse region, can be reconfigured to modify the resultant transmission spectrum. The reconfigurability nature of the modal interferometer offers scope to customize the sensitivity and range of operation of the system [30]. Besides, the signal of the sensing interferometer is conditioned by varying the field about the parallel combined reference interferometer. With the proposed system, the optical signal of SA towards an external field is amplified and filtered by controlling the field about RA, resulting in enhancement of the observed optical signal. By reconfiguring the sensitivity of modal interferometers, their resonant frequencies can be varied to enable noise reduction in operating micro-resonators [39]. Since the spatial position of RA can be altered, the proposed system can be handled remotely, offering a wider range of controllability and applicability than previously reported systems. The proposed system has the potential to develop an active technique for optical signal modulation for utility in real-time applications employing optical devices.

Acknowledgment. This work is partially supported by SERB STAR fellowship (Physical Science), DST-FIST, DST-TDT/DDP & NTTM. This work was developed within the scope of the projects CICECO and DigiAqua, financed by national funds through the Portuguese Science and Technology Foundation/MCTES (FCT I.P.). The research was co-funded by the financial support of the European Union under the REFRESH–Research Excellence for Region Sustainability and High-tech Industries via the Operational Programme Just Transition. This work was also supported by the Ministry of Education, Youth, and Sports of the Czech Republic conducted by the VSB-Technical University of Ostrava.