Open Access
Add to BibSonomy
Abstract
Photonic-assisted microwave frequency identification has been extensively studied in civil and defense applications due to its distinct features including wide frequency coverage, large instantaneous bandwidth, high frequency resolution, and immunity to electromagnetic interference. In this paper, we propose and experimentally demonstrate an approach for high-resolution frequency identification of wideband microwave signals by linearly mapping the microwave frequencies to the time delays of the optical pulses. In the proposed system, an ultrahigh-Q hybrid optical filter is a key component, which consists of a fiber ring resonator (FRR) and a silicon photonic racetrack micro-ring resonator (MRR). The FRR has an ultra-narrow bandwidth of 7.6 MHz and a small free spectral range (FSR) of 292.5 MHz, while the MRR has a bandwidth of 167.5 MHz and a large FSR of 73.8 GHz. By precisely matching the resonance wavelengths of the FRR and the MRR, a hybrid optical filter with an ultrahigh Q-factor and a large FSR is realized, which is much preferred to realizing a high resolution and a wide measurement range for microwave frequency identification. An experiment is performed and different types of microwave signals are identified. A frequency measurement range as broad as 33 GHz from 2 to 35 GHz, a frequency resolution as high as 15 MHz and a measurement accuracy as high as 5.6 MHz are experimentally demonstrated. The proposed frequency identification system holds great advantages including high frequency resolution, high measurement accuracy, and wide frequency coverage, which can find extensive applications in next-generation electronic warfare and cognitive radio systems.
© 2023 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
Fast and accurate frequency identification of wideband microwave or millimeter-wave signals is highly desirable in modern electronic systems, such as electronic countermeasure, radar warning and cognitive radio systems [1–3]. As these electronic systems are driven to have a higher operation frequency and wider bandwidth, the requirements for microwave frequency identification techniques are also pushed to an unprecedented level in terms of measurement bandwidth, frequency resolution, and the capability of measuring multiple frequencies [4]. Conventionally, microwave frequency identification is realized using electronic instruments, such as an electrical spectrum analyzer or an oscilloscope. These electronic-based techniques can achieve a high frequency resolution but suffer from a limited frequency range, low speed, and high susceptibility to electromagnetic interference (EMI).
To overcome these limitations, modern photonic technology is studied, and various photonic-assisted microwave frequency measurement techniques have been reported [5–14]. Generally, photonic-assisted microwave frequency measurement can be realized based on frequency-to-power mapping (FTPM) or frequency-to-time mapping (FTTM). For the FTPM technique, an amplitude comparison function (ACF) is usually constructed with the use of a complementary filter pair [15–19], an optical mixing unit [20,21], or a dispersive delay element [22–26]. By real-time monitoring the optical power variation, the frequency of the microwave signal can be identified. The key advantage of the FTPM technique is that an ultrahigh measurement speed is enabled, while its key disadvantage is that this technique can only be used in the scenario of a single-tone (ST) frequency identification. In some scenarios where wideband and complex microwave signals are needed to be identified, the FTPM technique is inapplicable.
To address this problem, various FTTM-based microwave frequency measurement techniques have been proposed, in which the microwave frequencies are linearly mapped to the time delays of the optical pulses. After photodetection, the optical pulses are converted to electrical pulses, and the microwave frequency can be extracted by measuring the time delays of the electrical pulses. To implement a FTTM system, a narrowband tunable optical filter is usually required [27–30], and a typical frequency resolution of 60 MHz has been demonstrated. The frequency resolution of the measurement system depends on the Q-factor of the optical filter. To further enhance the frequency resolution, it is required to increase the Q-factor of the optical filter, which is highly challenging. Moreover, for an optical filter with a periodic frequency response, there is a trade-off between its Q-factor and free spectral range (FSR). Generally, with the Q-factor increased, the FSR is decreased, which consequently limits the frequency measurement range of the identification system. In recent years, on-chip photonic-assisted microwave frequency measurement has been emerging with the advances of integrated photonic circuit (PIC), which enables a dramatic reduction in the system size, weight and power consumption (SWaP) [31,32]. Recently, a wideband adaptive microwave frequency identification technique has been proposed based on FTTM with the use of a silicon-on-insulator (SOI) micro-ring resonator (MRR) filter [33–35]. In addition to frequency identification, this technique can also identify the types of unknown microwave signals. However, the main problem is that a low frequency resolution of 375 MHz is resulted due to the limited Q-factor of the MRR filter. Due to the waveguide propagation loss caused by inevitable fabrication imperfections, it is extremely difficult to fabricate a silicon-based MRR with a 3-dB bandwidth in a scale of several MHz.
In this paper, we propose and experimentally demonstrate an approach for frequency identification of wideband microwave signals with a high frequency resolution based on an ultrahigh-Q hybrid optical filter. The work in this paper is an extension of our earlier work reported in [36]. Here, a more detailed study, including the measurement range, frequency resolution and measurement accuracy, is performed. In the proposed system, frequency identification is performed by linearly mapping the microwave frequencies to the time delays of the optical pulses with the use of the ultrahigh-Q hybrid optical filter consisting of a fiber ring resonator (FRR) and a silicon photonic racetrack MRR. The FRR has an ultra-narrow bandwidth of 7.6 MHz and a small FSR of 292.5 MHz, while the MRR has a bandwidth of 167.5 MHz and a large FSR of 73.8 GHz. By precisely matching the resonance peaks of the FRR and the MRR, a hybrid optical filter with an ultrahigh Q-factor and a large FSR is realized, which is of highly benefit to realizing a high frequency resolution and a wide measurement range for microwave frequency identification. An experiment is performed, and different types of microwave signals, including single-tone, multi-tone, chirped-frequency, and frequency-hopping signals, are identified. A frequency measurement range as broad as 33 GHz from 2 to 35 GHz, a frequency resolution as high as 15 MHz and a measurement accuracy as high as 5.6 MHz are experimentally demonstrated. The proposed frequency identification system holds great advantages including high frequency resolution, high measurement accuracy, and wide frequency coverage, which is potential to be widely used in next-generation electronic warfare and cognitive radio systems.
2. Principle
Figure 1(a) illustrates the perspective view of the proposed wideband microwave frequency identification system. A continuous-wave (CW) light wave generated by a tunable laser source (TLS) is launched into a Mach-Zehnder modulator (MZM) via a polarization controller (PC1), which is used to align the polarization state of the input light wave into the principal axis of the modulator. The incoming microwave signals to be measured are modulated on the optical carrier with the use of the MZM. At the output of the MZM, the modulated optical signal is launched into an ultrahigh-Q hybrid optical filter via the PC2. The hybrid optical filter is a key component, which determines the frequency resolution of the system. To enable a high resolution and large frequency range of the identification system, the optical filter is realized by cascading a FRR with a silicon photonic racetrack MRR. The FRR has an ultrahigh Q-factor but a small FSR, while the MRR has a large FSR but a limited Q-factor. As shown in Fig. 1(b), by precisely matching the resonance peaks of the FRR and the MRR, a hybrid optical filter with an ultrahigh Q-factor and a large FSR is realized, which entirely combines the distinct advantages of the FRR and on-chip MRR. After passing through the hybrid filter, the output optical signal is amplified by an erbium-doped fiber amplifier (EDFA), and then an optical bandpass filter (OBPF) is followed to remove the amplified spontaneous emission (ASE) noise. A photodetector (PD) is used for optical-to-electrical conversion. The recovered electrical signal is collected by an analog-to-digital converter (ADC). Due to the ultrahigh Q-factor and large FSR of the proposed hybrid optical filter, the proposed system can perform a high resolution and a wide measurement range for microwave frequency identification, which is highly preferred in the electronic warfare systems.
Fig. 1. (a) Schematic of the proposed microwave frequency identification system. TLS: tunable laser source; AFG: arbitrary function generator; PC: polarization controller; MZM: Mach-Zehnder modulator; FRR: fiber ring resonator; MRR: micro-ring resonator; EDFA: erbium-doped fiber amplifier; OBPF: optical bandpass filter; PD: photodetector; ADC: analog-to-digital converter; (b) Schematic of the ultrahigh-Q hybrid optical filter; (c) Linear mapping relationship between the microwave frequency and the time delay of the electrical signal.
Figure 1(c) illustrates the principle of the proposed microwave frequency identification system, which is realized by linearly mapping the microwave frequencies to different time delays. With the use of an MZM, the incoming microwave signal is modulated on the optical carrier, and by biasing the quadrature point an optical carrier and two 1st-order sidebands are generated. When a periodic driving signal generated by an arbitrary function generator (AFG) is applied to the TLS, the optical carrier is driven to have a wavelength sweep, and so are the sidebands. Since the center wavelength of the hybrid optical filter is fixed, during the wavelength sweeping, when the wavelength of the optical signal is aligned to that of the hybrid optical filter, at the output of the PD there is an electrical pulse generated. Consequently, a linear mapping between the frequency of the microwave signal to be measured and the time delay of the generated electrical pulse is created. At the output of the PD, there are two corresponding electrical pulses generated in time sequence. Since the initial wavelength of the optical carrier and wavelength sweep rate is known, by measuring the time interval between these two pulses, the frequency of the unknown microwave signal can be identified. In this way the effect of the initial wavelength fluctuation of the TLS and the center wavelength variation of the hybrid optical filter can be suppressed, and the measurement accuracy of the proposed system can be greatly improved. As illustrated in Fig. 1(c), when the frequency of the microwave signal to be measured exceeds half of the FSR of the hybrid optical filter, two electrical pulses corresponding to the two 1st-order sidebands will be generated, which leads to a large measurement error. Therefore, the existence of the −1st-order sideband limits the measurement range of the system to half of the FSR of the hybrid optical filter. Therefore, the OBPF is also used to remove the −1st-order sideband, which extends the measurement range to the whole FSR of the hybrid optical filter.
Mathematically, the frequency ${f_{RF}}$ of the microwave signal to be measured can be expressed as:
where B and T are the wavelength sweeping span of the TLS and the period of the driving signal applied to the TLS, and $\Delta t$ is the measured time interval between the generated two pulses. As can be seen, the frequency of the microwave signal to be measured is linearly mapped to the time delay of the electrical signal.The measurement range of the proposed microwave frequency identification system is determined by the FSR of the hybrid optical filter under the assumption that the MZM supports a bandwidth larger than this FSR. As can be seen in Fig. 1(c), due to the periodic feature of the FSR of the hybrid optical filter, a measurement error will be caused if the frequency of a microwave signal to be measured exceeds this FSR. The frequency measurement resolution of the proposed system depends on the bandwidth of the hybrid optical filter. The optical signal selected by the hybrid filter corresponds to an electrical pulse generated at the output of the PD. The frequency to be measured is linearly mapped to a specific time point to be located of the generated pulse peak. The narrower the bandwidth of the hybrid optical filter, the finer the time location of the generated electrical pulse and the higher frequency resolution of the proposed system. Thanks to its ultrahigh Q-factor and large FSR of the hybrid optical filter, a frequency identification system with a high frequency resolution and a wide frequency measurement range is realized.
3. Experimental demonstrations
An experiment is performed to demonstrate the proposed microwave frequency identification system. Figure 2 shows the experimental setup of the proposed microwave frequency identification system. A CW optical carrier with a power of 8 dBm is generated by a TLS and then sent to an MZM with a modulation bandwidth of 40 GHz via the PC1. To have a linear wavelength sweeping, a periodic sawtooth waveform generated by an AFG is applied to the TLS. The period and amplitude of the sawtooth waveform are set to be 1 ms and 1 V, respectively. The resulted wavelength sweeping range of the TLS is 0.29 nm from 1553.41 to 1553.70 nm, which corresponds to a frequency measurement range of 36.25 GHz. The microwave signal to be identified is generated by a wideband arbitrary waveform generator (AWG) and modulated on the optical carrier with the use of the MZM. The output modulated optical signal is launched into the ultrahigh-Q hybrid optical filter via the PC2. The hybrid optical filter is realized with an FRR and a silicon photonic racetrack MRR, both of which have a symmetrical add-drop configuration. The FRR is constructed using two 99:1 optical couplers (OCs), and its loop length is 68 cm. Thanks to the ultra-low optical propagation loss in the fiber, the FRR has an ultra-high Q-factor. The optical signal from the drop port of the FRR is launched into the silicon photonic racetrack MRR. The MRR is fabricated using the standard foundry muti-project wafer (MPW) process on a SOI wafer with a 220-nm-thick top layer and a 2-µm-thick buried oxide layer. The foundry name and process name are Chongqing United Microelectronics Center (CUMEC) and CSiP180Al, respectively. To improve the Q-factor of the racetrack MRR, multimode waveguides are employed to guide the optical signal with a length of 380 µm and a width of 2 µm, which are of benefit to reducing the optical propagation loss induced by sidewall roughness. To ensure single mode operation in the MRR, single-mode ridge waveguides with a width of 500 nm are used in the bending regions, and adiabatic tapers with a length of 50 µm are leveraged for mode conversion between the multi-mode waveguide and single mode wire waveguide. The gap between the bus waveguide and the MRR is designed to 460 nm. A metallic micro-heater is placed on top of the multimode waveguide to tune the resonance wavelength of the MRR based on the thermal-optic effect on silicon. At the drop port of the MRR, an EDFA is used to compensate the power loss and a tunable OBPF is followed to remove the amplified spontaneous emission (ASE) noise. The optical signal at the output of the OBPF is fed into a PD. The electrical signal generated by the PD is recorded using an ADC with a sampling rate of 25 MSa/s and a resolution of 12 bit.
In the proposed system, the ultrahigh-Q hybrid optical filter is a key component, which mainly consists of a FRR and a silicon photonic MRR. Figure 3(a) shows the measured magnitude responses of the FRR. As can be seen, the FRR has an ultra-narrow passband of 7.6 MHz and a small FSR of 292.5 MHz due to the long loop length. Figure 3(b) shows the measured transmission spectrum of the MRR. The MRR has a passband of 167.5 MHz and a large FSR of 73.8 GHz. To achieve a hybrid optical filter with an ultrahigh Q-factor and a large FSR, the resonance peaks of the FRR and the MRR are precisely matched by controlling the electrical power applied to the microheater of the MRR. Figure 3(c) shows the measured magnitude response of the hybrid optical filter and the inset gives the zoom-in view of the resonance peak. To clearly observe the transmission response of the hybrid optical filter, only one resonance peak is given in Fig. 3(c). The bandwidth of the resonance peak is measured to be 7.8 MHz, which corresponds to an ultrahigh Q-factor of 2.48 × 107. The FSR of the hybrid optical filter is determined by that of the silicon photonic MRR, which is measured to be 73.8 GHz. The measured transmission spectrum of the silicon photonic MRR and the measured magnitude responses of the hybrid optical filter show a slight asymmetry, which is caused by the nonlinear effect excited in the measurement due to the ultra-high Q-factor. By cascading the FRR and on-chip MRR, the hybrid approach entirely combines their distinct advantages, and a high frequency resolution and wide frequency measurement range can be simultaneously achieved by the proposed system.
Fig. 3. Measured normalized magnitude response of (a) the FRR and (b) the MRR; (c) Measured normalized magnitude response of the hybrid optical filter.
With the use of the hybrid optical filter, an experimental setup of the microwave frequency identification system is established. First, a single-tone (ST) signal measurement test is performed. An input ST microwave signal whose frequency is tuned from 2 to 35 GHz with a step of 3 GHz is sent into the measurement system. Figure 4(a) shows the measured temporal profiles of the generated electrical pulses. As can be seen, different frequencies of the input microwave signals have different time delays, and a linear mapping between the frequency and time is realized. Using Eq. (1), based on the measured time interval, the frequency of the microwave signal can be estimated. Figure 4(b) shows the estimated microwave frequencies in the red circle and the measurement errors in the blue diamond when the input microwave frequency is tuned from 2 to 35 GHz with a step of 0.375 GHz. As can be seen, the estimated microwave frequencies match the input microwave ones and the measurement frequency accuracy has a root-mean-square (RMS) error as small as 5.6 MHz. Based on the proposed system, a broad frequency measurement range of 33 GHz from 2 to 35 GHz is demonstrated, thanks to the large FSR of the hybrid optical filter. Theoretically, the measurement range of the proposed system is determined by the FSR of the hybrid optical filter. In the experiment, the frequency measurement range is limited by the working bandwidth of the MZM and the RF cables. In the future work, a wideband MZM and RF cables with a larger bandwidth can be employed to extend the frequency measurement range further. In the proposed system, the measurement time is determined by the period of the applied sawtooth waveform, which is set to be 1 ms. The measurement speed of the system can be improved by decreasing the sweeping period of the TLS, which can be realized based on external electro-optic modulation [37]. When the measurement speed of the proposed system is improved, real-time spectrum analysis of wideband microwave signals can be performed.
Fig. 4. (a) Measured temporal profiles of the generated electrical signals when the input microwave frequency is changed from 2 to 35 GHz with a step of 3 GHz; (b) Estimated microwave frequencies and measurement frequency accuracy when the input microwave frequency is changed from 2 to 35 GHz with a step of 0.375 GHz.
Then, a two-tone microwave signal with the frequencies of 18.00 GHz and 18.24 GHz is injected into the system for frequency identification. Figure 5(a) shows the measured temporal profiles of the generated electrical signal at the output of the PD in the blue line. As can be seen, two evidently separated peaks can be found, which again verifies that different frequencies are mapped to different time delays. For comparison, the hybrid optical filter is replaced by the same on-chip MRR, and the measured temporal profile is also given in the red line in Fig. 5(a). Since the MRR has a bandwidth much larger than that of the hybrid optical filter, the two peaks in the generated electrical signal are severely overlapped in the time domain, which would degrade the frequency resolution of the identification system. To further verify the frequency resolution of the identification system, a two-tone microwave signal with a much smaller frequency difference is injected. Figure 5(b) shows the measured temporal profiles of the generated electrical signal when a two-tone microwave signal with the frequencies of 18 GHz and 18.02 GHz is used. As shown in the figure, two separated peaks can be seen, which verifies the effectiveness of the identification system. Figure 5(c) shows the measured temporal profiles of the generated electrical signal when a two-tone microwave signal with the frequencies of 18 GHz and 18.015 GHz is used, which shows that a frequency measurement resolution better than 15 MHz is experimentally demonstrated. It is worth noting that the realized frequency resolution is slightly larger than the bandwidth of the hybrid optical filter, which may be caused by the interference between the two frequency tones when the frequency interval of the two-tone signal is close to the bandwidth of the optical filter [38].
Fig. 5. (a) Measured temporal profiles of the generated electrical signal when a two-tone microwave signal with the frequencies of (a) 18 GHz and 18.24 GHz, (b) 18 GHz and 18.02 GHz, (c) 18 GHz and 18.015 GHz is used.
To verify the capability of multiple frequency measurement over a wide range, a multi-tone (MT) microwave signal with different frequency intervals in a frequency span from 2 to 35 GHz is injected into the system. Figure 6(a) shows the measured temporal profiles of the generated electrical signal when the frequency interval of the input MT signal is set to be 6.6 GHz, in which six separated electrical peaks corresponding to six frequency components can be observed. Figure 6(b) shows the estimated microwave frequencies in the red circle and the measurement errors in the blue diamond. As can be seen, the estimated microwave frequencies match the input microwave ones and the frequency estimation errors have an RMS value as small as 6.7 MHz. Figure 6(c) shows the measured temporal profiles of the generated electrical signal when an eleven-tone microwave signal with a frequency interval of 3.3 GHz is injected, in which eleven separated electrical peaks can be seen. Figure 6(d) shows the estimated microwave frequencies in the red circle and the measurement errors in the blue diamond. As can be seen, the estimated microwave frequencies again match the input microwave ones and the frequency estimation errors have an RMS value of 6.1 MHz. Figure 6(e) shows the measured temporal profiles of the generated electrical signal when a 31-tone microwave signal with a frequency interval of 1.1 GHz is injected, in which 31 separated electrical peaks can be seen. Figure 6(f) shows the estimated microwave frequencies in the red circle and the measurement errors in the blue diamond. As can be seen, the estimated microwave frequencies again match the input microwave ones and the frequency estimation errors have an RMS value of 4.6 MHz. In the measurement of MT microwave signals, the decreased amplitude of the generated electrical peaks at high input frequency is caused by the limited bandwidth of the MZM and the RF cables. In the future work, a wideband MZM and RF cables with a larger bandwidth will be adopted to suppress the unevenness and extend the frequency measurement range further. The experimental results show that the proposed microwave frequency identification system has a strong capability of measuring MT signal with a wide measurement range and a high frequency accuracy.
Fig. 6. (a) Measured temporal profiles of the generated electrical signal and (b) Estimated microwave frequencies and the measurement errors when a six-tone microwave signal with a frequency interval of 6.6 GHz is injected; (c) Measured temporal profiles of the generated electrical signal and (d) Estimated microwave frequencies and the measurement errors when a eleven-tone microwave signal with a frequency interval of 3.3 GHz is injected; (e) Measured temporal profiles of the generated electrical signal and (f) Estimated microwave frequencies and the measurement errors when a 31-tone microwave signal with a frequency interval of 1.1 GHz is injected.
Since wideband chirped-frequency (CF) microwave signals are extensively used in the modern radar systems, a proof-of-concept experiment to identify a wideband chirped microwave signal is performed. In the experiment, CF microwave signals with different bandwidths and center frequencies are injected into the system. Figures 7(a), 7(b), and 7(c) show the measurement results of input CF microwave signal with frequencies ranging from 8 to 20 GHz, 5 to 23 GHz, and 2 to 26 GHz, respectively. With the use of the proposed ultrahigh-Q hybrid optical filter, the spectra of the input CF microwave signals are linearly mapped to the temporal waveform of the electrical signals. The measured duration time of the electrical waveforms are 148.89-370.643 µs, 92.908-425.534 µs, and 37.695-481.201 µs, corresponding to microwave frequencies of 8.025-19.979 GHz, 5.007-22.939 GHz, and 2.031-25.94 GHz, which match well with the theoretical values. Figures 7(d), 7(e), and 7(f) show the measurement results of the input CF microwave signal with frequencies ranging from 9 to 15 GHz, 15 to 21 GHz, and 21 to 27 GHz, respectively. The measured duration time of the electrical waveforms are 167.452-278.331 µs, 278.702-389.377 µs, and 390.165-500.84 µs, corresponding to microwave frequencies of 9.026-15.003 GHz, 15.023-20.99 GHz, and 21.032-26.998 GHz, respectively.
Fig. 7. Measured temporal profiles of the generated electrical signal when CF microwave signals with a center frequency of 15 GHz and different bandwidths of (a) 12 GHz, (b) 18 GHz and (c) 24 GHz, respectively; Measurement results of CF microwave signals with a bandwidth of 6 GHz and different center frequencies of (d) 12 GHz, (e) 18 GHz and (f) 24 GHz, respectively.
Table 1 gives the measurement results of the CF microwave signals. It can be seen that the measured microwave frequency matches well the input frequency. To quantitatively analyze the measurement accuracy, the parameters of bandwidth error ${\sigma _{band}}$ and center frequency error ${\sigma _{center}}$ are defined as [34]
Table 1. Measurement Results of CF Microwave Signals
In the experiment, a demonstration of frequency-hopping (FH) microwave signal measurement is also performed. FH microwave signals with different frequency spans and steps are launched into the system. Figure 8(a) shows the measurement result of the input FH microwave signal with a frequency range from 17 to 19 GHz and a frequency step of 0.5 GHz. As can be seen, 5 separated electrical peaks corresponding to 5 frequency levels are generated. Figure 8(b) shows the linear mapping relationship between the input frequency and the measured frequency. The RMS value of the measurement errors is 3.8 MHz. Figure 8(c) shows the measurement result when the frequency step of the input FH signal is reduced to 0.25 GHz, in which 9 separated electrical peaks can be observed. Figure 8(d) shows the linear mapping relationship between the input frequency and the measurement frequency. The RMS value of the measurement errors is 8.2 MHz. To verify the capability of measuring wideband FH microwave signal, a FH signal with a frequency range from 13 to 23 GHz and a frequency step of 0.5 GHz is injected into the system. Figure 8(e) shows the measurement result, in which 21 separated electrical peaks corresponding to 21 frequency levels can be observed. Figure 8(f) shows the linear mapping relationship between the input frequency and the measurement frequency. The RMS value of the measurement errors is 8.6 MHz. The experimental results verify that the proposed system is capable of measuring wideband FH microwave signals with a high accuracy.
Fig. 8. (a) Measured temporal profiles of the generated electrical signal when a two-tone microwave signal with the frequencies of (a) 18 GHz and 18.24 GHz, (b) 18 GHz and 18.02 GHz, (c) 18 GHz and 18.015 GHz is used. Measurement results of FH microwave signals from 17 to 19 GHz stepped by 0.5 GHz; (c) Measurement results of FH microwave signals from 17 to 19 GHz stepped by 0.25 GHz; (e) Measurement results of FH microwave signals from 13 to 23 GHz stepped by 0.5 GHz; (b), (d) and (f) show the measurement errors of FH microwave signals.
As described in the above demonstrations, the frequency of four types of microwave signals including ST signal, MT signal, FH signal, and CF signal have been identified with a high frequency resolution and accuracy. Actually, the proposed system is capable of identifying more complex microwave signals, such as a combination of multitype microwave signals. To verify this powerful function, frequency identification of different combinations of three-type microwave signals have been demonstrated. The parameters of the complex microwave signals can be divided into three groups: CF (5-10 GHz), MT (13-18 GHz, stepped by 1 GHz), and FH (21-26 GHz, stepped by 1 GHz); MT (5-10 GHz, stepped by 1 GHz), CF (13-18 GHz), and FH (21-26 GHz, stepped by 1 GHz); MT (5-10 GHz, stepped by 1 GHz), FH (13-18 GHz, stepped by 1 GHz), and CF (21-26 GHz). The measurement results of these three complex microwave signals are shown in Figs. 9(a), 9(b), and 9(c), respectively. Table 2 shows the measured frequencies of different groups of complex microwave signals. In the above measurement results, the maximum RMS errors for MT and FH signals are calculated to be 3.7 MHz and 11.4 MHz, and the maximum bandwidth error ${\sigma _{band}}$ and center frequency error ${\sigma _{center}}$ for CF signals are 0.77% and 14.6 MHz, respectively. Besides, the proposed system can also be used to classify the type of the input microwave signal by exploiting the distinguishable characteristics of the measured temporal electrical waveform. In the proposed system, time-varying signals such as the CF signal and the FH signal are measured based on statistical method, therefore the measured temporal electrical waveforms are filled with random power [34]. On the contrary, the measured temporal electrical waveforms of the time-invariant signals such as the ST signal and the MT signal are hollow without being filled with random power. Figures 9(d) and 9(e) show the zoom-in views of one frequency component of the MT signal and the FH signal given in Fig. 9(c). As can be seen, the measured electrical peal of the MT signal is hollow while that of the FH signal is filled with random powers. Therefore, the proposed frequency identification system provides a power capability of measuring the frequency spectrum and classify the type of complex wideband microwave signals.
Fig. 9. Measured temporal profiles of the generated electrical signal when the combination of multitype microwave signals is injected into the system. (a) FH (21-26 GHz, stepped by 1 GHz), CF (5-10 GHz) and MT (13-18 GHz, stepped by 1 GHz); (b) FH (21-26 GHz, stepped by 1 GHz), CF (13-18 GHz) and MT (5-10 GHz, stepped by 1 GHz); (c)FH (13-18 GHz, stepped by 1 GHz), CF (21-26 GHz) and MT (5-10 GHz, stepped by 1 GHz); (d) and (e) are the zoom-in views of frequency components of the MT signal and the FH signal given in Fig. (c).
Table 2. Measurement Results Analysis of Multitype Microwave Signals
In this work, a silicon photonic integrated racetrack MRR is used to construct the hybrid optical filter, while other devices such as the TLS, MZM, OBPF, and PD are commercially available products. Fortunately, high-performance integrated TLS, MZM, OBPF, and PD have been reported in previous research works [39–42]. Therefore, the proposed system is potential to be implemented on an integrated chip by leveraging hybrid integration of different photonic platforms.
4. Conclusion
In conclusion, we proposed and experimentally demonstrated an approach for frequency identification of wideband microwave signals with a high frequency resolution based on an ultrahigh-Q hybrid optical filter. In the proposed system, the hybrid optical filter is a key component, which was realized by cascading an FRR with a silicon photonic racetrack MRR. The FRR has an ultra-narrow bandwidth of 7.6 MHz and a small FSR of 292.5 MHz, while the MRR has a bandwidth of 167.5 MHz and a large FSR of 73.8 GHz. By precisely matching the resonance peaks of the FRR and the MRR, a hybrid optical filter with an ultrahigh Q-factor and a large FSR was realized. Benefiting from the hybrid optical filter, a high resolution and a wide measurement range for microwave frequency identification can be realized. An experiment was performed, and different types of microwave signals, including ST, MT, FH, and CF signals, were identified with a high frequency resolution. A frequency measurement range of 2-35 GHz, a frequency resolution as high as 15 MHz and a measurement accuracy as high as 5.6 MHz were experimentally demonstrated. The proposed frequency identification system holds great advantages including high frequency resolution, high measurement accuracy, and wide frequency coverage, which is potential to be widely used in next-generation electronic warfare and cognitive radio systems.
Funding
National Key Research and Development Program of China (2018YFE0201800); National Natural Science Foundation of China (62005018, 62071042, U22A2018); China Postdoctoral Science Foundation (2020M680381).
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. A. E. Spezi, “Electronic warfare systems,” IEEE Trans. Microw. Theory Tech. 50(3), 633–644 (2002). [CrossRef]
2. P. W. East, “Fifty years of instantaneous frequency measurement,” IET Radar Sonar Navig. 6(2), 112–122 (2012). [CrossRef]
3. K. Domdouzis, B. Kumar, and C. Anumba, “Radio-frequency identification (RFID) applications: a brief introduction,” Adv. Eng. Inform. 21(4), 350–355 (2007). [CrossRef]
4. X. H. Zou, B. Lu, W. Pan, et al., “Photonics for microwave measurements,” Laser Photonics Rev. 10(5), 711–734 (2016). [CrossRef]
5. J. Yao, “Microwave photonics,” J. Lightwave Technol. 27(3), 314–335 (2009). [CrossRef]
6. J. Capmany and D. Novak, “Microwave photonics combines two worlds,” Nat. Photonics 1(6), 319–330 (2007). [CrossRef]
7. S. Pan and J. Yao, “Photonics-based broadband microwave measurement,” J. Lightwave Technol. 35(16), 3498–3513 (2017). [CrossRef]
8. L. V. Nguyen and D. B. Hunter, “A photonic technique for microwave frequency measurement,” IEEE Photonics Technol. Lett. 18(10), 1188–1190 (2006). [CrossRef]
9. M. Attygalle and D. B. Hunter, “Improved photonic technique for broadband radio-frequency measurement,” IEEE Photonics Technol. Lett. 21(4), 206–208 (2009). [CrossRef]
10. W. Li, N. H. Zhu, and L. X. Wang, “Brillouin-assisted microwave frequency measurement with adjustable measurement range and resolution,” Opt. Lett. 37(2), 166–168 (2012). [CrossRef]
11. H. Y. Jiang, D. Marpaung, M. Pagani, et al., “Wide-range high-precision multiple microwave frequency measurement using a chip-based photonic Brillouin filter,” Optica 3(1), 30–34 (2016). [CrossRef]
12. Y. Chen, W. Zhang, J. Liu, et al., “On-chip two-step microwave frequency measurement with high accuracy and ultra-wide bandwidth using add-drop micro-disk resonators,” Opt. Lett. 44(10), 2402–2405 (2019). [CrossRef]
13. Y. Ma, D. Liang, D. Peng, et al., “Broadband high-resolution microwave frequency measurement based on low-speed photonic analog-to-digital converters,” Opt. Express 25(3), 2355–2368 (2017). [CrossRef]
14. H. Wang, S. Zhang, X. Zou, et al., “Photonic microwave frequency measurement based on frequency-configurable pilot tones,” IEEE Photonics Technol. Lett. 30(4), 363–366 (2018). [CrossRef]
15. H. Chi, X. Zou, and J. Yao, “An approach to the measurement of microwave frequency based on optical power monitoring,” IEEE Photonics Technol. Lett. 20(14), 1249–1251 (2008). [CrossRef]
16. X. Zou, H. Chi, and J. Yao, “Microwave frequency measurement based on optical power monitoring using a complementary optical filter pair,” IEEE Trans. Microw. Theory Tech. 57(2), 505–511 (2009). [CrossRef]
17. B. Zhu, W. Zhang, S. Pan, et al., “High-Sensitivity Instantaneous Microwave Frequency Measurement Based on a Silicon Photonic Integrated Fano Resonator,” J. Lightwave Technol. 37(11), 2527–2533 (2019). [CrossRef]
18. H. Emami and M. Ashourian, “Improved dynamic range microwave photonic instantaneous frequency measurement based on four-wave mixing,” IEEE Trans. Micro. Theory Tech. 62(10), 2462–2470 (2014). [CrossRef]
19. S. Pan and J. Yao, “Instantaneous microwave frequency measurement using a photonic microwave filter pair,” IEEE Photonics Technol. Lett. 22(19), 1437–1439 (2010). [CrossRef]
20. M. Pagani, B. Morrison, Y. Zhang, et al., “Low-error and broadband microwave frequency measurement in a silicon chip,” Optica 2(8), 751–756 (2015). [CrossRef]
21. M. Burla, X. Wang, M. Li, et al., “Wideband dynamic microwave frequency identification system using a low-power ultracompact silicon photonic chip,” Nat. Commun. 7(1), 13004 (2016). [CrossRef]
22. J. Zhou, S. Fu, S. Aditya, et al., “Instantaneous microwave frequency measurement using photonic technique,” IEEE Photonics Technol. Lett. 21(15), 1069–1071 (2009). [CrossRef]
23. J. Li, S. Fu, K. Xu, et al., “Photonic-assisted microwave frequency measurement with higher resolution and tunable range,” Opt. Lett. 34(6), 743–745 (2009). [CrossRef]
24. X. Zou, S. Pan, and J. Yao, “Instantaneous microwave frequency measurement with improved measurement range and resolution based on simultaneous phase modulation and intensity modulation,” J. Lightwave Technol. 27(23), 5314–5320 (2009). [CrossRef]
25. X. Zhang, H. Chi, X. Zhang, et al., “Instantaneous microwave frequency measurement using an optical phase modulator,” IEEE Microw. Wireless Compon. Lett. 19(6), 422–424 (2009). [CrossRef]
26. W. Li, N. H. Zhu, and L. X. Wang, “Reconfigurable instantaneous frequency measurement system based on dual-parallel Mach-Zehnder modulator,” IEEE Photonics J. 4(2), 427–436 (2012). [CrossRef]
27. S. Zheng, S. Ge, X. Zhang, et al., “High-resolution multiple microwave frequency measurement based on stimulated Brillouin scattering,” IEEE Photonics Technol. Lett. 24(13), 1115–1117 (2012). [CrossRef]
28. T. A. Nguyen, E. H. W. Chan, and R. A. Minasian, “Instantaneous high-resolution multi-tone measurement system based on frequency-to-time mapping technique,” Opt. Lett. 39(8), 2419–2422 (2014). [CrossRef]
29. T. Hao, J. Tang, W. Li, et al., “Microwave photonics frequency-to-time mapping based on a fourier domain mode locked optoelectronic oscillator,” Opt. Express 26(26), 33582–33591 (2018). [CrossRef]
30. T. Hao, J. Tang, N. Shi, et al., “Multi-tone measurement based on a Fourier domain mode-locked optoelectronic oscillator operating around oscillation threshold,” Opt. Lett. 44(12), 3062–3065 (2019). [CrossRef]
31. D. Marpaung, J. Yao, and J. Capmany, “Integrated microwave photonics,” Nat. Photonics 13(2), 80–90 (2019). [CrossRef]
32. D. Marpaung, “On-chip photonic-assisted instantaneous microwave frequency measurement system,” IEEE Photonics Technol. Lett. 25(9), 837–840 (2013). [CrossRef]
33. F. Zhou, H. Che, X. Wang, et al., “Photonic multiple microwave frequency measurement based on frequency-to-time mapping,” IEEE Photonics J. 10(2), 1–7 (2018). [CrossRef]
34. X. Wang, F. Zhou, D. Gao, et al., “Wideband adaptive microwave frequency identification using an integrated silicon photonic scanning filter,” Photonics Res. 7(2), 172–181 (2019). [CrossRef]
35. Y. Yao, Y. Zhao, Y. Wei, et al., “Highly integrated dual-modality microwave frequency identification system,” Laser Photonics Rev. 16(10), 2200006 (2022). [CrossRef]
36. H. Liu, X. Hong, Y. Cheng, et al., “High-resolution frequency identification of wideband microwave signal using a hybrid optical filter,” Optical Fiber Communication Conference, W2B. 24 (2023).
37. P. Zuo, D. Ma, and Y. Chen, “Short-time Fourier transform based on stimulated Brillouin scattering,” J. Lightwave Technol. 40(15), 5052–5061 (2022). [CrossRef]
38. L. Wang, T. Hao, M. Guan, et al., “Compact multi-tone microwave photonic frequency measurement based on a single modulator and frequency-to-time mapping,” J. Lightwave Technol. 40(19), 6517–6522 (2022). [CrossRef]
39. Y. Fan, A. van Rees, P. J. M. van der Slot, et al., “Hybrid integrated InP-Si3N4 diode laser with a 40-Hz intrinsic linewidth,” Opt. Express 28(15), 21713–21728 (2020). [CrossRef]
40. M. He, M. Xu, Y. Ren, et al., “High-performance hybrid silicon and lithium niobate Mach–Zehnder modulators for 100 Gbit s-1 and beyond,” Nat. Photonics 13(5), 359–364 (2019). [CrossRef]
41. S. Cui, K. Cao, Z. Pan, et al., “Compact microring resonator based on ultralow-loss multimode silicon nitride waveguide,” Adv. Photonics Nexus 2(4), 046007 (2023). [CrossRef]
42. A. Beling, X. Xie, and J. C. Campbell, “High-power, high-linearity photodiodes,” Optica 3(3), 328–338 (2016). [CrossRef]
