Open Access
Add to BibSonomyAbstract
Optical serial coherent analyzer of radio-frequency is a novel scheme that enables fast-scanning microwave signal measurements in a large bandwidth. The measurements are performed based on serial channelization realized by using a fast scanning laser source as the local oscillator to down-convert the to-be-measured radio-frequency (RF) signals. Optical coherent detection effectively removes interferences induced by RF’s self-beating and guarantees the accuracy of measurements. In the experimental demonstration, instantaneous multi-frequency measurements and vector information acquisition of RF signals can be achieved by this scheme within 2.8 μs over 14 GHz bandwidth.
© 2014 Optical Society of America
1. Introduction
Radio-frequency (RF) measurements have attracted much attention for its applications in various fields such as electronic warfare and modern communications. Recently, by taking advantage of photonics, such as large bandwidth, low loss, immunity to electromagnetic interference et al., many photonic-assisted multiple-frequency microwave measurement schemes have been proposed. For multi-frequency measurements, or further signal analysis, approaches based on channelized receiving or frequency scanning have been hotspots [1–11]. Optical filters, such as filter bank, phase-shifted gratings, diffraction or waveguide gratings, are usually used to slice spectrum and construct parallel channels in channelized receivers [1–5]. Channelizers with reconfigurable operating bandwidth and channel spacing are realized based on tunable FP etalon, adjustable filters and so on [6–8]. But multiple transmission paths and receiving devices are necessary in these schemes, which leads to complicated structures. On the other hand, novel scanning receiving systems with large operating bandwidth have been proposed for multi-frequency measurements [9–11]. Scanning receivers are realized by controlling a tunable FP interferometer to produce power maxima [9], or using stimulated Brillouin scattering to selectively convert phase modulation to amplitude modulation [10], to identify to-be-measured RF signals. But the resolution and measurement speed are limited by external controlling signals. An electrically tuned fiber Bragg grating are used to realize the scanning receivers in [11], but it is realized by heating the fiber’s internal electrodes and thus the system is easily impacted by the environment.
In this paper, we propose and demonstrate a novel photonic approach for RF measurement based on fast laser scanning, called optical serial coherent analyzer of radio-frequency (OSCAR). Every individual frequency of the laser scanner exists in corresponding time window, so channelized slicing in frequency domain and time domain are realized simultaneously when the scanning signal is used as the local oscillator (LO) for down-conversion. Thus RF channels are constructed serially in time domain. At the receiver end, by an optical coupler and a balanced photodetector (BPD), the interferences from RF’s self-beating could be removed. This scheme is simple in construction and enables signal analysis in the span of 14 GHz bandwidth in just a few microseconds, which is a considerable improvement compared to common scanning receiver with millisecond-order scan rate.
2. Principle and experimental setup
The schematic of proposed scheme is described in Fig. 1. The continuous wavelength (CW) light is divided into two branches. In one branch, the CW light is injected to pump the scanning structure to generate scanning laser source with a constant frequency shifting step. While in the other branch, an incoming RF signal is modulated on the CW source by a Mach-Zehnder modulator (MZM) at the condition of carrier-suppressed modulation. Then the two branches signals are coupled into a narrow-band BPD and beat with each other. When the frequency difference between them is within the receiver bandwidth, the down-converted RF signal will be detected after O/E conversion. Finally, frequency components of the test RF signal are calculated simultaneously by analyzing the received beat notes. To understand the operation of the proposed system, we start from the principle of the laser scanner.
Fig. 1 Schematic of the proposed OSCAR. CW: continuous wavelength; PC: polarization controller; MZM: Mach-Zehnder modulator; ADC: analog-to-digital convertor.
As shown in Fig. 2, the laser scanner is realized based on a recirculating frequency shifting (RFS) loop [12] and an optical switch [13]. Firstly, the input CW is carved to pulses by the optical switch. And the switch is externally electric-controlled to ensure that duration of an individual pulse is equal to the round-trip time of the RFS loop. Then the optical pulses are injected into the RFS loop and modulated with a RF LO signal by a complex I/Q modulator. Under the condition of single sideband carrier-suppressed modulation, as the optical signal travels around the loop, its instantaneous frequency will shift with a step equal to the RF LO in every round trip. A band-pass filter (BPF) is used to limit the bandwidth of the scanning signal and an erbium doped fiber amplifier (EDFA) is used to compensate the loop loss. As a result, the output of the loop is a scanning signal whose frequency shift interval is equal to RF LO, hold time of an individual frequency is equal to loop round-trip time and scanning period is equal to optical switch’s on-off period. Clearly, the bandwidth of the BPF and RF LO collectively determine the scanning range, which corresponds to the scheme’s measurement range and can be easily readjusted.
Fig. 2 The structure of laser scanner based on optical switch and RFS. I/Q: in-phase/quadrature; BPF: band-pass filter; δt: hold time of switch opening; N·δt: one scanning period.
Next, we discuss the principle of the OSCAR scheme. The time-frequency characteristic of the scanning laser source and test RF is diagramed in Fig. 3(a). Colorful solid lines in Fig. 3(a) represent the time-varying frequencies of the laser scanner and blue arrows are the discrete frequencies of carrier-suppressed up-converted RF signal. The laser scanner’s frequency shifts from fc to fc + (N-1)·δf with a step of δf in a scanning period of N·δt, where fc is the CW frequency and δt is the hold time of an individual frequency. When receiver bandwidth is set to be exactly equal to the frequency step of the scanning laser source, different frequency components of RF signal will be received in different time windows of δt. That is, δf-width channels are serial in time domain and the diagonal areas show the time-frequency range of each channel in Fig. 3. Assuming that the modulated RF signal lies in the positions shown in Fig. 3(a), received beat notes should be located in the 2nd and the 3rd channel as illustrated in Fig. 3(b). In fact, they can be expressed as Eq. (1),
where fx is the frequency of test RF signal while f1 and f2 are detected beat frequencies in the two adjacent k-th and (k + 1)-th channels (here in the 2nd and 3rd channel) respectively. And the down-converted beat notes of one certain RF frequency should satisfy the following relation of f1 + f2 = δf. The parameter k could be decided by searching the time windows in which beat note exists. So fx can be identified by solving Eq. (1) as Fig. 3(c) shows. Since power of an individual scanning frequency is uniform and controllable, the relative power of the test RF signal could be achieved as well. If the RF signal contains a wideband spectrum of several hundreds of MHz, its amplitude-frequency information also can be identified by splitting spectrum into discretization.Fig. 3 Time-frequency diagrams of (a) the laser scanner’s frequencies (colorful solid lines) and modulated RF (blue arrows); (b) received beat notes (blue lines); (c) identified RF frequencies in corresponding channels (crossed-over blue arrows are incorrect identifications). Diagonal lined areas represent time-frequency range of different channels.
In order to eliminate the vagueness in identifying test frequency components, bandwidth of the receiver should be equal to or greater than the scanning frequency step of laser scanner. Otherwise, it is equivocal that a certain beat note is from (fc + fx)-fk or fk-(fc + fx) where fk represents the k-th frequency of laser scanner and fk = fc + (k-1)·δf, k = 1,2,…,N. It is better to set receiver bandwidth to be δf, which ensures simple digital signal processing.
Finally, coherent detection carried out by BPD is analyzed as follows. Assuming that the modulated RF signal contains multiple frequencies and fxm represents the m-th frequency, then it can be expressed as Eq. (2). Meanwhile the scanning LO signal fk is expressed as Eq. (3),
Then the output of BPD will be expressed as Eq. (4) [14],
So it’s known that received baseband components are exactly the beat notes between instantaneous scanning frequency and modulated RF when the BPD is ideal balanced. Compared to direct detection [15, 16], in which PD’s output is expressed as Eq. (5) and RF’s self-beat notes exist as well,
Coherent detection guarantees “clean” beat notes. For a certain frequency component whose time-domain expression (shown as Eq. (6)) and center frequency fxm and instantaneous amplitude Am(t) have been achieved, its relative phase could be calculated according to Eq. (7),
In which h(t) is the transmission function of a low-pass filter and the symbol “*” represents convolution. Thus, vector information (frequency, amplitude, phase) of x(t) can be obtained without ambiguity.
3. Measurement results and discussions
In the experiment, the RF LO and the receiver bandwidth are both set to be 2 GHz, which is also the channel bandwidth. The hold time of a single frequency is 400 ns in our experiment, so the resolution is 2.5 MHz in theory. The laser scanner contains 7 wavelengths scanning in a whole period of 2.8 μs, meaning that the system holds 7 channels with 357-kHz scanning rate in a frequency measurement range of 14 GHz. The bandwidth of the BPF can be adjusted according to different scanning bandwidth. The frequency of the laser scanner hops 2 GHz per shift and there is no tones in 14-16 GHz in the experiment, thus the bandwidth of the BPF can set to be greater than 14 GHz and less than 16 GHz. So it is set to be around 15 GHz in our experiment, which is a little wider than OSCAR’s measurement range to avoid the slope effect. The single round output of RFS loop is shown in Fig. 4(a) (measured by APEX AP2050 and the resolution bandwidth is 20 MHz). It is found that only + 1st sideband is retained and other sidebands, such as carrier frequency, −1st sideband and harmonic components, are well suppressed. The amplitude-frequency characteristic of the laser scanner detected by an oscilloscope (Agilent DSO91204A) is shown in Fig. 4(b) and the time-frequency feature is shown in Fig. 4(c).
Fig. 4 (a) The carrier-suppressed single-sideband output of I/Q modulator. (b) The amplitude-frequency characteristic of the laser scanner. (c) The time-frequency feature of the laser scanner.
Figure 5 shows the experimental results of multi-frequency RF signal which consists of discrete frequencies at 1.2, 2.4, 3.7, 4.5 and 5.6 GHz. As expected, the corresponding beat notes are received in the 1st channel at 1.2 GHz, the 2nd channel at 0.4, 0.8 and 1.7 GHz, the 3rd channel at 0.3, 0.5 and 1.6 GHz, and the 4th channel at 0.4 and 1.5 GHz. So the frequencies can be identified correctly according to Eq. (1) as shown in Fig. 5(b).
Fig. 5 (a) Beat notes in some channels of multi-frequency signal. (b) The recovered multi-frequency RF signal.
Afterwards, RF signals with a broad spectrum are tested in the experiment as well. When the RF signal contains 40 MHz frequency-components of 9.97 GHz-10.01 GHz, the down-converted baseband components are expected to be located in adjacent 5th, 6th and 7th channel. And it is found in Fig. 6 that only those 3 channels have received beat frequencies while no tone is detected in other channels. The recovered frequency is shown in Fig. 6(d).
Fig. 6 Beat notes received in the (a) 5th, (b) 6th, (c) 7th channel of wideband signal from 9.97 to 10.01 GHz. (d) The recovered wide-band RF signal.
In order to test OSCAR’s capability of complex signal analysis, we also measured a binary-phase-shift-keying (BPSK) signal whose center frequency is 9.4 GHz and bit rate is 50 Mb/s. The modulated signal is shown in Fig. 7(a) and the recovered information is in Fig. 7(b), which is calculated by carrier-demodulation and a low-pass filter according to Eq. (6).
At last, we measured the spurious free dynamic range (SFDR) to investigate the system’s noise and nonlinearity and evaluate its performance. In the experiment, the noise floor is −135 dBm and the SFDR achieved is 83 dB·Hz2/3 which is shown in Fig. 8.
The scheme is convenient to extend measurement range and change measurement resolution. By changing the RF LO and BPF’s bandwidth in the loop, and certainly the optical switch is adjusted accordingly, measurement range will be changed. And by changing time length of the RFS loop and the optical switch correspondingly, the resolution could be changed as well. Thus, this system is reconfigurable easily and promising in various RF vector measurements applications.
4. Summary
In this paper, a novel photonic RF measurement scheme called OSCAR has been proposed and experimentally verified. The measurement is performed based on serial channelization and coherent detection. The serial channelization is realized by using fast laser scanning structure as LO to down-convert the modulated RF signals. By this scheme, not only instantaneous multi-frequency measurement could be achieved, but also phase information of RF signals can be acquired successfully. The measurement covers 14 GHz range and is finished in 2.8 μs. Furthermore, this system is promised to have extensible measurement range and further improvement in various RF vector measurements applications.
Acknowledgments
This work is s supported by the National Program on Key Basic Research Project (973) under Contract 2012CB315703; NSFC under Contracts 61322113, 61090391, 61120106001; Tsinghua National Laboratory for Information Science and Technology (TNList) cross-discipline Foundation. The authors also thank the young top-notch talent program sponsored by Ministry of Organization, China.
References and links
1. X. Zou, W. Pan, B. Luo, L. Yan, and Y. Jiang, “Photonic approach to microwave frequency measurement with digital circular-code results,” Opt. Express 19(21), 20580–20585 (2011). [CrossRef] [PubMed]
2. D. B. Hunter, L. G. Edvell, and M. A. Englund, “Wideband microwave photonic channelised receiver,” in International Topical Meeting on Microwave Photonics (2005), pp. 249–252.
3. V. Borja, M. Teresa, and M. Javier, “Photonic technique for the measurement of frequency and power of multiple microwave signals,” IEEE Trans. Microw. Theory Tech. 58(11), 3103–3108 (2010). [CrossRef]
4. W. Wang, R. L. Davis, T. J. Jung, R. Lodenkamper, L. Lembo, J. Brook, and M. Wu, “Characterization of a coherent optical RF channelizer based on a diffraction grating,” IEEE Trans. Microw. Theory Tech. 49(10), 1996–2001 (2001). [CrossRef]
5. J. M. Heaton, C. D. Watson, S. B. Jones, M. M. Bourke, C. M. Boyne, G. W. Smith, and D. R. Wight, “16-channel (1- to 16-GHz) microwave spectrum analyzer device based on a phased array of GaAs/AlGaAs electro-optic waveguide delay lines,” Integrated Optic Devices II, SPIE 3278, 245–251 (1998). [CrossRef]
6. C. S. Brès, S. Zlatanovic, A. O. J. Wiberg, and S. Radic, “Reconfigurable parametric channelized receiver for instantaneous spectral analysis,” Opt. Express 19(4), 3531–3541 (2011). [CrossRef] [PubMed]
7. X. Zou and J. Yao, “Optical approach to microwave frequency measurement with adjustable measurement range and resolution,” IEEE Photon. Technol. Lett. 20(23), 1989–1991 (2008). [CrossRef]
8. X. Zou, W. Pan, B. Luo, and L. Yan, “Photonic approach for multiple-frequency-component measurement using spectrally sliced incoherent source,” Opt. Lett. 35(3), 438–440 (2010). [CrossRef] [PubMed]
9. S. T. Winnall and A. C. Lindsay, “A Fabry-Perot scanning receiver for microwave signal processing,” IEEE Trans. Microw. Theory Tech. 47(7), 1385–1390 (1999). [CrossRef]
10. S. Zheng, S. Ge, X. Zhang, H. Chi, and X. Jin, “High-resolution multiple microwave frequency measurement based on stimulated Brillouin scattering,” IEEE Photon. Technol. Lett. 24(13), 1115–1117 (2012). [CrossRef]
11. P. Rugeland, Z. Yu, C. Sterner, O. Tarasenko, G. Tengstrand, and W. Margulis, “Photonic scanning receiver using an electrically tuned fiber Bragg grating,” Opt. Lett. 34(24), 3794–3796 (2009). [CrossRef] [PubMed]
12. T. Kawanishi, T. Sakamoto, S. Shinada, and M. Izutsu, “Optical frequency comb generator using optical fiber loops with single-sideband modulation,” IEICE Electron. Express 1(8), 217–221 (2004). [CrossRef]
13. C. Lei, H. Chen, M. Chen, S. Yang, and S. Xie, “High-speed laser scanner with tunable scan rate, wavelength resolution and spectral coverage,” in Conference on Lasers and Electro-Optics, (Optical Society of America, 2013), paper JM3O.3. [CrossRef]
14. I. P. Kaminow, T. Li, and A. E. Willner, Optical Fiber Telecommunications V B: Systems and Networks (Elsevier, 2008), Chap. 3.
15. R. Li, H. Chen, Y. Yu, M. Chen, S. Yang, and S. Xie, “Multiple-frequency measurement based on serial photonic channelization using optical wavelength scanning,” Opt. Lett. 38(22), 4781–4784 (2013). [CrossRef] [PubMed]
16. R. Li, C. Lei, Y. Liang, H. Chen, M. Chen, S. Yang, and S. Xie, “Serial photonic channelized RF frequency measurement based on optical coherent frequency scanning,” in OptoElectronics and Communications Conference held jointly with 2013 International Conference on Photonics in Switching, pp. 1–2(2013).
