Open Access
Add to BibSonomy
Abstract
Extracting precise target characteristics from microwave image is needed and calls for high-resolution microwave imaging radar systems. In this paper, a Ka-band ultra-wideband microwave photonic (MWP) imaging radar is developed and experimentally demonstrated. In the transmitter, continuous ultra-wideband linear frequency modulation (LFM) wave is generated based on optical frequency sextupling technique. In the receiver, a combination of optical frequency mixer with fiber delay lines and electric analog-to-digital converter (ADC) is capable of receiving target echoes and imaging targets with different distances. The maximum instantaneous bandwidth of the transmitted waveform is measured to be 10.02 GHz and corresponding range resolution is calibrated to be 1.68 cm. Out-field tests with demonstrator working at synthetic aperture radar (SAR) or inverse synthetic aperture radar (ISAR) mode are carried out. Different targets such as an unmanned aerial vehicle (UAV), airliner and Leifeng pagoda are imaged. Based on corresponding high-resolution microwave images, quantitative information of the targets can be identified, which shows the great potential of the radar demonstrator for various remote sensing applications.
© 2018 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
Microwave imaging, for its unique ability to provide an imagery of target at very long range in all-weather and all-time conditions, has been playing more and more important roles in military and civil applications [1,2]. Realizing exact target characteristics extraction based on microwave image is the development goal of microwave imaging technique. Currently, the best high-resolution microwave imaging radar based on traditional microwave technique is the Haystack Ultrawideband Satellite Imaging Radar (HUSIR) proposed by American Lincoln laboratory, which has an 8 GHz synthesized bandwidth at W frequency band [3]. However, its high cost and large size inhibits its applications in air-based, spaceborne or unmanned aerial vehicle (UAV) platforms. As the rapid development of intelligent recognition technique and precise defense technique, extracting quantitative information of target desired by remote sensing for earth and aerospace targets detecting calls for more high-resolution microwave images. Increasing bandwidth is an effective way to improve the imaging resolution, while the traditional electric technique suffers bottleneck for generating large bandwidth radar waveforms. Microwave photonics provides a small-size and low-cost way to generate and process microwave signal with large bandwidth [4,5].
Microwave photonics has been continuously introduced to improve the performance of modern radar systems [6–13]. With fully photonics-based single signal core, dual frequency-band inverse synthetic aperture radar (ISAR) operating in S and X band was demonstrated by a research group in Italy [14–16], which showed the distinct advantage of microwave photonics for electromagnetic compatibility. Though data fusion was adapted in the above coherent dual-band radar to improve the resolution, the synthetic bandwidth was still limited to several tens of MHz for the speed of optical serial-to-parallel converter and the repetition rate of the mode-locked laser [15]. To improve the bandwidth, a different microwave photonic (MWP) radar architecture was proposed by Zou’ s group [17], in which the bandwidth of the generated and received waveforms could be tens of GHz. However, the nanosecond-level temporal period limited by the chromatic dispersion of fiber inhibited its application for long-range detection. A K-band photonics-based ISAR system was established by Fangzheng Zhang etc [18,19], in which the temporal period and bandwidth of generated waveforms were 5 μs and 8 GHz respectively. Laboratory experiment of imaging a rotary fan at 2.6 m away was demonstrated to show the fast imaging ability of the system. To show the similar field-trial ability as traditional microwave electric radar, a Ku-band MWP imaging radar with a temporal period of 500μs was presented by Li etc. and evaluated by ISAR imaging of non-cooperative passenger aircraft at 800 m away [20]. However, its bandwidth was only 600 MHz, which is still much less than HUSIR. Overall, though MWP imaging radar systems have shown various potential advantages for radar implementation, key parameters such as resolution are still no better than traditional radar under the premise of enabling long-range detection for field-trial application.
In this paper, we propose and experimentally demonstrate an ultra-wide bandwidth imaging radar based on microwave photonics. The generation of the ultra-wideband waveforms is realized by optical frequency sextupling based on single Mach-Zehnder modulator (MZM). And the reception of the ultra-wideband echo is dechirped by a photonic mixer, in which optical local oscillator signal is delayed by an optical delay line before mixing course to solve the contraction between long-range detection and high-speed analog-to-digital converter (ADC) with high effective number of bits (ENOB). A Ka-band MWP imaging radar with ultra-wideband bandwidth up to 10.02 GHz is realized. To the best of our knowledge, this is the largest bandwidth that microwave photonic imaging radar has realized. With two trihedral corner reflectors, the range resolution of the radar system is calibrated to be 1.68 cm. Based on waveforms with temporal periods of 150 μs and 1.2 ms, synthetic aperture radar (SAR) experiment for Leifeng pagoda and ISAR experiments for aircraft targets such as UAV and airliner are conducted at long ranges away. For the first time, microwave images with single-centimeter-level resolution are obtained and quantitative information of the targets is extracted benefiting from the high resolution. The bandwidth of our radar is larger than HUSIR, which indicates the surpassing of photonics-based radar over traditional electric imaging radar in resolution aspect and will enhance the development of high-resolution microwave imaging techniques.
2. Principles
Figure 1 shows the schematic of the presented MWP imaging radar, which can be divided into two parts: the transmitter (shown on blue board) and the receiver (shown on orange board). In the transmitter, the light signal generated by a laser source is modulated by a MZM1, which is biased at null point through a potentiometer to suppress the optical carrier. An electric linear frequency modulation (LFM) wave at intermediate frequency (IF) generated by a direct digital synthesizer (DDS) is applied to the radio-frequency (RF) port of the MZM1 to modulate the intensity of the light wave. The output signal of the MZM1 can be written as:
where and are the amplitude and angular frequency of the lightwave from laser; , and are the center frequency, chirp rate and temporal period of the IF signal from DDS; is the order Bessel function of the first kind, is the modulation index of the MZM1 and can be expressed , in which is the half-wave voltage of the MZM1 and is the amplitude of the IF signal from DDS. When the value of is set to satisfy, we have, which means the 1st order optical sideband is suppressed. In this case, only the ± 3rd and higher order sidebands exist at the output of MZM1. By eliminating the ± 5th and higher order sidebands using an optical band-pass filter, there will be nothing but the ± 3rd sidebands in the optical spectrum, which can be expressed as:
Fig. 1 Schematic of the MWP ultra-wideband imaging radar. MZM: Mach-Zehnder modulator; DDS: direct digital synthesizer; OF: optical filter; PD: photodetector; PA: power amplifier; TA: transmitting antenna; OTD: optical time delayer; EDFA: erbium doped fiber amplifier; LPF: low-pass filter; ADC: analog-digital converter; LNA: low-noise amplifier; RA: receiving antenna.
Then the optical signal is divided into two parts: one is sent to the radar receiver as the local oscillator signal, and the other is sent to PD1 to beat and generate the radar detection signal, which can be given by:
In the transmitter, after photoelectric conversion by the PD1, not only the center frequency but also the bandwidth are all six times to the original signal from DDS. However, the temporal time remains unchanged. Then the high-frequency ultra-wideband radar signal with temporal period equal to the signal generated by electric techniques is amplified by a power amplifier and then sent to the targets through a transmitting antenna.
In the receiver, the local oscillator signal is modulated by the incoming echoes reflected from the measuring targets through a quadrature-point-biased optical intensity modulator (MZM2). The incoming echoes is collected by a receiving antenna and amplified by a RF amplifier which is composed of a low noise amplifier and a power amplifier before applying to the MZM2. To solve the contraction between the detection range and bandwidth of ADC with high ENOB, a tunable optical time delayer (OTD) is used to delay the optical local oscillator signal. The delayed optical signal can be expressed as:
where and are the electric field intensity of optical oscillator signal and the time delay of OTD, respectively. The incoming echo can be expressed as:where represent the time delay that incoming echo experienced in free space and is the amplitude of the incoming echo. After modulating by the incoming echo, the light signal at the output of MZM2 can be written as:where represents the phase difference between two arms of MZM2 and is equal to while the MZM2 is biased at quadrature point. Let , , and assume the incoming echo is a small signal, the above equation can be expanded in terms of Bessel functions of the first kind,The optical sideband represented by first term is closed to the one represented by the fifth term in frequency domain, as well as the optical sidebands represented by the third and fourth terms. As a result, the beating frequencies of two groups of optical sidebands mentioned above are low while the others are high. After filtered by an electric low-pass filter (LPF), the electric signal can be written as:
As can be seen, the frequency of the final electric signal is proportional to the difference between time delays caused by the optical delay line and the target. Since the bandwidth of the LPF is limited, the detection can be switched through the OTD according to the delay of target so as to ensure the final signal frequency is located within the bandwidth of the LPF.
3. Results and discussions
3.1 Demonstrator and its performance
To identify the ability of the proposed MWP radar for processing ultra-wideband radar signal and high-resolution imaging, a Ka-band microwave imaging radar demonstrator is constructed, as shown in Fig. 2(a). The light source of the whole demonstrator is a continuous-wave laser (CoBrite DX4, IDphotonics), with the output wavelength and optical power are 1550 nm and 16 dBm, respectively. To reduce the requirement for the value to satisfy, a MZM with low (AX-0MSS-20-PFA-PFA-LV, EOSPACE) is adapted to modulate the light. For precisely biasing the MZM1 at null point, a customized DC voltage supplier is used. An electrical LFM signal from a DDS is sent to the RF port of MZM1, center frequency, bandwidth and tunable temporal period range of which are 5.83 GHz, 1.67GHz and 150 μs-1.2 ms, respectively. The modulated optical wave is then filtered by a customized optical filter (Terxion). When the period of the electrical LFM wave is 150 μs, the output optical spectrum of the filter is shown in Fig. 2(b), which is measured by an optical spectrum analyzer (AQ6370D, YOKOGAWA). Clearly, the two highest optical sidebands are the third-order sidebands, which is 0.28 nm (corresponding to 35 GHz in frequency) away. Other sidebands are relatively very weak, and the 1st order sideband and the carrier are 22 dB and 39 dB lower than the 3rd sidebands, respectively, which means high suppression effect for other undesired sidebands and ensures a good signal-to-noise ratio (SNR) of the emitted waveforms. A 50:50 optical coupler is used to split the optical sidebands into two ways: one goes into the receiver as optical local oscillator signal, the other stays in the transmitter to generate the LFM signal by a high-frequency PD (Finisar, XPDV 2120RA). An electrical power amplifier enhances the power of the LFM signal to be 47 dBm for target detection at long distance away. Figure 2(c) shows the spectrum of the final LFM signal measured by an electric spectrum analyzer (Keysight, N9030A). The bandwidth, power deviation and SNR of the LFM signal are identified to be 10.02GHz, ± 2.02 dB and higher than 40 dB, respectively. The time-bandwidth product (TBWP) of the generated radar waveform is 1,503,000. Results introduced above show the excellent ability of the established radar for generating ultra-wideband radio-frequency signal based on MWP techniques.
Fig. 2 The photograph of the radar demonstrator (a) and its performance test results: (b) is the optical spectrum of light output from MZM1, (c) is the spectrum of the generated LFWM signal and (d) is the resolution calibration result, in which P1 and P2 represent the reflectors.
The emission and receive of ultra-wideband RF signal are completed by two conical horn antennas, which can formulate radar beams with not only large beam width (14°) but also high main-to-sidelobe ratio. The amplifier group in the receiver consists of a low noise amplifier and an adjustable gain amplifier, the maximum gains of which are both 40 dB. The bandwidth of MZM2 is 65 GHz and its working point is controlled by a bias controller (YYLabs, mini-MBC3). Before the optical local oscillator signal enters MZM2, optical line with proper length will be configured according to the target distance, which may cause optical power fluctuation for the varied optical losses of different-length optical lines. To avoid the influence of this fluctuation on the stability of the receiver, an EDFA is adapted to amplify the power of the output signal from the optical delay line to a designated value (15 dBm). The delayed optical local oscillator signal modulated by the incoming electric echo signal beats at PD2 (Discovery, DSC50S) to generate the final dechirped signal, which is then filtered by an IF electric bandpass filter (2MHz-200MHz,BJQX,PSIPA1733001). The filtered signal is sampled and stored by a 500MSa/s 12-bit ADC.
The performance of the MWP imaging radar is tested by standard samples. The range resolution of the radar demonstrator is calibrated by two trihedral corner reflectors with a distance of 1.6 cm away from each other, and about 14 m away from the antennas in radial direction. To reduce the interference from ground clutters, the reflectors are placed on plain marble floor. Figure 2(d) shows the resolution test result, which is obtained by applying data processing of quantization, interpolation and Fourier transform to one-period echo signal from the reflectors. In the figure, the distance between two highest peaks is calculated to be 1.68 cm, which is close to the real value and the theoretical resolution of the demonstrator (1.5 cm). To our knowledge, this is the highest resolution microwave photonic imaging radar that has been realized.
3.2 Field-trial experiment results
Field trial ability of the demonstrator is firstly investigated by ISAR imaging non-cooperative moving targets. In modern society, wide application of unmanned aircrafts desires strict monitoring urgently. A small UAV (AOBO,JX1000 X6) is first chose to be imaged. Figure 3(a) shows the overhead view of the imaging experiment and Fig. 3(b) is the photograph of the used six-rotor UAV, the diameter and arm width of which are 1.2 m and 1.9 cm, respectively. The distance of the moving UAV was about 120 m away, the temporal period of the imaging pulses was 1.2 ms (corresponding TBWP is 12,024,000) and the length of optical delay line was set to be 1 km. In the experiment, the UAV flew across the radar beam in a straight line and with no self-rotation. Thus, the relative rotation angle between the UAV and the radar was 14°. According to the calculation equation mentioned in the Ref [13], an ISAR image of the UAV with a cross-range resolution of 1.75 cm might be obtained. To recover image from the incoming echoes, an improved ISAR algorithm based on MATLAB is introduced to deal with the high-resolution data. Figure 3(c) shows the ISAR image of the UAV after calibration process. Benefiting from the high resolution of the demonstrator, the rotor number of the UAV is clearly identified to be 6. And center battery (illustrated by red rectangle), two 1.8 cm-width legs (white rectangles) and two1.9 cm-width arms vertical to the radar beam direction are all easily distinguished. Discrimination of so many detail structures from such a small target shows not only the excellent range and cross-range resolutions but also the effectiveness of the radar demonstrator for monitoring low slow small targets.
Fig. 3 The diagram of the UAV ISAR experiment in overhead view (a), the photograph of the six-rotor UAV used in the experiment (b) and its ISAR image (c) obtained by the demonstrator.
Further, to demonstrate the radar imaging ability for long-range targets, ISAR imaging experiments for Boeing 737 in take-off are carried out. The distance between the demonstrator and the plane was about 1.2 km, the temporal period of the imaging pulses was150 μs and the length of the optical delay line was set to be 1 km. The data acquisition lasts as the plane crosses the radar beam. The data processing is similar to the one used in processing UAV image except that two-dimensional error correction process is employed. Figure 4(a) shows the ISAR image of the Boeing 737, and the inset is a photo of this kind plane. As can be seen, outlines of the whole plane are obvious. Its wings, tail and body can all be clearly distinguished. Compared with former ISAR images of planes obtained by other radar systems [20], images obtained by this MWP radar demonstrator show a sharp edge, which will be useful for target recognition. Figure 4(b) shows the enlarged view of the down wing in the Fig. 4(a). Figure 4(c) is the wing photo corresponding to Fig. 4(b). By comparison, one engine and three flat tracks are identified, which matches the real number. In addition, many strong scatter points arranged at the edge of wings can be seen. The inset of Fig. 4(b) shows some of them, which corresponds to the small track structures (as shown in inset in Fig. 4(c) controlling the leading edge slats when the plane takes off. The fine discrimination of this fine structure shows the excellent resolution of the proposed demonstrator and the identification of the number of flat tracks demonstrates the ability for providing quantitative information of target.
Fig. 4 Experimental imaging results of Boeing 737. (a) ISAR imaging result of the whole plane, the inset is photograph of a Boeing 737; (b) enlarged view of the down wing part in (a), the inset is the enlarged view of the rectangle in white dashed line; (c) the photograph of the wing part, the inset corresponds to the enlarged view in (b).
Beside ISAR, synthetic aperture radar (SAR) imaging is another important field of microwave imaging. Based on vehicle platform, Leifeng pagoda, which is a landmark building of Hangzhou city in China, was imaged to show the SAR imaging ability of the radar demonstrator. As shown in Fig. 5(a), the octagonal pagoda consists of five layers. When the data acquisition was carried out, the pagoda was about 230 m away, the elevation angle of the antennas was about 13°, the speed of the vehicle was 8 km/h, the temporal period of the radar pulse was 150 μs and the optical delay line was set to be 0 km. An improved SAR algorithm is introduced to recover the image, in which the adverse effect of the vehicle jolt is specially treated. The photograph of the pagoda is shown in Fig. 5(a), and corresponding SAR image of the pagoda is shown in Fig. 5(b). As can be seen, the body and spire of the pagoda are vividness in the image. The five floors of the body and the right-angle structures at the corridor roof of each floor can be recognized (shown by the four white lines). Figure 5(c) is the enlarged view of the spire part in Fig. 5(b), and Fig. 5(d) is the corresponding photo of the part. By comparing these two images, two chains are identified as indicated by the red arrows in both images. In addition, pairs of strong scatter points distribute symmetrically along the center axis, which corresponds to the layers of metal wheels in Fig. 5(d). The pair number of the scatter points can be counted to 7, which also matches well with the fact. The capturing of the chain structures exhibits the resolution of the demonstrator, and the accurate counting for the metal wheels in the spire part indicates that the radar can be used to provide quantitative information of targets at long range by SAR imaging.
Fig. 5 The photograph of the Leifeng pagoda and the analysis of its SAR image. (a) the photograph of the Leifeng pagoda; (a) SAR imaging results of the pagoda; (c) enlarged view of the spire part in (b); (d) photograph of the spire part corresponding to (c).
Resolution calibration and out-field imaging experiments presented above identify the excellent ability of MWP radar demonstrator. And the 10.02 GHz bandwidth and 14° beam width of the antennas also ensure the centimeter-level imaging resolution. It worth noted that this two-dimensional centimeter-level image resolution is the key for extracting quantitative target information from ISAR or SAR microwave images. Compared to former microwave images [14,20-21], target images obtained by this MWP radar demonstrator have sharper edges and show more detailed structure. All the above can dramatically improve the accurateness of target recognition based on microwave imaging. The significance of the proposed MWP demonstrator can summarized into three aspects: firstly, to our knowledge, this is the first time that MWP radar with centimeter-level resolution images non-cooperative targets, which shows its high value and potential for truly application; Secondly, based on MWP techniques, the bandwidth of the proposed demonstrator surpasses the most advanced electric microwave imaging radar-HUSIR, which indicates that microwave photonics provide an alternative way for realizing ultra-wideband imaging radar; Thirdly, although results in this paper are obtained by improved ISAR or SAR algorithms, challenges still exists for presenting more effective algorithms. Although the work distance currently is limited by the power amplifier, our work does show the application advantages of MWP radar and can promote the development of microwave imaging techniques.
4. Conclusions
A Ka-band MWP ultra-wideband imaging radar is proposed and experimentally demonstrated, which can generate and process radar signals with 10.02 GHz bandwidth and temporal periods of 150 μs-1200 μs. Based on standard test, the range resolution of the radar demonstrator was calibrated to be 1.68 cm. By ISAR imaging UAV and Boeing 737, and SAR imaging Leifeng pagoda in field-trail, the effectiveness and advantages of ultra-wideband MWP radar for obtaining quantitative target information are verified for microwave imaging applications. In addition, imaging distant targets with excellent resolution and surpassing the bandwidth of traditional microwave electric radar (HUSIR) indicate the great development prospect of the MWP radar technique.
Funding
National Natural Science Foundation of China (NSFC) (61701532); Natural Science Foundation of Hubei Province (2018CFB411, 2018CFB539, and 2018CFB331).
Acknowledgments
The authors thank Yitang Dai and Feifei Yin for their assistances with in-field experiments and Jixiang Fu and Mengdao Xing for their guidance in algorithm modification.
References
1. V. Chen and M. Martorella, Inverse Synthetic Aperture Radar Imaging: Principles, Algorithms, and Applications (SciTech Publishing, 2014)
2. W. L. Melvin and J. A. Scheer, Principles of Modern Radar: Vol. II Advanced Techniques (SciTech Publishing, 2012).
3. https://www.ll.mit.edu//publications/technotes/TechNote_HUSIR.pdf.
4. J. Capmany and D. Novak, “Microwave photonics combines two worlds,” Nat. Photonics 1(6), 319–330 (2007). [CrossRef]
5. J. Yao, “Microwave Photonics,” J. Lightwave Technol. 27(3), 314–335 (2009). [CrossRef]
6. J. D. McKinney, “Photonics illuminates the future of radar,” Nature 507(7492), 310–312 (2014). [CrossRef] [PubMed]
7. P. Ghelfi, F. Laghezza, F. Scotti, G. Serafino, A. Capria, S. Pinna, D. Onori, C. Porzi, M. Scaffardi, A. Malacarne, V. Vercesi, E. Lazzeri, F. Berizzi, and A. Bogoni, “A fully photonics-based coherent radar system,” Nature 507(7492), 341–345 (2014). [CrossRef] [PubMed]
8. 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]
9. L. Maleki, “The optoelectronic oscillator,” Nat. Photonics 5(12), 728–730 (2011). [CrossRef]
10. A. Wang, J. H. Wo, J. Zhang, X. Luo, X. Xu, D. M. Zhang, P. F. Du, and L. Yu, “Radio-frequency arbitrary waveform generation based on dispersion compensated tunable optoelectronic oscillator with ultra-wide tunability,” Chin. Opt. Lett. 15(10), 100603 (2017). [CrossRef]
11. J. Yao, “Arbitrary waveform generation,” Nat. Photonics 4(2), 79–80 (2010). [CrossRef]
12. X. S. Yao and L. Maleki, “Optoelectronic microwave oscillator,” J. Opt. Soc. Am. B 13(8), 1725–1735 (1996). [CrossRef]
13. S. Peng, S. Li, X. Xue, X. Xiao, D. Wu, X. Zheng, and B. Zhou, “High-resolution W-band ISAR imaging system utilizing a logic-operation-based photonic digital-to-analog converter,” Opt. Express 26(2), 1978–1987 (2018). [CrossRef] [PubMed]
14. F. Laghezza, F. Scotti, D. Onori, and A. Bogoni, “ISAR Imaging of Non-Cooperative Targets via Dual Band Photonics-Based Radar System,” Int Radar Symp Proc (2016). [CrossRef]
15. F. Scotti, D. Onori, and F. Laghezza, “Fully Coherent S- and X-Band Photonics-Aided Radar System Demonstration,” IEEE Microw. Wirel. Co. 25(11), 757–759 (2015). [CrossRef]
16. F. Scotti, F. Laghezza, D. Onori, and A. Bogoni, “Field trial of a photonics-based dual-band fully coherent radar system in a maritime scenario,” IET Radar Sonar Nav 11(3), 420–425 (2017). [CrossRef]
17. W. Zou, H. Zhang, X. Long, S. Zhang, Y. Cui, and J. Chen, “All-optical central-frequency-programmable and bandwidth-tailorable radar,” Sci. Rep. 6(1), 19786 (2016). [CrossRef] [PubMed]
18. F. Zhang, Q. Guo, Z. Wang, P. Zhou, G. Zhang, J. Sun, and S. Pan, “Photonics-based broadband radar for high-resolution and real-time inverse synthetic aperture imaging,” Opt. Express 25(14), 16274–16281 (2017). [CrossRef] [PubMed]
19. F. Z. Zhang, Q. S. Guo, Y. Zhang, Y. Yao, P. Zhou, D. Y. Zhu, and S. L. Pan, “Photonics-based real-time and high-resolution ISAR imaging of non-cooperative target,” Chin. Opt. Lett. 15(11), 95–98 (2017).
20. R. Li, W. Li, M. Ding, Z. Wen, Y. Li, L. Zhou, S. Yu, T. Xing, B. Gao, Y. Luan, Y. Zhu, P. Guo, Y. Tian, and X. Liang, “Demonstration of a microwave photonic synthetic aperture radar based on photonic-assisted signal generation and stretch processing,” Opt. Express 25(13), 14334–14340 (2017). [CrossRef] [PubMed]
21. J. Sheng, M. Xing, L. Zhang, M. Q. Mehmood, and L. Yang, “ISAR Cross-Range Scaling by Using Sharpness Maximization,” IEEE Geosci. Remote Sens. 12(1), 165–169 (2015). [CrossRef]
