Expand this Topic clickable element to expand a topic
Skip to content
Optica Publishing Group

Remote broadband RF signal down-conversion with stable phase and high efficiency using a sideband optical phase-locked loop

Open Access Open Access

Abstract

We propose a phase-stable high-efficiency down-conversion approach for broadband radio frequency signals transmitted from a remote site. A high power coherent optical local oscillator signal is used at the local site to increase the conversion gain and the spurious-free dynamic range (SFDR) at the same time. A sideband optical phase-locked loop ensures the suppression of the phase noise induced by the fiber transmission and the relative frequency drift of the remote and local lasers, which are essential for the signal transmission and the down-conversion. We first experimentally demonstrate the down-conversion of a single frequency signal at 16.45 GHz to a 250 MHz intermediate frequency (IF) signal with 3 dB gain and 103 dB/Hz2/3 SFDR after 10 km fiber transmission. Then we show the broadband down-conversion capability by down-converting a 1 GHz wide linear frequency modulated pulse signal centered at 11 GHz to 1 GHz with 3 dB gain. Along with a positive gain, the SFDR of the IF signal down-converted from 5 GHz to 40 GHz has reached 97.6 dB/Hz2/3 on average. This approach is suitable for weak broadband remote signal down-conversion with a simple-structured remote end.

© 2020 Optical Society of America under the terms of the OSA Open Access Publishing Agreement

1. Introduction

With the development of modern communication, the frequency of the radio frequency (RF) signal is increasing rapidly for larger bandwidth and better anti-interference performance. Meanwhile, in many antenna-related applications, such as the remote antenna for radars [1,2], distributed antenna system for wireless communication [3,4] and antenna remoting for radioastronomy [5,6], the RF signal received by the remote antenna is miles away from the central station. Therefore, it is essential to first transmit the high-frequency RF signal (tens of GHz) from the remote end to the local site and then convert it to a lower-frequency intermediate frequency (IF) signal (hundreds of MHz) for the analog to digital conversion. Limited by the cost and the power consumption, the remote end should be as simple as possible. Besides, for most remote antenna applications, the power of the RF signal received by the remote site is weak [7]. Further considering the long-distance transmission loss, large conversion gain with high signal fidelity is highly desired.

Thanks to the advantages of low loss, broadband and immunity to electromagnetic interference in microwave photonic links [8], microwave photonic down-conversion is considered to be a superior technique for remote RF signal transmission and conversion with a broadband capability and high isolation [9]. Many microwave photonic down-conversion approaches are proposed to improve the conversion performance including the conversion gain, noise figure (NF) and spurious-free dynamic range (SFDR). The methods using dual-parallel Mach-Zehnder modulators (DPMZM) [1012] have a good performance on conversion gain and SFDR, but it requires a complex remote end with an additional local oscillator (LO) signal. In many approaches suitable for remote signal down-conversion, the method based on cascaded Mach-Zehnder modulators [1315] is first studied for the simplicity of the structure and the excellent coherence in the optical domain, but the conversion efficiency of this method is limited due to the low laser power and modulation efficiency. To get higher conversion gain, methods based on the nonlinearity of a semiconductor optical amplifier (SOA) [16,17] or an electro-absorption modulator [18,19] have been proposed. The cross-phase modulation and four-wave mixing at SOA can be used to realize frequency conversion, and the nonlinearity of the electro-absorption modulator can also get a photocurrent related to the beat signal between RF and LO signal. For the SOA-based approach, the conversion gain improvement is not due to the high efficiency of the nonlinear effects but the amplification of the optically carried RF signal and the LO carrier in the SOA. However, the nonlinearity induced undesired harmonic wave deteriorates the down-conversion SFDR. In virtue of the good linearity of the phase modulator, some approaches use phase modulators to modulate the RF signal at the remote site and generate coherent LO signal by injection locking [20,21], but the locking range of the slave laser is limited, and the frequency of the down-converted IF signal is hard to adjust flexibly. Currently, it is difficult to find a down-conversion approach for the remote RF signal with large conversion gain, high SFDR and flexible IF signal.

In this paper, we propose a down-conversion method for the remote RF signal based on a sideband optical phase-locked loop (OPLL). A high-power optical LO signal is used at the local end to increase the conversion gain. Thanks to the OPLL, the local laser is phase-locked to the remote laser, resulting in the reduction of the phase noise induced by the two lasers and the fiber transmission. The bandwidth of the remote RF signal can be large within the modulation range of 40 GHz while the frequency of the required microwave LO signal is only less than half of that of the remote RF signal. Owing to the flexibility of the proposed structure, the frequency of the IF signal can be adjusted arbitrarily to adapt to different situations. We experimentally demonstrate remote RF signal down-conversion from 16.45 GHz to 250 MHz with a 3 dB conversion gain, and 103 dB/Hz2/3 SFDR over a 10 km fiber link. Furthermore, a linear frequency modulated (LFM) pulse signal sweeping from 10.5 GHz to 11.5 GHz is down-converted to an IF signal sweeping from 0.5 GHz to 1.5 GHz with a 3 dB conversion gain and high sweep linearity. The SFDR of the whole conversion range from 5 GHz to 40 GHz is 97.6 dB/Hz2/3 on average, showing high linearity of the down-conversion system.

2. Principle

2.1 Structure

The structure of the proposed down-conversion method is shown in Fig. 1. At the remote site, we use a fiber laser (Laser1) and an intensity modulator (IM1) to modulate the RF signal received by the remote antenna, which is then transmitted to the local site through a fiber link. At the local site, we use another high-power fiber laser (Laser2) as the optical LO signal, which is modulated by a tunable microwave signal in a local intensity modulator (IM2) with carrier-suppressed double sideband modulation. After coupled together, the right sideband of the LO signal is heterodyne phase-locked with the carrier of Laser1, and the left sideband of the LO signal is used to down-convert the RF signal with a photodetector (PD). Thanks to the well-designed OPLL, the phase of the local Laser2 is locked to the transmitted remote Laser1. Thus, the phase noise induced by the two lasers and the fiber transmission can be reduced to the minimum. Meanwhile, the high power coherent LO signal results in evident improvement on the conversion gain without the loss of the signal dynamic range, ensuring high signal fidelity even if the received RF signal is very weak at the remote end. The central frequency and the instantaneous bandwidth of the RF signal can be arbitrary within the range of 40 GHz, which is only limited by the bandwidth of IM1. The IF signal can also be set at any frequency flexibly by adjusting the frequency of the local microwave LO signal. It should be noted that the frequency of the LO microwave signal is less than half of the RF signal, lowering the requirement for the high-frequency LO source and reducing the system cost. The LO frequency can be further decreased if higher-order sidebands of the coherent LO signal is used. Nevertheless, in order to remain the same down-conversion performance when using higher-order sideband locking, a higher-power local laser and a microwave LO source with less phase noise are required.

 figure: Fig. 1.

Fig. 1. The principle of the down-conversion using a sideband OPLL.

Download Full Size | PDF

2.2 Phase noise suppression

Here, we analyze how the sideband OPLL reduces the phase noise in theory. For simplicity, we first ignore the amplitude information of the lasers, which has no contribution to the phase stability of the final IF signal. In the proposed structure, the remote signal is transmitted along the fiber link with a time delay τ. At the local site, the local laser first passes an acousto-optic frequency shifter (AOFS) for close-loop frequency adjustment and is then modulated by the microwave LO signal. The optically carried RF signal is heterodyne beaten with the left sideband of the LO signal. The IF signal can be expressed as:

$$\begin{aligned}{E_{IF}} &= \exp \{ j[({\omega _1} + {\omega _{RF}})(\textrm{t} - \tau ) + {\varphi _1}(\textrm{t} - \tau ) + {\varphi _{RF}}(\textrm{t} - \tau )]\} \\ &- \exp \{ j[({\omega _2} + {\omega _{LO}} + {\omega _{AOFS}})t + {\varphi _2}(\textrm{t}) + {\varphi _{LO}}(\textrm{t}) + {\varphi _{AOFS}}(\textrm{t})]\} \\ & = \exp \{ j[({\omega _1} - {\omega _2} + {\omega _{RF}} - {\omega _{LO}} - {\omega _{AOFS}})t - {\omega _1}\tau - {\omega _{RF}}\tau \\ & + {\varphi _{RF}}(\textrm{t} - \tau ) + {\varphi _1}(\textrm{t} - \tau ) - {\varphi _2}(\textrm{t}) - {\varphi _{LO}}(\textrm{t}) - {\varphi _{AOFS}}(\textrm{t})]\} , \end{aligned}$$
where ${\omega _{1,2}}$, ${\omega _{RF,LO}}$, and ${\omega _{AOFS}}$ represent the angular frequency of Laser1 and Laser2, the RF and LO signal, and the AOFS, respectively; ${\varphi _{1,2}}$, ${\varphi _{RF,LO}}$, and ${\varphi _{AOFS}}$ represent their phase; ${\varphi _1}({t - \tau } )- {\varphi _2}(t )$ represents the phase noise induced by the two lasers. Since the transmission delay varies with environment changes [22], the term ${\omega _1}\tau $ is the phase noise induced by the remote laser during the fiber link transmission. The term ${\omega _{RF}}\tau $ represents the phase noise of RF signal after transmission which is negligible compared with ${\omega _1}\tau $ because the frequency of the remote laser (hundreds of THz) is way larger than the frequency of the RF signal (tens of GHz). All these terms indicate that the two different lasers and the fiber link will induce frequency drift and phase noise into the IF signal. When phase-locking the beating signal of right sideband LO signal and the carrier of Laser1 with a standard reference, we can get:
$$\begin{array}{l} [({\omega _2} - {\omega _{LO}} + {\omega _{AOFS}})t + {\varphi _2}(t) - {\varphi _{LO}}(t) + {\varphi _{AOFS}}(t)] - [{\omega _1}(t\textrm{ - }\tau ) + {\varphi _1}(t - \tau )]\\ = ({\omega _2} - {\omega _{LO}} + {\omega _{AOFS}} - {\omega _1})t + {\varphi _2}(t) - {\varphi _{LO}}(t) + {\varphi _{AOFS}}(t)\textrm{ + }{\omega _1}\tau - {\varphi _1}(t - \tau )\\ = {\omega _{ref}}t + {\varphi _{ref}}(t), \end{array}$$
where ${\omega _{ref}}$,${\; }{\varphi _{ref}}$ represent the angular frequency and the phase of the standard reference signal, respectively. Substitute Eq. 2 into Eq. 1, we can get:
$${E_{IF}} = \exp \{ j[({\omega _{RF}} - 2{\omega _{LO}} - {\omega _{ref}})t + {\varphi _{RF}}(t - \tau ) - 2{\varphi _{LO}}(t) - {\varphi _{ref}}(t) - {\omega _{RF}}\tau ]\} .$$
Equation 3 shows that the phase noise induced by two lasers and the fiber transmission is eliminated, and the phase of the final IF signal is mainly dependent on the phase of the LO signal and the reference signal. If a highly-stable reference signal and microwave LO signal are used, a phase-stable IF signal can be obtained. Note that the last term of Eq. 3, ${\omega _{RF}}\tau $ represents the residual phase variation of the RF signal after transmission, which is negligible compared with the noise induced by two lasers and the fiber link.

2.3 Conversion gain increment

Conversion gain is one of the most important indicators of a down-conversion system, especially when the received signal is very weak. The definition of the conversion gain is given by

$$G = \frac{{{P_{IF}}}}{{{P_{RF}}}},$$
where ${P_{IF}}$ and ${P_{RF}}$ are the power of the IF signal and RF signal, respectively. In the proposed approach, we assume that the microwave signal loaded on the modulators are ${v_{RF}}$ and ${v_{LO}}$, where ${v_{RF,LO}}(t )= {V_{RF,LO}}\sin ({\omega _{RF,LO}}t + {\varphi _{RF,LO}})$. When the modulator IM1 and IM2 working at the quadrature bias point which is the same as that in the cascaded modulator based method, the electrical power of the IF signal got from a photodetector with responsivity $\gamma $ and 50 $\Omega $ load can be expressed as:
$${I_{IF}} = 2\gamma \sqrt {\frac{{{P_{laser1}}{P_{laser2}}{\alpha _1}{\alpha _2}{\alpha _{link}}}}{4}} {J_1}({X_{RF}}){J_1}({X_{LO}})\cos ({\omega _{RF}} - {\omega _{LO}})t,$$
$${P_{IF - OPLL}} = \frac{{50}}{2}{P_{laser1}}{P_{laser2}}{\alpha _1}{\alpha _2}{\alpha _{link}}{[{J_1}({X_{RF}}){J_1}({X_{LO}})\gamma ]^2},$$
where ${X_{RF.LO}} = \frac{{\pi {V_{RF,LO}}}}{{{V_{\pi 1,2}}}}$ is the modulation depth. ${P_{laser1}}$ and ${P_{laser2}}$ are the optical output power of the two lasers. ${J_1}$ is the first-order Bessel function. ${\alpha _1}{\alpha _2}{\alpha _{link}}$ represents the insertion loss of the two modulators and the fiber link. ${V_{\pi 1,2}}$ is the half-wave voltage of the two modulators. We compare it with the cascaded modulator based method in terms of the conversion gain, which is a classic approach suitable for remote RF signal down-conversion. At the same working condition including ${v_{RF,LO}}$, the bias point of modulators, the responsivity of the photodetector, the electrical power of the IF signal in the cascaded modulator based method is given by [13]:
$${P_{IF - cascaded}} = \frac{{50}}{2}{\left( {\frac{{\gamma {\alpha_1}{\alpha_2}{\alpha_{link}}{P_{laser1}}}}{2}{J_1}({X_{RF}}){J_1}({X_{LO}})} \right)^2}.$$
According to the definition of conversion gain, the ratio of these two conversion methods is:
$$\frac{{{G_{OPLL}}}}{{{G_{cascaded}}}} = \frac{{{P_{IF - OPLL}}}}{{{P_{IF - cascaded}}}} = \frac{{4{P_{laser2}}}}{{{P_{laser1}}{\alpha _1}{\alpha _2}{\alpha _{link}}}}.$$

The ratio of these two conversion methods is dependent on two terms, namely the term $\frac{{{P_{laser2}}}}{{{P_{laser1}}}}$, representing the power ratio of the local and remote lasers, and the term ${\alpha _1}{\alpha _2}{\alpha _{link\; }}$, representing the loss of the system. The higher power the local laser has, the more superior the proposed method will be. This is beneficial to the real scenario where a high power LO laser is often used and the power consumption limits the power of the remote laser. It should be noted that in the above deduction, we bias IM2 at the quadrature point for a fair comparison with the cascaded modulator based method. In the following experiments, we actually bias IM2 at the null point to obtain even higher LO power. In the cascaded modulator based method, however, the local modulator can not work at the null point because the carrier-suppressed modulation will decrease the power of the optical RF component. Besides, the greater the loss of the system has, the greater the ratio of these two methods is. Because in the proposed structure, the RF and LO signal only goes through one modulator rather than two modulators in the cascaded modulator based method. Thus, it experiences less loss of the modulators and fiber links which is advantageous when the loss induced by the modulators and fiber link is considerable. Compared with the classic method, the proposed method has evident gain increment and is less affected by the loss of the system, making it a promising candidate for remote signal down conversion.

3. Experiment and result

The experimental setup is shown in Fig. 2. A microwave source is used to generate the RF signal, which is first modulated by an intensity modulator on the light and is then transmitted through a 10 km single-mode fiber link. The output power of Laser1 is 12 dBm and the half-wave voltage of IM1and IM2 is 6.5 V. The RF signal is modulated at the quadrature bias point to ensure modulation linearity. At the local site, the remote signal is coupled with the local laser (Laser2) which is modulated by an LO microwave signal with suppressed-carrier modulation. The optical signals are split into two parts through two 9:1 fiber couplers. 10% of the signals are used for the phase-locking and 90% of the signals are for down-converting. To further increase the conversion gain, an EDFA with a noise figure of about 5.5 dB is used to boost the optical power. An optical filter with a bandwidth of about 10 GHz is followed to only select the left sideband of the LO signal and the optically carried RF signal for down-conversion. The down-converted signal is obtained from the PD1. For the phase-locking branch, the right sideband of the LO signal beats the central wavelength of Laser1 in PD2. The bandwidth of PD1 and PD2 is 1.6 GHz. This beating signal is sent to a phase-frequency detector (PFD) as a feedback signal whose phase is compared with a 200 MHz standard reference signal generated by an arbitrary waveform generator (AWG). The phase error signal is first adjusted by the loop filters and then controls the frequency and the phase of the local laser. In this experiment, we use a composite phase-locked loop to ensure the phase-locking of the two lasers. In the fast loop, a voltage-controlled oscillator (VCO) driven by the adjusted phase error signal is used to change the frequency of the local laser via an acousto-optic frequency shifter (AOFS). It ensures the fast phase cloning of the local laser from the remote laser, leading to a stable heterodyne beat signal. Considering the large drift between the two lasers and the limited frequency tracking range of the AOFS, a slow loop consisting of a piezoelectric transducer (PZT) is further used. The PZT can adjust the wavelength of the local laser in a slower pace however with a larger range. Thus, it can expand the locking frequency range and ensure that the frequency difference of the two lasers is always within the locking range of the fast loop. The composite phase-locked loop keeps the optical LO signal synchronizing with the remote laser and finally provides a coherent LO signal to down-convert the RF signal with considerable conversion gain and stable phase. Note that most parts of the OPLL including the PFD and loop filters are integrated on a single circuit, making this solution compact and cost-efficient.

 figure: Fig. 2.

Fig. 2. Schematic diagram of the experimental setup. IM, intensity modulator. AOFS, acousto-optic frequency shifter. PFD, phase-frequency detector. VCO, voltage-controlled oscillator. PZT, piezoelectric transducer. EDFA, erbium-doped optical fiber amplifier. OBPF, optical bandpass filter.

Download Full Size | PDF

Since the right sideband of the LO signal is heterodyne phase-locked with the carrier of the remote laser, the frequency difference between these two optical signals is the same with the frequency of a local reference signal, which is a standard sinusoidal signal of 200 MHz in the experiment, resulting in a phase-stable IF signal. The optical spectra of the transmitted remote RF signal and the LO signal at the local site are shown in Fig. 3(a). Here we use a 16.45 GHz RF signal with -3 dBm power at the remote site, and the modulation depth of the RF signal is 10.7% (${X_{RF}} = \frac{{\pi {V_{RF}}}}{{{V_{\pi 1}}}}$). Note that it is in the linear modulation region of the remote modulator. The relatively weak power of the RF signal is consistent with the actual situation at the remote antenna site. And then we use an 8 GHz LO signal with 16 dBm power at the local site to generate a powerful optical LO signal. A phase-stable 250 MHz IF signal with 0 dBm power is observed by a real-time signal analyzer and a phase noise analyzer as shown in Fig. 3(b) and 3(c), respectively. Here, the conversion gain is 3 dB after a 10 km fiber link, which is limited by the maximum power of the two lasers (Laser1 and Laser2), the microwave source, and the EDFA. Higher conversion gain can be obtained if higher-power sources (lasers and microwave source) are used as long as the PD does not reach the saturation region. The red curve in Fig. 3(b) is the spectrum of the 250 MHz IF signal obtained at PD1 in phase-locked condition, and for comparison, the black curve is the spectrum of the IF signal in the free-running condition without phase locking (note that the frequency of IF signal in the free-running condition is constantly drifting as shown by the dotted lines, we shift its central frequency to 250 MHz for clearer comparison). It can be seen that the phase noise of the IF signal induced by two lasers at free-running condition is reduced obviously in the phase-locked condition, resulting in higher frequency stability of the IF signal. The platform of about 400 kHz wide at the center of the red curve is caused by the OPLL’s limited loop bandwidth within which the phase characteristic depends on the local 200 MHz reference signal. The phase noise spectra of the IF signal and the reference signal are compared in Fig. 3(c). The phase noise spectrum of the IF signal represented by the blue curve almost overlaps that of the reference signal below 104 Hz, indicating that the IF signal acquires the same phase stability as the reference signal. Similarly, the phase noise platform of the IF signal near 400 kHz is mainly caused by the OPLL’s limited loop bandwidth. Note that the phase noise spikes around 150 Hz are induced by the phase-locking circuit. In the free-running condition, the phase noise spectrum of the IF signal cannot be observed because the frequency is drifting rapidly with an average speed of a few MHz per second. We also test the instantaneous frequency deviation of the IF signal shown in Fig. 3(d), and the frequency deviation variance is 0.147 Hz2. The results show that the conversion method can generate an IF signal with high phase stability and very small frequency deviation which is extremely important for phase-sensitive applications.

 figure: Fig. 3.

Fig. 3. Experimental results of the down-conversion of a single frequency signal. a) The optical spectrum at the local site; b) The spectrum of the IF signal; c) The phase noise spectra of the IF signal and the reference signal; d) The frequency deviation of the IF signal.

Download Full Size | PDF

The LFM pulse signal is regarded as a typical high-frequency wideband RF signal, which can be widely applied in many situations such as linear frequency modulated continuous wave (LFM-CW) radar. To demonstrate the system’s ability to down-convert the actual remote wideband RF signals, we use a high sampling rate AWG to generate an LFM pulse signal sweeping from 10.5 GHz to 11.5 GHz with a 4 µs pulse width in a 5 µs period. The waveform detected by a high-sample rate oscilloscope is shown in Fig. 4(a), and its time-frequency diagram is shown in Fig. 4(b). After 10 km fiber transmission and down-converting, the IF signal sweeping from 0.5 GHz to 1.5 GHz is observed. Its waveform and time-frequency diagram are shown in Figs. 4(c) and 4(d). The amplitude of the down-converted signal is about 0.2 V and the conversion gain is about 3 dB which demonstrates the positive conversion gain of our experimental setup. Because the overall frequency response of the modulators, the EDFA and the PD we use is not flat on the sweep range, there exists some deterioration on amplitude flatness, but the frequency sweep linearity of the down-conversion signal shown in Fig. 4(d) keeps the same with the original pulse signal. The noise floor of the down-converted signal is increased slightly due to the SNR loss of electro-optical modulation and the EDFA induced ASE noise.

 figure: Fig. 4.

Fig. 4. Experimental results of the down-conversion of an LFM signal. a) The waveform of the LFM pulse signal; b) The time-frequency diagram of the pulse signal; c) The waveform of the down-converted signal; d) The time-frequency diagram of the down-converted signal.

Download Full Size | PDF

SFDR is an important parameter that represents the linearity of the down-conversion. Many approaches with high conversion gain result in low SFDR (below 90 dB/Hz2/3) because of the undesirable harmonic distortion. To test the SFDR of the proposed method, a 16.45 GHz and a 16.451 GHz RF signal are modulated on the light at quadrature bias-point at the remote site. With the increase of the RF power, not only the IF signals at 250 MHz and 251 MHz but also the third-order-intermodulation (IM3) signals at 249 MHz and 252 MHz can be observed. The result of the two-tone test is shown in Fig. 5(a). The platform under the IF signals is also caused by OPLL’s loop bandwidth. At the quadrature point, the SFDR of the down-conversion approach is 103 dB/Hz2/3 as demonstrated in Fig. 5(b), showing good linearity of the down-conversion system. The SFDR of the proposed method is mainly limited by the nonlinearity of the modulators. In addition, the amplified spontaneous emission (ASE) noise of the EDFA will increase the noise floor thereby also slightly decreasing the SFDR.

 figure: Fig. 5.

Fig. 5. Experimental results of the system dynamic property. a) The spectrum of the Two-tone test, RBW: 50 Hz; b) SFDR of the down-conversion at the quadrature point

Download Full Size | PDF

In order to measure the performance of the proposed down-conversion in the whole conversion range, we test the SFDR by changing the frequency of the RF signal from 5 GHz to 40 GHz, which is limited by the frequency range of the microwave source and the modulator bandwidth. When measuring the SFDR with different RF frequencies, we keep the experimental configuration unchanged including the power of Laser1 and Laser2, the working point of IM1 and IM2, the power of the EDFA, etc. And we test the SFDR of our system at the quadrature point by just adjusting the RF frequency in a 5 GHz step in the whole conversion range as shown in Fig. 6. The SFDR is 97.6 dB/Hz2/3 on average in the whole range and the peak value reaches 104 dB/Hz2/3. It proves that the proposed approach has good linearity for remote RF signal transmission and down-conversion. The slight SFDR decreasing with the increase of the RF frequency is mainly caused by the decrease of the conversion gain while the noise floor remaining the same.

 figure: Fig. 6.

Fig. 6. The SFDR of the down-conversion from 5 GHz to 40 GHz

Download Full Size | PDF

Finally, we compare different conversion methods based on microwave photonics on some important indicators as shown in Table 1. As can be seen from the table, for many microwave photonic down-conversion approaches, it is difficult to have both good performances on conversion gain and SFDR. Meanwhile, not all approaches are suitable for down-converting a broadband signal transmitted from a remote site. The proposed approach achieves flexible down-conversion with a positive gain and good linearity by using merely a lower-frequency microwave LO signal, showing evident advantages in converting broadband remote RF signal.

Tables Icon

Table 1. Comparison of different conversion methods based on microwave photonics.

4. Conclusion

In summary, we down-convert the remote broadband RF signal with a stable phase, high efficiency and high linearity. By using a high power coherent LO signal at the local site whose sideband is locked with the remote laser, large conversion gain has been reached and the phase noise induced by the fiber link and the laser drifting has been eliminated. In the proposed structure, the instantaneous bandwidth of the remote RF signal can be large within the modulation range and the frequency of the IF signal can be adjusted flexibly to meet different situations. In addition, the frequency of the required microwave LO signal is reduced to less than half of the RF signal, easing the requirement for the local microwave source and expanding the conversion range. The down-conversion of a 16.45 GHz RF signal to a 250 MHz IF signal is demonstrated over a 10 km fiber link with a 3 dB gain, 0.147 Hz2 frequency deviation variance and 103 dB/Hz2/3 SFDR. We also down-convert an LFM pulse signal sweeping from 10.5 GHz to 11.5 GHz into an IF signal sweeping from 0.5 GHz to 1.5 GHz with 3 dB conversion gain and good linearity. The SFDR of the whole conversion range from 5 GHz to 40 GHz is up to 104 dB/Hz2/3 and 97.6 dB/Hz2/3 on average. The experimental results show that the proposed down-conversion system has high efficiency, a stable phase, and high linearity, which is very suitable for remote broadband RF signal down-conversion.

Funding

National Natural Science Foundation of China (61690193, 61827807).

Disclosures

The authors declare no conflicts of interest.

References

1. M. Zhang, Y. Ji, and Y. Zhang, “Remote Radar based on Chaos Generation and Radio over Fiber,” IEEE Photonics J. 6(5), 1–12 (2014). [CrossRef]  

2. D. Grodensky, D. Kravitz, and A. Zadok, “Ultra-Wideband Microwave-Photonic Noise Radar Based on Optical Waveform Generation,” IEEE Photonics Technol. Lett. 24(10), 839–841 (2012). [CrossRef]  

3. M. Crisp, R. V. Penty, I. H. White, and A. Bell, “Wideband Radio over Fiber Distributed Antenna Systems for phase-locked In-Building Wireless Communications,” IEEE Vehicular Technology Conference, 1–5 (2010).

4. J. Wang, H. Zhu, and N. J. Gomes, “Distributed Antenna Systems for Mobile Communications in High-Speed Trains,” IEEE J. Select. Areas Commun. 30(4), 675–683 (2012). [CrossRef]  

5. S. Montebugnoli, M. Boschi, F. Perini, P. Faccin, G. Brunori, and Pirazzini, “Large antenna array remoting using radio over fiber techniques for radio astronomical application,” Microw. Opt. Technol. Lett. 46, 48–54 (2005). [CrossRef]  

6. E. Ackerman, C. Cox, J. Dreher, K. Davis, and D. Deboer, “Fiber-optic antenna remoting for radioastronomy applications,” General Assembly of International Union of Radio Science, Maastricht, pp. 1–4 (2004).

7. M. Matsuura and J. Sato, “Bidirectional Radio-Over-Fiber Systems Using Double-Clad Fibers for Optically Powered Remote Antenna Units,” IEEE Photonics J. 7(1), 1–9 (2015). [CrossRef]  

8. R. Esman, S. Pappert, B. Krantz, and G. Gopalakrishnan, “Photonics for Microwave Generation, Transmission, and Processing,” OSA Optical Fiber Communication Conference, San Diego, p. OTuE5, (2009).

9. R. W. Ridgway and C. L. Dohrman, “Microwave Photonics Programs at DARPA,” J. Lightwave Technol. 32(20), 3428–3439 (2014). [CrossRef]  

10. L. Huang, R. Li, D. Chen, and P. Wang, “Photonic Downconversion of RF Signals With Improved Conversion Efficiency and SFDR,” IEEE Photonics Technol. Lett. 28(8), 880–883 (2016). [CrossRef]  

11. E. H. W. Chan and R. A. Minasian, “Microwave Photonic Downconverter With High Conversion Efficiency,” J. Lightwave Technol. 30(23), 3580–3585 (2012). [CrossRef]  

12. H. Li, Y. Wang, and D. Wang, “High dynamic range microwave photonic down-conversion based on dual-parallel Mach-Zehnder modulator,” International Symposium on Optoelectronic Technology and Application, 1–6 (2016).

13. G. K. Gopalakrishnan, R. P. Moeller, M. M. Howerton, W. K. Burns, K. J. Williams, and R. D. Esman, “A low-loss downconverting analog fiber-optic link,” IEEE Trans. Microwave Theory Tech. 43(9), 2318–2323 (1995). [CrossRef]  

14. R. Helkey, J. C. Twichell, and C. Cox, “A down-conversion optical link with RF gain,” J. Lightwave Technol. 15(6), 956–961 (1997). [CrossRef]  

15. A. Karim and J. Devenport, “High Dynamic Range Microwave Photonic Links for RF Signal Transport and RF-IF Conversion,” J. Lightwave Technol. 26(15), 2718–2724 (2008). [CrossRef]  

16. C. Bohemond, T. Rampone, and A. Sharaiha, “Performances of a Photonic Microwave Mixer Based on Cross-Gain Modulation in a Semiconductor Optical Amplifier,” J. Lightwave Technol. 29(16), 2402–2409 (2011). [CrossRef]  

17. H. Song and J. Song, “Simultaneous all-optical frequency downconversion technique utilizing an SOA-MZI for WDM radio over fiber (RoF) applications,” J. Lightwave Technol. 24(8), 3028–3034 (2006). [CrossRef]  

18. J. Seo, C. Choi, W. Choi, Y. Kang, and Y. Chung, “Remote optoelectronic frequency down-conversion using 60-GHz optical heterodyne signals and an electroabsorption Modulator,” IEEE Photonics Technol. Lett. 17(5), 1073–1075 (2005). [CrossRef]  

19. D. S. Shin, G. L. Li, C. K. Sun, S. A. Pappert, K. K. Loi, W. S. C. Chang, and P. K. L. Yu, “Optoelectronic RF signal mixing using an electroabsorption waveguide as an integrated photodetector/mixer,” IEEE Photonics Technol. Lett. 12(2), 193–195 (2000). [CrossRef]  

20. C. Lin, B. M. L. Pascoguin, S. B. Jester, D. C. Evans, J. R. Adleman, and E. W. Jacobs, “Coherent detection RF-IF down-converting link using injection-locked laser design,” Asia-Pacific Microwave Photonics Conference (APMP), Sendai, 71–74 (2014).

21. J. R. Adleman and C. L. Lin, “Photonic RF-IF wideband down conversion using optical injection locking,” Micro and Nanotechnology Sensors, Systems, and Applications VII, 946721 (2015).

22. Y. Dong, W. Xie, W. Wei, Z. Liu, X. Wang, and N. Deng, “Highly stable dissemination of microwave and millimeter-wave signals over fiber-optic links,” CLEO Pacific Rim Conference, paper Th4F.1(2018).

23. B. M. Haas and T. E. Murphy, “Linearized Downconverting Microwave Photonic Link Using Dual-Wavelength Phase Modulation and Optical Filtering,” IEEE Photonics J. 3(1), 1–12 (2011). [CrossRef]  

Cited By

Optica participates in Crossref's Cited-By Linking service. Citing articles from Optica Publishing Group journals and other participating publishers are listed here.

Alert me when this article is cited.


Figures (6)

Fig. 1.
Fig. 1. The principle of the down-conversion using a sideband OPLL.
Fig. 2.
Fig. 2. Schematic diagram of the experimental setup. IM, intensity modulator. AOFS, acousto-optic frequency shifter. PFD, phase-frequency detector. VCO, voltage-controlled oscillator. PZT, piezoelectric transducer. EDFA, erbium-doped optical fiber amplifier. OBPF, optical bandpass filter.
Fig. 3.
Fig. 3. Experimental results of the down-conversion of a single frequency signal. a) The optical spectrum at the local site; b) The spectrum of the IF signal; c) The phase noise spectra of the IF signal and the reference signal; d) The frequency deviation of the IF signal.
Fig. 4.
Fig. 4. Experimental results of the down-conversion of an LFM signal. a) The waveform of the LFM pulse signal; b) The time-frequency diagram of the pulse signal; c) The waveform of the down-converted signal; d) The time-frequency diagram of the down-converted signal.
Fig. 5.
Fig. 5. Experimental results of the system dynamic property. a) The spectrum of the Two-tone test, RBW: 50 Hz; b) SFDR of the down-conversion at the quadrature point
Fig. 6.
Fig. 6. The SFDR of the down-conversion from 5 GHz to 40 GHz

Tables (1)

Tables Icon

Table 1. Comparison of different conversion methods based on microwave photonics.

Equations (8)

Equations on this page are rendered with MathJax. Learn more.

E I F = exp { j [ ( ω 1 + ω R F ) ( t τ ) + φ 1 ( t τ ) + φ R F ( t τ ) ] } exp { j [ ( ω 2 + ω L O + ω A O F S ) t + φ 2 ( t ) + φ L O ( t ) + φ A O F S ( t ) ] } = exp { j [ ( ω 1 ω 2 + ω R F ω L O ω A O F S ) t ω 1 τ ω R F τ + φ R F ( t τ ) + φ 1 ( t τ ) φ 2 ( t ) φ L O ( t ) φ A O F S ( t ) ] } ,
[ ( ω 2 ω L O + ω A O F S ) t + φ 2 ( t ) φ L O ( t ) + φ A O F S ( t ) ] [ ω 1 ( t  -  τ ) + φ 1 ( t τ ) ] = ( ω 2 ω L O + ω A O F S ω 1 ) t + φ 2 ( t ) φ L O ( t ) + φ A O F S ( t )  +  ω 1 τ φ 1 ( t τ ) = ω r e f t + φ r e f ( t ) ,
E I F = exp { j [ ( ω R F 2 ω L O ω r e f ) t + φ R F ( t τ ) 2 φ L O ( t ) φ r e f ( t ) ω R F τ ] } .
G = P I F P R F ,
I I F = 2 γ P l a s e r 1 P l a s e r 2 α 1 α 2 α l i n k 4 J 1 ( X R F ) J 1 ( X L O ) cos ( ω R F ω L O ) t ,
P I F O P L L = 50 2 P l a s e r 1 P l a s e r 2 α 1 α 2 α l i n k [ J 1 ( X R F ) J 1 ( X L O ) γ ] 2 ,
P I F c a s c a d e d = 50 2 ( γ α 1 α 2 α l i n k P l a s e r 1 2 J 1 ( X R F ) J 1 ( X L O ) ) 2 .
G O P L L G c a s c a d e d = P I F O P L L P I F c a s c a d e d = 4 P l a s e r 2 P l a s e r 1 α 1 α 2 α l i n k .
Select as filters


Select Topics Cancel
© Copyright 2024 | Optica Publishing Group. All rights reserved, including rights for text and data mining and training of artificial technologies or similar technologies.