Open Access
Add to BibSonomy
Abstract
Multi-channel microwave photonic (MWP) signal processing can simultaneously perform different task operations on multiple signals carried by multiple wavelengths, which holds great potential for ultrafast signal processing and characterization in a wavelength-division-multiplexed (WDM) network. As emerging telecommunication services create more data, an elastic optical network, which has a flexible and non-uniform spectrum channel spacing, is an alternative architecture to meet the ever-increasing data transfer need. Here, for the multi-channel ultra-fast signal processing in the elastic optical network, we propose and demonstrate an on-chip non-uniformly spaced multi-channel microwave photonic signal processor based on an ultrahigh-Q multimode micro-disk resonator (MDR). In the proposed signal processor, an MDR supporting multiple different order whispering-gallery modes (WGMs) with an ultrahigh Q-factor is specifically designed. Benefiting from the large and different free spectral ranges (FSRs) provided by the different order WGMs, a non-uniformly spaced multi-channel microwave photonic signal processor is realized, and various processing functions are experimentally demonstrated including bandpass filtering with a narrow passband of 103 MHz, a rejection ratio of 22.3 dB and a frequency tuning range from 1 to 30 GHz, multiple frequency measurement with a frequency measurement range from 1 to 30 GHz, a frequency resolution better than 200 MHz and a measurement accuracy of 91.3 MHz, and phase shifting with a phase tuning range from -170°∼160°, an operational bandwidth of 26 GHz from 6 GHz to 32 GHz and a small power variation of 0.43 dB. Thanks to the coexistence of different order WGMs supported by the MDR, non-uniformly spaced multi-channel signal processing is enabled with the key advantages including a broad operation bandwidth, an ultra-narrow frequency selectivity, and a large phase tuning range with a small power variation. The proposed signal processor is promising to be widely used in future elastic optical networks with flexible spectrum grids.
© 2023 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
Thanks to large frequency tuning range offered by modern photonics, microwave signal photonic processing is a topic of interest in recent years [1–6], and numerous processing solutions have been proposed and demonstrated, including microwave filtering [7–13], microwave frequency measurement [14–20], and microwave phase-shifting [21–27]. Benefiting from the rapid development of photonic integrated circuit (PIC) technologies, microwave photonics (MWP) signal processing has been coming into an integration era. Much effort has been directed into the study, and various on-chip MWP signal processors have been reported [28–31]. In particular, due to the distinct advantages of small footprint, low power consumption and high frequency selectivity, on-chip optical cavities, as a key component, have found extensive applications in integrated MWP signal processors [32,33].
Due to the planner process offered by the integrated photonic fabrication platform, there are two main on-chip optical micro-cavities: micro-ring resonators (MRRs) and micro-disk resonators (MDRs). The planner fabrication imperfections cause the waveguide sidewall roughness, which leads to a main source of the optical waveguide propagation loss. Specifically, in an MRR, the sidewall roughness of the inner and outer sidewalls can contribute to an optical propagation loss, which limits the frequency selectivity of the resonator. To have a high frequency selectivity, a large radius of an MRR is usually required, of which the cost is a small free spectral range (FSR), limiting the frequency tuning range of the realized MWP processor. Furthermore, the MRR waveguide usually supports single-mode operation and thus the MRR features a periodic spectrum. With the MRR, a uniformly spaced frequency channel of the MWP processor is resulted, in which multiple laser sources having a rigorously uniform spaced wavelength are required.
Unlike an MRR that suffers from scattering loss caused by sidewall roughness from both inner and outer sidewalls, an MDR experiences scattering loss only from the outer sidewall, thus the overall scattering loss is highly reduced. Therefore, compared with the MRR, the MDR can have a higher Q-factor and a larger FSR simultaneously, which is much preferred in a wide frequency tuning range and high-frequency-resolution MWP signal processor. Moreover, since the disk waveguide can support multiple mode, the multiple different order whispering-gallery modes (WGMs) with different FSRs co-exist in a single resonator. With the exponential growth of data traffic due to the multimedia services, an elastic optical network (EON) architecture is considered as a promising solution for next-generation optical network [34]. Distinct from the uniform spectrum grid in the current optical networks, the spectrum grid in an EON is usually non-uniform, which depends on the user application request. To address the high-speed signal processing need of non-uniformly spaced frequency channels in an EON, a MWP processor based on an MDR is potentially a strong candidate.
In this paper, we propose and demonstrate an on-chip non-uniformly spaced multi-channel microwave photonic signal processor based on a high-Q multimode micro-disk resonator. In the proposed MWP signal processor, an ultrahigh-Q multimode MDR supporting multiple different order WGMs with an ultrahigh Q-factor is specifically designed. Benefiting from the large and different FSRs provided by the different order WGMs, a non-uniformly spaced multi-channel microwave photonic signal processor can be realized, and various processing functions are experimentally demonstrated including bandpass filtering with a narrow passband of 103 MHz, a rejection ratio of 22.3 dB and a frequency tuning range from 1 to 30 GHz, multiple microwave frequency measurement with a frequency measurement range from 1 to 30 GHz, a frequency resolution better than 200 MHz and a measurement accuracy of 91.3 MHz, and phase shifting with a phase tuning range from -170°∼160°, an operational bandwidth of 26 GHz from 6 GHz to 32 GHz and a small power variation of 0.43 dB. Thanks to the coexistence of different order WGMs supported, this specifically designed ultrahigh-Q MDR features non-uniformly spaced multi-channel signal processing with the key advantages including a broad operation bandwidth, an ultra-narrow frequency selectivity, and a large phase tuning range with a small power variation. Our proposed on-chip signal processor is promising to be widely used in future elastic optical networks with flexible spectrum grids.
2. Principle
Figure 1 illustrates the perspective view of the proposed on-chip non-uniformly spaced multi-channel microwave photonic signal processor. Continuous-wave (CW) light waves with different wavelengths generated by a set of tunable laser sources (TLSs) are launched into the on-chip wideband dual-parallel Mach-Zehnder modulators (DP-MZMs) via polarization controllers (PCs), which are used to align the polarization states of the input light waves into that of the waveguide mode of the modulators. The incoming microwave signals to be processed are modulated on the optical carriers with the use of the DP-MZMs. By specifically controlling the bias condition of the modulator, phase-modulated or single-sideband intensity modulated optical signal can be generated [35]. At the output of the DP-MZMs, different modulated optical signals are combined using an optical coupler (OC) and then launched into a specifically designed ultrahigh-Q multimode MDR, which is leveraged to perform various signal processing functions including bandpass filtering, frequency measurement, and phase shifting. The processed optical signals are guided into an optical filter bank which divides the signal into different sub-channels with a non-uniformly frequency spacing. After photodetection by the high-speed photodetectors (PDs), the processed microwave signals are recovered and collected by a set of high-speed analog-to-digital converters (ADCs). Thanks to the coexistence of different order WGMs supported by the MDR, this specifically designed ultrahigh-Q MDR supports non-uniformly spaced multi-channel signal processing with the key advantages including a broad operation bandwidth, an ultra-narrow frequency selectivity, and a large phase tuning range with a small power variation.
2.1 Tunable bandpass filtering
As one of the fundamental processing functions, microwave filtering with a high frequency selectivity and a wide frequency tuning range is highly needed. Based on the proposed on-chip MWP signal processor, a widely tunable bandpass microwave photonic filter (MPF) with a high frequency selectivity can be implemented. As shown in Fig. 1, a CW light wave is coupled into the DP-MZM via an input grating coupler. By biasing the two sub-Mach-Zehnder modulators (sub-MZMs) and the main Mach-Zehnder modulator (main-MZM) at the minimum transmission point, the maximum transmission point, and the quadrature transmission point respectively, the DP-MZM works as an equivalent phase modulator. At the output of the DP-MZM, the phase-modulated optical signal can be expressed under the small signal condition:
2.2 Multiple microwave frequency measurement
Frequency measurement of microwave or millimeter-wave signals is a very important measurement task in modern electronic systems, such as electronic countermeasure, radar warning, and electronic intelligence systems. Conventionally, microwave frequency measurement is realized based on electronic components, which suffers from a limited frequency measurement range and vulnerability to electromagnetic interference (EMI). Recently, photonics-assisted microwave frequency measurement techniques have attracted considerable attention, which can perform the frequency measurement of the microwave signals in the optical domain, holding unique advantages including wide frequency measurement range, ultra-fast measurement speed, low power consumption, and immunity to EMI.
With our proposed MWP signal processor, a high-resolution multiple microwave frequency measurement system can be done based on linearly frequency sweeping of the realized bandpass MPF. By applying a periodic driving signal on the TLS to have a wavelength sweeping, the resulted MPF also has a periodic linear center frequency sweeping. Thus, a linear mapping between the time and microwave signal frequency is created, which can be used in the frequency measurement. Specifically, a triangle waveform generated by an arbitrary waveform generator (AWG) is applied to the TLS to have a roundtrip wavelength sweeping, which leads to a roundtrip center frequency sweep of the resulted bandpass MPF. Supposing an unknown microwave signal is launched into this measurement system, at the output of the system a temporal electrical waveform will be recorded with an ADC. At two specific times there are a peak. By measuring the time interval between the two peaks, the frequency of the unknown microwave signal can be identified based on the linear mapping relationship between the sweeping time and center frequency of the MPF. Mathematically, when the frequency sweeping range and period of the MPF are $\Delta f$ and T, respectively, and the time interval is measured to be $\Delta t$, the measured microwave frequency can be given by
With our proposed MWP signal processor, the frequency measurement resolution and range are determined by the bandwidth and frequency tuning range of the MPF. Thanks to the ultra-high Q-factor of the MDR, a high resolution is achieved, and a wide frequency measurement range is enabled using the photonics approach. In addition to the single tone measurement, due to its fast sweeping speed over a wide frequency range, the proposed system can perform multi-tone measurements, which is highly preferred in the electronic warfare systems.
2.3 Wideband microwave phase shifting
Microwave phase shifting is another key signal processing function that fulfill a tunable phase shifting of an input microwave signal over a wide frequency range. with our proposed on-chip MWP signal processor, a wideband MWP phase shifter could also be done. A CW optical signal generated by a TLS is launched into the DP-MZM. The two sub-MZMs and the main-MZM are biased at the quadrature transmission point to perform single-sideband (SSB) modulation. Under small signal condition, the SSB optical signal at the output of DP-MZM can be expressed as:
Then the SSB optical signal is guided into the ultrahigh-Q MDR. To realize the phase shifting, the optical carrier is chosen to be located at the resonance wavelength of MDR. Thus, the optical signal at the output of MDR is given by:
3. Experimental demonstrations
In the proposed microwave photonic signal processor, the specially designed ultrahigh-Q multimode MDR is a key component, whose transmission and phase responses determine the performance of the microwave photonic signal processor. To verify the feasibility of the microwave photonic signal processor, the ultrahigh-Q multimode MDR is firstly designed and fabricated, and a microwave photonic signal processing system is implemented with the help of other commercially available optical devices. Due to the number limitation of the tunable laser sources available in the lab, in the experimental demonstration different signal processing functions in the different channel is performed sequentially. In the proposed system, a phase-modulated signal is required to perform bandpass filtering and frequency measurement, while a single-sideband intensity modulated signal is required to perform phase shifting. By specifically controlling the bias condition of a DP-MZM, phase-modulated or single-sideband intensity modulated optical signal can be generated. In the experimental demonstration, due to the lack of wideband DP-MZM in the lab, a wideband phase modulator (PM) and a tunable flat-top optical bandpass filter (OBPF) with an ultrahigh spectral slope rate are jointly used in the system to generate phase-modulated and single-sideband intensity modulated optical signals. The experimental setup is shown in Fig. 2. An CW optical carrier with an output power of 10 dBm is generated by the TLS and sent to a PM with a bandwidth of 40 GHz via the PC1. The microwave signal to be processed is modulated on the optical carrier with the use of the PM. The phase-modulated optical signal is then launched into the OBPF. At the output of the OBPF, the optical signal is injected into the integrated MDR via the PC2, which is used to minimize the polarization-dependent loss. An erbium-doped fiber amplifier (EDFA) is connected after the MDR to compensate the coupling loss between the fiber array and the chip. The amplified optical signal is split into two parts by an optical coupler. One part is launched into the optical spectrum analyzer (OSA) for real-time monitoring the optical spectrum, while the other part is sent to the high-speed PD with a bandwidth of 33 GHz. The microwave signal generated at the output of the PD is then collected by a high-speed ADC. Thanks to the multimode operation supported by the MDR, non-uniformly spaced multi-channel signal processing is enabled, which would highly relax the wavelength restriction of the multiple laser sources.
In the proposed on-chip MWP signal processor, the ultrahigh-Q multimode MDR is a key component. An ultrahigh-Q multimode MDR is designed and fabricated. As shown in the inset of Fig. 2, the MDR has an all-pass configuration, of which the radius is designed to be 40 µm. A slab waveguide with a height of 70 nm is incorporated to wrap the disk and the bus waveguide with an aim to weaken the disk sidewall roughness and to increase the confinement of the optical field in the disk [36]. To fulfill the phase-matching conditions, the width of the bus waveguide and the coupling gap are designed to be 850 nm and 200 nm, respectively. The designed MDR is fabricated on a standard SOI wafer with a 220-nm-thick top layer and a 2-µm-thick buried oxide layer.
Figure 3(a) shows the measured transmission spectrum in the blue line and its phase response in the red line of the fabricated MDR. As can be seen, thanks to the disk waveguide, the MDR supports multiple WGMs. Figure 3(b) shows the zoom-in view of the transmission spectrum of the WGM near 1554.269 nm in the blue line. The bandwidth of the notch is measured to be 0.79 pm, corresponding to an ultrahigh Q-factor of 1.97 × 106, which is of highly benefit to realizing a high frequency selectivity of the MPF. The red line gives the phase response of the WGM in which the phase shift at the resonance wavelength is measured to be 0.37 rad. Figure 3(c) shows the zoom-in view of the transmission spectrum of the WGM near 1564.626 nm in the blue line. The bandwidth of the notch is measured to be 2.88 pm, corresponding to an ultrahigh Q-factor of 5.43 × 105. The red line gives the phase response of the WGM in which the phase shift at the resonance wavelength is measured to be 5.93 rad. Thanks to the multi-mode operation supported by the MDR, different WGMs feature different frequency responses, which provides a rich channel source for optical signal processing. In addition, benefiting from the small footprint of the MDR, a large FSR of the WGMs can be achieved, which can ensure the broad frequency tuning range of the MWP signal processor. For the fabricated MDR with a radius of 40 µm, the theoretical FSR of the fundamental MDR is calculated to be as large as 2.6 nm.
Fig. 3. (a) Measured transmission spectrum and its phase response of the MDR. Zoom-in view of the WGM near (b) 1554.269 nm and (c) 1564.626 nm.
3.1 Widely tunable bandpass filtering
Thanks to its ultra-large Q-factor, the WGM at 1554.269 nm of the MDR is used as the ultra-narrow optical notch filter to remove the -1st-order sideband, and the wavelength of the optical carrier is chosen at 1554.421 nm. Fig. 4(a) shows the realized bandpass MPF with a center frequency at 19 GHz. The bandwidth of the bandpass MPF is measured to be 103 MHz, which matches well with the bandwidth of the notch of the WGM at 1554.269 nm. Figure 4(b) shows the rejection ratio of the realized MPF as large as 22.3 dB. The rejection ratio could be further improved by optimizing the coupling design of the WGM for critical coupling. When the wavelength of the optical carrier is tuned from 1554.277 to 1554.509 nm, the center frequency of the MPF is tuned from 1 to 30 GHz accordingly, as shown in Fig. 4(c). As can be seen, the shape of the MPF remains almost unchanged during the tuning of the center frequency. The frequency tuning range of the realized MPF is limited to half of the frequency spacing between the adjacent WGMs, which is about 32 GHz. Figure 4(d) shows the measured bandwidth and rejection ratio of the bandpass MPF when its center frequency is tuned from 1 to 30 GHz. As can be seen, during the center frequency tuning, the narrow passband and high rejection ratio of the MPF maintain, which verifies the fact that the MPF holds the key advantage of a wide frequency tuning range. In the experiment, a thermoelectric-cooler is used to stabilize the silicon chip temperature at 26 °C. The stability of the MPF is experimentally evaluated within a duration of 30 minutes. The root-mean-square (RMS) value of the center frequency change is calculated to be 10.3 MHz, which is mainly caused by the wavelength fluctuation of the TLS and the resonance wavelength change of the MDR.
Fig. 4. (a) Measured frequency response of the MPF at 19 GHz. (b) Measured rejection ratio of the MPF at 19 GHz. (c) Frequency responses of the realized MPF when the center frequency is tuned from 1 to 30 GHz. (d) Measured bandwidth and rejection ratio of the MPF at different center frequencies.
Table 1 shows the performance comparison between the realized MPF and other existing integrated bandpass MPFs in terms of tuning range, bandwidth, and rejection ratio. As can be seen, the proposed MPF based on silicon photonic MDR is capable of simultaneously providing a wide tuning range and a narrow passband, while the rejection ratio is yet to be improved.
Table 1. Performance Comparison of Existing Integrated Bandpass MPFs
3.2 Multiple microwave frequencies measurement
In the experimental demonstration of multiple microwave frequencies measurement, a triangle waveform generated by an AWG is applied to the TLS to sweep the wavelength of the optical carrier, which leads to the center frequency tuning of the realized bandpass MPF. The temporal period of the triangle waveform is set to be 500 µs, and its amplitude is set to be 150 mV, which has a wavelength sweeping range of the TLS as large as 0.25 nm, corresponding to a frequency measurement range over 31.25 GHz. The output electrical signal of the measurement system is acquired by a 12-bits ADC with a sampling rate of 10 MSa/s.
A single-tone frequency measurement test is firstly performed. An input microwave signal whose frequency is changed from 5 to 25 GHz with a step of 5 GHz is sent into the measurement system. Fig. 5(a) shows the measured electrical signal collected by the ADC at the output of the measurement system. As can be seen, thanks to the symmetrical round-trip frequency sweeping, the recorded electrical signal is also symmetrical. By measuring the time intervals between the two symmetrical peaks, the frequency of the input microwave signal can be identified. The relationship between the microwave frequency f (GHz) and the measured time interval $\Delta t$ (µs) can be determined by
where ${f_{offset}}$ is the constant frequency offset caused by the wavelength interval between the optical carrier and the WGM of the MDR. In our measurement system, ${f_{offset}} = {0.66}\,GHz$.Fig. 5. (a) Measured output electrical signal when the input microwave signal frequency is changed from 5 GHz to 25 GHz with a step of 5 GHz. (b) Frequency measurement result over 1-30 GHz. (c) Frequency measurement errors over 1-30 GHz, showing an RMS of 91.3 MHz.
In Fig. 5(a), the time intervals are measured to be 70.8 µs, 151.0 µs, 230.2 µs, 309.2 µs, and 389.5 µs, corresponding to the microwave frequencies of 5.09 GHz, 10.10 GHz, 15.05 GHz, 19.99 GHz, and 25.01 GHz, which verifies the effectiveness of the frequency measurement. To evaluate the measurement bandwidth and accuracy, a frequency estimation measurement over 1-30 GHz is performed. Figure 5(b) shows the frequency measurement result when an input microwave signal with a frequency from 1 to 30 GHz with a step of 1 GHz is injected. After linear fitting, an R2 value as high as 0.9997 is achieved, which demonstrates the good match between the estimated frequencies and the real ones. Fig. 5(c) gives the frequency estimation errors over 1-30 GHz, of which the RMS value is calculated to be 91.3 MHz.
Then, a multi-tone frequency measurement is performed. A three-tone microwave signal including the frequencies of 5 GHz, 9 GHz, and 13 GHz is launched into the system simultaneously. Fig. 6(a) shows the measured output electrical signal of the system. Again, three pairs of high peaks corresponding to three different frequency components are experimentally captured. The time intervals between the peaks are measured to be 69.4 µs, 133.3 µs, and 196.6 µs, corresponding to the microwave frequencies of 5.00 GHz, 8.99 GHz, and 12.95 GHz. To estimate the frequency resolution of the system, a six-tone signal with frequencies ranging from 9.9 to 10.9 GHz with a step of 200 MHz is sent to the microwave frequency measurement system. The measurement result is shown in Fig. 6(b), and the inset gives the zoom-in view of the measured electrical peaks. As can be seen, six separated peaks corresponding to six frequency components can be observed. The time intervals between the peaks are measured to be 148.0 µs, 151.0 µs, 154.2 µs, 157.4 µs, 160.6 µs, and 163.7 µs, corresponding to the microwave frequencies of 9.91 GHz, 10.10 GHz, 10.30 GHz, 10.50 GHz, 10.70 GHz, and 10.89 GHz. Theoretically, the frequency measurement resolution is mainly determined by the bandwidth of the MPF realized in this work. However, when multi-tone microwave signal with a small frequency interval is launched into the system, the electrical outputs corresponding to different frequency components will interfere with each other and leads to a distorted temporal electrical waveform. Therefore, the frequency resolution is lightly larger than the bandwidth of the MPF [37]. In the proposed system, a frequency measurement resolution better than 200 MHz is realized.
Fig. 6. (a) Measured electrical outputs when a three-tone microwave signal with the frequencies of 5 GHz, 9 GHz, and 13 GHz is sent to the system. (b) Measured electrical outputs when a six-tone microwave signal with the frequencies of 9.9 to 10.9 GHz with a step of 200 MHz is sent to the system.
In the experiment, a chirped-frequency microwave signal measurement is also performed. Figure 7(a) shows the measurement results of two chirped-frequency microwave signals with a center frequency of 4 GHz (CF1) and 8 GHz (CF2), and a bandwidth of 1 GHz. As can be seen, the generated electrical peaks have the similar temporal durations. The measured time intervals of the CF1 at the inner and outer edges are 46.5 µs and 62 µs, corresponding to the frequency from 3.57 GHz to 4.54 GHz. The measured time intervals of the CF2 at the inner and outer edges are 109.2 µs and 124.9 µs, corresponding to the frequency from 7.49 GHz to 8.47 GHz, which matches well with the input waveform. Then, two CF microwave signals with a center frequency of 10 GHz and a bandwidth of 2 GHz (CF3) and 4 GHz (CF4) are experimentally launched into the measurement system, and the results are shown in Fig. 7(b). The measured time intervals of the CF3 at the inner and outer edges are 134.7 µs and 165.5 µs, corresponding to the frequency from 9.08 GHz to 11.01 GHz. The measured time intervals of the CF4 at the inner and outer edges are 118.8 µs and 189.4 µs, corresponding to the frequency from 8.09 GHz to 11.94 GHz. The measured data match well with the input waveform, which verifies the effectiveness of the measurement again.
Fig. 7. (a) Measurement results of two chirped-frequency microwave signals with a center frequency of 4 GHz (CF1) and 8 GHz (CF2), and a bandwidth of 1 GHz. (b) Measurement results of two chirped-frequency microwave signals with a center frequency of 10 GHz and different bandwidths of 2 GHz (CF3) and 4 GHz (CF4).
Table 2 shows the performance comparison between the realized photonic-assisted microwave frequency measurement system and other existing integrated frequency measurement system in terms of measurement range, accuracy, and signal type. As can be seen, the proposed microwave frequency measurement system is capable of measuring single-tone, multi-tone, chirped-frequency signals over a wide frequency range. The measurement accuracy is 91.3 MHz, which is determined by the finite 3 dB bandwidth of the MPF (103 MHz) and the limited signal-to-noise ratio (SNR) of the system. The measurement accuracy can be further improved by narrowing the bandwidth of the MPF and improving the SNR of the system.
Table 2. Performance Comparison of Existing Integrated Frequency Measurement System
3.3 Wideband microwave phase shifting
In the experiment of wideband MWP phase shifter, the +1st order sideband is removed by the OBPF and the generated SSB modulation signal is injected to the MDR. The optical carrier is located in the notch of the WGM around 1564.626 nm, where a phase shift is introduced into the optical carrier. The -1st-order sideband is located out of the notch, where no phase shift is introduced. Then, the phase-shifted optical carrier and the -1st-order sideband are beat in the high-speed PD, generating a phase-shifted microwave signal. By tuning the wavelength of the TLS, the phase of the output microwave signal can be tuned.
As shown in Fig. 8(a), when the wavelength of the optical carrier is tuned from 1564.6176 nm to 1564.6339 nm, the phase of the microwave signal can be tuned from -170° to 160°. The operational bandwidth of the MWP phase shifter is as wide as 26 GHz from 6 GHz to 32 GHz, which is mainly limited by the bandwidth of the PD. Figure 8(b) shows the measured power variations of the microwave signals with different phase shifting. During the phase shift, the power variation is measured to be less than 0.43 dB. The mechanism that leads to a small power variation mainly includes two aspects: the nonlinear gain characteristics of the EDFA and the saturation characteristics of the PD. The nonlinear gain characteristics exist in most optical amplification devices such as EDFA, in which the gain of the amplification device depends on the input power level and a self-equalization gain correction effect occurs to minimize the fluctuations in the output power [27]. Besides, the saturation characteristics of the PD is also of benefit to reduce the output power variation, in which the responsivity of the PD is dependent on the optical power injected into the PD [38].
Table 3 shows the performance comparison between the realized MWP phase shifter and other existing MWP phase shifter in terms of operational bandwidth, phase tuning range, and power variation. As can be seen, the proposed MWP phase shifter has a wide operational bandwidth and a small power variation. The phase tuning range of the proposed phase shifter can be further increased by cascading two MDRs [39].
Table 3. Performance Comparison of Existing Integrated MWP Phase Shifter
Overall, by incorporating the specifically-designed MDR in the signal processor, three signal processing functions are experimentally demonstrated. Thanks to the resonator nature of the MDR, more processing functions can be made including temporal differentiation, Hilbert transformation, and time delay. With a multiple wavelength laser source, by aligning the optical carrier wavelengths into the resonance wavelengths of the different WGM mode, a non-uniform multi-channel optical signal processor could be realized. The processing function of each channel could be reconfigured upon request.
4. Discussion
In the proposed microwave photonic signal processor, the frequency resolution is determined by the Q-factor of the multimode MDR, which is used to perform signal processing in the optical domain. For MDR, the Q-factor is limited by the intrinsic material absorption loss and the scattering loss induced by surface roughness due to fabrication imperfections [36]. In this work, an ultrahigh-Q multimode MDR is specially designed, in which a slab waveguide is incorporated to wrap the disk and the bus waveguide with an aim to weaken the disk sidewall roughness and to reduce the scattering loss. However, when the Q-factor of the MDR becomes high, optical bistability may occur at a low input optical power level due to two-photon absorption induced nonlinear thermal-optic effect [40], which leads to a distorted spectral response for the MDR and deteriorates the performance of the microwave photonic signal processor. Therefore, to achieve a higher frequency resolution for signal processing, a chip fabrication process with a higher precision and materials with a lower absorption loss are highly desired to improve the Q-factor of the MDR, and the optical power launched into the MDR should be properly controlled to avoid optical bistability.
In this work, a multi-channel microwave photonic signal processor consisting of several wideband DP-MZMs, an OC, an ultrahigh-Q multimode MDR, an optical filter bank, and several wideband PDs has been proposed. Among them, the specially designed ultrahigh-Q multimode MDR is a key component, whose transmission and phase responses determine the performance of the microwave photonic signal processor. Therefore, the ultrahigh-Q multimode MDR is firstly designed and fabricated, and a microwave photonic signal processing system is implemented with the help of other commercially available optical devices. In the future work, a silicon photonic integrated microwave photonic signal processor including all the photonic devices will be fabricated, and a multi-wavelength laser array will be used to perform multi-channel microwave photonic signal processing simultaneously.
5. Conclusion
In conclusion, we proposed and experimentally demonstrated an on-chip non-uniformly spaced multi-channel microwave photonic signal processor based on a high-Q multimode micro-disk resonator. In the proposed MWP signal processor, an ultrahigh-Q multimode MDR supporting multiple different order WGMs with an ultrahigh Q-factor was specifically designed and fabricated. Benefiting from the large FSR provided by the different order WGMs of the MDR, a non-uniformly spaced multi-channel microwave photonic signal processor can be realized, and different processing functions are experimentally demonstrated including bandpass microwave filtering with a narrow passband of 103 MHz, a rejection ratio of 22.3 dB and a frequency tuning range from 1 to 30 GHz, multiple microwave frequency measurement with a frequency measurement range from 1 to 30 GHz, a frequency resolution better than 200 MHz and a measurement accuracy of 91.3 MHz, phase shifting with a phase tuning range from -170°∼160°, an operational bandwidth of 26 GHz from 6 GHz to 32 GHz and a small power variation of 0.43 dB. Thanks to the coexistence of different order WGMs supported by the MDR, non-uniformly spaced multi-channel signal processing is enabled with the key advantages including a broad operation bandwidth, ultra-narrow frequency selectivity, large phase tuning range at the cost of a small power variation. The proposed signal processor is promising to be widely used in future elastic optical networks with flexible spectrum grids.
Funding
National Key Research and Development Program of China (2018YFE0201800); National Natural Science Foundation of China (62005018, 62071042, U22A2018); China Postdoctoral Science Foundation (2020M680381); Natural Science Foundation of Chongqing (cstc2020jcyj-msxmX0673).
Disclosures
The authors declare that there are no conflicts of interest related to this article.
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. J. Capmany and D. Novak, “Microwave photonics combines two worlds,” Nat. Photonics 1(6), 319–330 (2007). [CrossRef]
2. J. P. Yao, “Microwave photonics,” J. Lightwave Technol. 27(3), 314–335 (2009). [CrossRef]
3. R. A. Minasian, “Photonic signal processing of microwave signals,” IEEE Trans. Microwave Theory Tech. 54(2), 832–846 (2006). [CrossRef]
4. R. A. Minasian, E. H. W. Chan, and X. Yi, “Microwave photonic signal processing,” Opt. Express 21(19), 22918–22936 (2013). [CrossRef]
5. J. Capmany, J. Mora, I. Gasulla, J. Sancho, J. Lloret, and S. Sales, “Microwave photonic signal processing,” J. Lightwave Technol. 31(4), 571–586 (2013). [CrossRef]
6. R. A. Minasian, “Ultra-wideband and adaptive photonic signal processing of microwave signals,” IEEE J. Quantum Electron. 52(1), 1–13 (2016). [CrossRef]
7. Y. Liu, A. Choudhary, D. Marpaung, and B. J. Eggleton, “Integrated microwave photonic filters,” Adv. Opt. Photonics 12(2), 485–555 (2020). [CrossRef]
8. Y. Xie, A. Choudhary, Y. Liu, D. Marpaung, K. Vu, P. Ma, D.-Y. Choi, S. Madden, and B. J. Eggleton, “System-level performance of chip-based Brillouin microwave photonic bandpass filters,” J. Lightwave Technol. 37(20), 5246–5258 (2019). [CrossRef]
9. A. Choudhary, I. Aryanfar, S. Shahnia, B. Morrison, K. Vu, S. Madden, B. Luther-Davies, D. Marpaung, and B. J. Eggleton, “Tailoring of the Brillouin gain for on-chip widely tunable and reconfigurable broadband microwave photonic filters,” Opt. Lett. 41(3), 436–439 (2016). [CrossRef]
10. X. Xu, M. Tan, J. Wu, T. G. Nguyen, S. T. Chu, B. E. Little, R. Morandotti, A. Mitchell, and D. J. Moss, “Advanced adaptive photonic RF filters with 80 taps based on an integrated optical micro-comb source,” J. Lightwave Technol. 37(4), 1288–1295 (2019). [CrossRef]
11. H. Yang, J. Li, G. Hu, B. Yun, and Y. Cui, “Hundred megahertz microwave photonic filter based on a high Q silicon nitride multimode microring resonator,” OSA Continuum 3(6), 1445–1455 (2020). [CrossRef]
12. H. Qiu, X. Zhang, F. Zhou, J. Qie, Y. Yao, X. Hu, Y. Zhang, X. Xiao, Y. Yu, and J. Dong, “A continuously tunable sub-gigahertz microwave photonic bandpass filter based on an ultra-high-Q silicon microring resonator,” J. Lightwave Technol. 36(19), 4312–4318 (2018). [CrossRef]
13. L. Zhang, S. Hong, Y. Wang, H. Yan, Y. Xie, T. N. Chen, M. Zhang, Z. Yu, Y. Shi, L. Liu, and D. Dai, “Ultralow-loss silicon photonics beyond the singlemode regime,” Laser Photon. Rev. 16(4), 2100292 (2022). [CrossRef]
14. M. Pagani, B. Morrison, Y. Zhang, A. Casas-Bedoya, T. Aalto, M. Harjanne, M. Kapulainen, B. J. Eggleton, and D. Marpaung, “Low-error and broadband microwave frequency measurement in a silicon chip,” Optica 2(8), 751–756 (2015). [CrossRef]
15. M. Burla, X. Wang, M. Li, L. Chrostowski, and J. Azaña, “Wideband dynamic microwave frequency identification system using a low-power ultracompact silicon photonic chip,” Nat. Commun. 7(1), 13004 (2016). [CrossRef]
16. H. Jiang, D. Marpaung, M. Pagani, K. Vu, D. Y. Choi, S. J. Madden, L. Yan, and B. J. Eggleton, “Wide-range high-precision multiple microwave frequency measurement using a chip-based photonic Brillouin filter,” Optica 3(1), 30–34 (2016). [CrossRef]
17. X. Wang, F. Zhou, D. Gao, Y. Wei, X. Xiao, S. Yu, J. Dong, and X. Zhang, “Wideband adaptive microwave frequency identification using an integrated silicon photonic scanning filter,” Photonics Res. 7(2), 172–181 (2019). [CrossRef]
18. W. Jiao, M. Cheng, K. Wang, and J. Sun, “Demonstration of photonic-assisted microwave frequency measurement using a notch filter on silicon chip,” J. Lightwave Technol. 39(21), 6786–6795 (2021). [CrossRef]
19. Y. Tao, F. Yang, Z. Tao, L. Chang, H. Shu, M. Jin, Y. Zhou, Z. Ge, and X. Wang, “Fully on-Chip microwave photonic instantaneous frequency measurement system,” Laser Photon. Rev. 16(11), 2200158 (2022). [CrossRef]
20. Y. Yao, Y. Zhao, Y. Wei, F. Zhou, D. Chen, Y. Zhang, X. Xiao, M. Li, J. Dong, S. Yu, and X. Zhang, “Highly integrated dual-modality microwave frequency identification system,” Laser Photon. Rev. 16(10), 2200006 (2022). [CrossRef]
21. M. Burla, L. R. Cortés, M. Li, X. Wang, L. Chrostowski, and J. Azaña, “On-chip programmable ultra-wideband microwave photonic phase shifter and true time delay unit,” Opt. Lett. 39(21), 6181–6184 (2014). [CrossRef]
22. J. Tang, M. Li, S. Sun, Z. Li, W. Li, and N. Zhu, “Broadband microwave photonic phase shifter based on a feedback-coupled microring resonator with small radio frequency power variations,” Opt. Lett. 41(20), 4609–4612 (2016). [CrossRef]
23. C. Porzi, G. Serafino, M. Sans, F. Falconi, V. Sorianello, S. Pinna, J. E. Mitchell, M. Romagnoli, A. Bogoni, and P. Ghelfi, “Photonic integrated microwave phase shifter up to the mm-wave band with fast response time in silicon-on-insulator technology,” J. Lightwave Technol. 36(19), 4494–4500 (2018). [CrossRef]
24. S. X. Chew, D. Huang, L. Li, S. Song, M. A. Tran, X. Yi, and J. E. Bowers, “Integrated microwave photonic phase shifter with full tunable phase shifting range (> 360°) and RF power equalization,” Opt. Express 27(10), 14798–14808 (2019). [CrossRef]
25. L. McKay, M. Merklein, A. C. Bedoya, A. Choudhary, M. Jenkins, C. Middleton, A. Cramer, J. Devenport, A. Klee, R. DeSalvo, and B. J. Eggleton, “Brillouin-based phase shifter in a silicon waveguide,” Optica 6(7), 907–913 (2019). [CrossRef]
26. D. Lin, X. Xu, P. Zheng, G. Hu, B. Yun, and Y. Cui, “A high-performance microwave photonic phase shifter based on cascaded silicon nitride microrings,” IEEE Photon. Technol. Lett. 32(19), 1265–1268 (2020). [CrossRef]
27. S. X. Chew, S. Song, L. Li, L. Nguyen, and X. Yi, “Inline microring resonator based microwave photonic phase shifter with self-mitigation of RF power variations,” J. Lightwave Technol. 40(2), 442–451 (2022). [CrossRef]
28. D. Marpaung, J. Yao, and J. Capmany, “Integrated microwave photonics,” Nat. Photonics 13(2), 80–90 (2019). [CrossRef]
29. L. Zhuang, C. G. Roeloffzen, M. Hoekman, K. J. Boller, and A. J. Lowery, “Programmable photonic signal processor chip for radiofrequency applications,” Optica 2(10), 854–859 (2015). [CrossRef]
30. W. Zhang and J. Yao, “A fully reconfigurable waveguide Bragg grating for programmable photonic signal processing,” Nat. Commun. 9(1), 1396 (2018). [CrossRef]
31. D. Pérez, I. Gasulla, L. Crudgington, D. J. Thomson, A. Z. Khokhar, K. Li, W. Cao, G. Z. Mashanovich, and J. Capmany, “Multipurpose silicon photonics signal processor core,” Nat. Commun. 8(1), 636 (2017). [CrossRef]
32. W. Bogaerts, P. De Heyn, T. Van Vaerenbergh, K. DeVos, S. K. Selvaraja, T. Claes, P. Dumon, P. Bienstman, D. Van Thourhout, and R. Baets, “Silicon microring resonators,” Laser & Photon. Rev. 6(1), 47–73 (2012). [CrossRef]
33. W. Zhang and J. Yao, “Photonic integrated field-programmable disk array signal processor,” Nat. Commun. 11(1), 406 (2020). [CrossRef]
34. B. C. Chatterjee, N. Sarma, and E. Oki, “Routing and spectrum allocation in elastic optical networks: A tutorial,” IEEE Commun. Surv. Tutorials 17(3), 1776–1800 (2015). [CrossRef]
35. Y. Chen, S. Liu, and S. Pan, “Multi-format signal generation using a frequency-tunable optoelectronic oscillator,” Opt. Express 26(3), 3404–3420 (2018). [CrossRef]
36. W. Zhang and J. Yao, “Silicon-based single-mode on-chip ultracompact microdisk resonators with standard silicon photonics foundry process,” J. Lightwave Technol. 35(20), 4418–4424 (2017). [CrossRef]
37. L. Wang, T. Hao, M. Guan, G. Li, M. Li, N. Zhu, and W. Li, “Compact multi-tone microwave photonic frequency measurement based on a single modulator and frequency-to-time mapping,” J. Lightwave Technol. 40(19), 1–6 (2022). [CrossRef]
38. P.-L. Liu, K. J. Williams, M. Y. Frankel, and R. D. Esman, “Saturation characteristics of fast photodetectors,” IEEE Trans. Microwave Theory Tech. 47(7), 1297–1303 (1999). [CrossRef]
39. M. Pu, L. Liu, W. Xue, Y. Ding, H. Ou, K. Yvind, and J. M. Hvam, “Widely tunable microwave phase shifter based on silicon-on-insulator dual-microring resonator,” Opt. Express 18(6), 6172–6182 (2010). [CrossRef]
40. T. J. Johnson, M. Borselli, and O. Painter, “Self-induced optical modulation of the transmission through a high-Q silicon microdisk resonator,” Opt. Express 14(2), 817–831 (2006). [CrossRef]
