Open Access
Add to BibSonomy
Abstract
Camouflage is a strategy that animals utilize for concealment in their habitat, making themselves invisible to their predators and preys. In RF systems, steganography or stealth transmission is the camouflage of information – a technology of hiding and transmitting secret messages in public media. Steganography conceals the secret message in publicly available media such that the eavesdropper or attacker will not be able to tell if there is a secret message to look for. Marine hatchetfish have two effective camouflage skills to help them hide from their predators – silvering and counterillumination. Silvering in marine hatchetfish uses its microstructured skin on its sides to achieve destructive interference at colors that could indicate the presence of the fish, while they also emit light at their bottom part to match its color and intensity to its surrounding, making it invisible from below, referred to as counterillumination. In this work, we borrow the two underwater camouflage strategies from marine hatchetfish, mimic them with photonic phenomena, and apply the camouflage strategies for physical stealth transmission of a 200 MBaud/s 16QAM OFDM secret signal at 5 GHz over a 25-km of optical fiber. The proposed bio-inspired steganography strategies successfully hid the secret signal in plain sight in temporal, RF spectral, and optical spectral domains, by blending in using counterillumination and turning invisible using silvering techniques. The stealth signal can only be retrieved with the precise and correct parameter for constructive interference at the secret signal frequency to unmask the silvering.
© 2021 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
Information security has always demanded new strategies to protect sensitive information since most aspects of our society from online social media, mobile computing, security service, health care service, smart city, to online banking are intensively connected using optical and radio frequency (RF) networks [1–3]. Massive number of sensitive, confidential, and personal information are being transmitted, processed, and stored every second; thus, networks and information have to be protected against malicious attacks to prevent attacker from intercepting or stealing the sensitive information. While encryption is mainly used to protect sensitive information, the presence of the information has not been concealed – meaning that the attacker can still see the presence of the encrypted information and it is just a matter of time for them to decrypt it through approaches like brute force attack. Therefore, effective cryptography requires two major components, encryption and steganography. Encryption scrambles the sensitive information so that it is unreadable without the key, while steganography hides the sensitive information within ordinary information to maintain its secrecy during transmission so that the attacker will not even know there is a signal to look for. Therefore, an effective cryptography scheme could ensure confidentiality, integrity, and authentication of information.
Fiber optics is the backbone of most communication networks that connects cities and nations across oceans and continents, as well as supporting radio-over-fiber transmission of mobile radio frequency signal (i.e. 5G and WiFi) [4,5]. Therefore, there is a critical need to physically secure the sensitive information during transmission in the data link using physical cryptography. Physical encryption techniques [6] have been studied intensively and are well developed; however, steganography in the physical layer [7–10] have always been overlooked. Physical encryption can be achieved using either optical or electronic approaches. In electronic scheme, chaos mapping can be done by masking [6] or scrambling [11] using digital chaotic signal in passive optical network (PON) at the physical layer. In optical scheme, encryption can be achieved via chaotic lasers [12–14], including nonlinear dynamics for high-quality chaos synchronization [15], semiconductor laser with optical feedback for key space enhancement [16], and double-random phase encoding in the fractional Fourier domain [17]. Electrical RF steganography [18] was proved successfully to hide the digitally modulated communication information via linear chirp radar signals; however, electrical RF steganography systems have a low bandwidth, their electromagnetic wave nature makes them easily suffering from electromagnetic interference and are vulnerable to steganalysis – detecting messages hidden using steganography. On the other hand, optical steganography is mainly done by spreading the stealth signal over an extremely wide optical spectrum (i.e. stealth optical pulse train from mode-locked laser with a wide spectrum [8]), such that the pulse amplitude would be low and can be buried underneath system noise when passing through a dispersive device. To allow the optical carrier to mimic and blend into the system noise, the optical pulse can also be processed to be a noise-like signal at the stealth transmitter using a combination of nonlinear spectral broadening, temporal spreading and power equalizing [19]. To tackle the challenges in conventional optical steganography, direct usage of broadband amplified spontaneous emission (ASE) noise [9] is thus a good candidate such that the stealth signal could be fully buried under the system noise with the same noisy signature. Transmission of optical stealth signal in optical fiber makes it very challenging for the attacker to pick up any signature of the stealth signal without being detected.
Turning to nature for solution, animals conceal their presence in their surroundings via camouflage – an extremely efficient way to ensure their survival. Borrowing camouflage strategies from animals could be an effective solution towards steganography in RF and optical systems. Among different types of camouflage, underwater camouflage is powerful because of the multi-dimensional concealment it can achieve. Underwater camouflage helps sea animals to hide from predators from above the water, being invisible from its side, and removing its dark appearance when seen from below. marine hatchetfish [20–22] has some of these powerful camouflage skills for survival – silvering and counterillumination, as illustrated in Fig. 1. Marine hatchetfish has microstructured skin on its sides to achieve destructive interference at colors that could indicate the presence of the fish, while constructive interference occurs at colors that is similar to their surroundings, this technique is known as silvering [Fig. 1(a)]. At the same time, marine hatchetfish also emits light from the bottom part of their body to match its color and intensity to its surrounding to make them invisible from below, referring as counterillumination [Fig. 1(b)].
Fig. 1. Illustration of the two camouflage skills in marine hatchetfish. (a) Side view (i) no camouflage - fish is visible (ii) silvering - fish is destructively interfered at colors that could indicate the presence of the fish; (b) Bottom view (i) no camouflage – fish appears darker against the bright water surface when seen from below (ii) counterillumination – fish illuminates itself to the same color and intensity as the background.
In this work, we borrow the two camouflage techniques from marine hatchetfish – silvering and counterillumination, mimic the techniques using photonics, and apply them as optical steganography techniques for stealth transmission of OFDM signal in radio-over-fiber networks. The mimicking of silvering – the use of destructive interference at the fish color to make it invisible to predators, is achieved using an optical finite impulse response (FIR) structure such that the stealth signal is destructively interfered to the attacker at any point of the transmission. While the mimicking of counterillumination – the production of light to match their backgrounds in both brightness and color, is achieved using a wideband low intensity light as the optical carrier to match with the background noise in the system. Unlike most existing optical steganography schemes, the proposed bio-inspired steganography does not just bury the stealth signal underneath system noise, but it also using destructive interference at the stealth signal frequency to make the stealth signal to be invisible and disappear in the attacker’s eyes.
2. Principle
Figure 2 is the design of the bio-inspired optical steganography scheme. To achieve counterillumination, a broadband ASE source generated from an erbium doped fiber amplifier (EDFA) is used as the optical carrier for the stealth OFDM signal to “illuminate” at the same wavelength and intensity as the background system noise, such that no distinct optical spectral component is observed. The broadband ASE source is spectrally sliced for achieving a desired photonic RF FIR in RF domain [23] when passing through dispersive medium for “silvering”, such that destructive interference at the stealth signal frequency (fs) is resulted in the attacker’s view, hiding the signal in both RF spectral domain and temporal domain. The constructively interfered frequency (fc) will be way above the frequency range of interest – i.e. nothing will be observed since there is no signal transmission at that frequency range. The stealth-modulated broadband optical carrier is combined with the public signal and system noise for transmission, resulting in a complete submerging of the stealth signal underneath the system noise, hiding the stealth signal in the optical spectral domain. The transmission fiber has a positive dispersion sign that will further move the constructively interfered frequency to an even higher frequency (fc+) as the signal propagates – much further away from the stealth signal frequency. In most optical network, dispersion compensating fiber with negative dispersion is placed at the last section of the transmission to correct any temporal spreading caused by the transmission fiber. The presence of the dispersion compensating fiber will move the constructive interference frequency back to fc. The attacker will not be able to observe any trace of the stealth signal in the optical spectrum, RF spectrum, or time domain at any point of the transmission. As a result, there is no reason for the attacker to attack the “empty” channel, successfully achieving steganography. At the intended receiver, a precisely matched dispersion is needed to achieve constructive interference at the stealth signal frequency fs, revealing the stealth signal. Due to the broadcast characteristic of PON, other public receiver will just treat the secret channel as system noise.
Fig. 2. Illustration of the proposed bio-inspired optical steganography for RF signal transmission over the fiber. (i) Silvering – photonic RF FIR creates destructive interference condition at the stealth signal frequency (fs); (ii) Transmission in optical fiber will only push the constructive interference condition to a much higher frequency (fc+); (iii) Dispersion compensation fiber at the last section of the transmission will move the constructive interference condition back to fc; (iv) Correct dispersion at the stealth receiver allows constructive interference condition to occur at the stealth signal frequency fs.
3. Experimental details
Experimental setup of the proposed bio-inspired optical steganography for OFDM-PON is shown in Fig. 3. Optical broadband source from an erbium doped fiber amplifier (EDFA) that covers the wavelength range from 1528 to 1568 nm is shaped to 20 nm wide and sliced by an optical wave shaper (Finisar 1000S). The dual channel 12 Gb/s arbitrary waveform generator (AWG, Keysight 8190A) is used to generate both the stealth signal and the public signal. The stealth RF signal is modulated onto the shaped optical source using a 12 GHz electro-optic intensity modulator (MZM). The stealth signal-modulated optical carrier is then launched to a dispersion compensation fiber (DCF1) to introduce proper time delay between taps for generating the photonic RF FIR. It is important to note that while we are using a 1.5 km DCF1 with dispersion coefficient of −255 ps/nm at the stealth transmitter in the experiment, the dispersion can be chosen to any value as long as the resultant FIR constructive interference condition occurs at a higher frequency than the stealth signal frequency. Meanwhile, the public signal is directly modulated onto a DFB laser output at 1553.33 nm. In our experiment, amplified spontaneous emission (ASE) noise is added to mimic the wideband system noise that are normally found in long-haul transmission system. The stealth signal, public signal, and system noise are combined and transmitted through a 25-km standard single mode fiber (SMF) followed by a 2.5-km DCF2 for dispersion compensation in a PON. A three-port thin film filter (TFF) with 3-dB bandwidth of 0.3 nm and center wavelength at 1553.33 nm is used to drop the public channel to the public receiver, leaving the hidden stealth signal for the stealth receiver. To observe the stealth signal, precise knowledge of the amount of dispersion is needed for constructive interference condition to occur at the stealth signal frequency, which is only known to the stealth receiver but not the attacker. After passing through DCF3 at the stealth receiver, the stealth signal is converted back to the electrical domain using a photodetector and is captured using a 128 GSa/s real-time sampling oscilloscope (Keysight UXR0334A Infiniium) with 33 GHz bandwidth for offline digital signal processing.
Fig. 3. Experimental demonstration of the proposed bio-inspired steganography scheme. BBS: broadband optical source; WS: optical wave shaper; MZM: electro-optic intensity modulator; DCF1-3: dispersion compensating fiber; ATT: optical attenuator; AWG: arbitrary waveform generator; ASE: amplified spontaneous emission; DFB: distributed feedback laser diode; SMF: single mode fiber; TFF: thin-film filter; PD: photodetector; OSC: real-time sampling oscilloscope.
To design the desired photonic RF finite impulse response with center frequency fFIR the corresponding optical comb should have a FSR ΔλFSR governed by the following equation:
where D, LDCF1, LDCF3 denote dispersion coefficient, the length of the DCF1 and DCF3, respectively. The center frequency (fFIR) will be right at the stealth signal frequency fs after passing through all the optical fiber (i.e. DCF1, SMF, DCF2, and DCF3). However, the FIR center frequency fFIR will be away from fs at any point after the transmitter and before the end of DCF3. The large key space provided by the total dispersion make it difficult to discover, search, or retrieve the stealth signal without the precise knowledge of the correct dispersion. Then, the overall shape of the optical comb is designed such that it is the correct RF filter profile at the stealth receiver for retrieving the stealth signal [24]. Therefore, the needed optical spectral shaping function can be expressed as:Fig. 4. (a) Relationship between constructive interference frequency (fFIR) and the designed FSR of the optical comb carrier at different fiber combinations; (b) Measured optical spectrum of the shaped optical comb carrier; (c) Tunable and reconfigurable constructive interference peaks at different frequencies.
In this experiment, our goal is to perform stealth transmission of a 200 MBaud/s 16 QAM-OFDM signal at 5 GHz. Optical comb carrier with spacing of 0.317 nm is used such that destructive interference condition is achieved at 5 GHz (stealth signal frequency) but a constructive interference condition at 13.22 GHz [solid red in Fig. 5(a)] at the transmitter. Constructive interference condition can only be achieved at 5 GHz (stealth signal frequency) when the accumulated dispersion is −680 ps/nm at the stealth receiver, as shown by the dashed brown line in Fig. 5(a). Thus, no stealth signal can be observed at the starting point of the transmission. As the stealth signal propagates in the SMF, the positive dispersion value of SMF will always reduce the total amount of dispersion. Thus, the constructive interference frequency will always be moved further up to a higher frequency fc+ (solid orange) while keeping the stealth signal invisible at any point along the SMF. During standard dispersion compensation in a PON or radio over fiber network, the amount of dispersion compensation will move the constructive interference frequency back to a slightly lower frequency (solid green and purple) but will not be enough to move it back all the way to the stealth signal frequency, keeping the stealth signal invisible at any location of the transmission. The constructive interference condition can only be moved back to the stealth signal frequency if the correct dispersion is used at the stealth receiver [dashed brown curve in Fig. 5(a)]. The noise floor at the RF spectrum is at −50 dBm and can be improved by 10 dB using balanced detection [25].
Fig. 5. Experimental results of the bio-inspired steganography scheme for the concealment of stealth signal. (a) State1(red): constructive interference condition occurs at 13.22 GHz after DCF1, State 2(Orange): positive dispersion of the SMF will move constructive interference condition to higher frequency; State 3(green/purple): DCF 2 is used to compensate the dispersion of SMF in public transmission nodes that moves the constructive interference peak to a slightly lower frequency; State 4(dashed brown): the matched dispersion at stealth receiver will shift the constructive interference condition to the stealth signal frequency at 5 GHz; (b) RF spectrum and constellation diagram measured during transmission without a correct stealth receiver (c) RF spectrum and constellation diagram measured at the stealth receiver with correct dispersion.
When the public signal is removed, we examined the received signal at the stealth receiver without applying a correct dispersion. The stealth signal is not visible in neither time domain nor RF spectral domain, as shown in Fig. 5(a) inset and Fig. 5(b), respectively. Furthermore, the corresponding constellation diagram as shown by the inset of Fig. 5(b) does not show any sign of the stealth signal – a noise like diagram is resulted. Even if the eavesdropper tries to tap into the transmission fiber directly at different locations of the transmission, no trace of the stealth signal can be observed because the 5 GHz stealth frequency is at destructive interference along the whole transmission [as shown by the solid color curves in Fig. 5(a)], just like the silvering camouflage technique that marine hatchetfish is using. At the stealth receiver, the stealth signal can be retrieved only if the correct dispersion is used, such that constructive interference condition occurs precisely at 5 GHz, i.e. the stealth signal frequency. Figure 5(c) shows the resultant constructive interference at 5 GHz (brown dashed) with the RF spectrum of the stealth signal shown in blue, and a clear constellation diagram shown in the inset.
Just like the marine hatchetfish where multiple camouflage skills are used to conceal its appearance in different view, the proposed optical steganography scheme also utilize a second stealth technique to conceal its appearance in the optical spectral domain too. Marine hatchetfish uses counterillumination to illuminate itself to the same color and intensity as the background such that it will not appear dark against the bright water surface when seen from below. Here, the use of incoherent broadband optical comb source acts like the counterillumination in marine hatchetfish, such that the stealth signal’s appearance is similar to the ASE system noise. The broadband optical comb source has successfully concealed the trace of stealth signal in optical spectral domain under the system noise, as shown in Fig. 6(a). In case the eavesdropper attempts to use coherent detection technique to detect the phase of the transmission that includes both the system noise and the stealth signal, the bandwidth of coherent detection is limited by the photodiode and analog-to-digital conversion capability, which is in the order of 100 GHz [26], corresponding to only 2% of the ASE noise and broadband comb source bandwidth used for the stealth signal, preventing the eavesdropper from digitizing or recording signal that is buried under the noise.
Fig. 6. (a) Measured optical spectra of the transmission with (blue) and without (orange) the stealth signal. (b) Simulated mesh plot of the BER change with respect to the dispersion mismatch and the stealth signal frequency.
The goal of steganography is to conceal the stealth signal such that the attacker will not know if there is a signal to look for. It is important to know whether it is easy to unintentionally get the dispersion close enough to unconcealed the stealth signal. Therefore, we study the dispersion tolerance of the proposed bio-inspired optical steganography scheme by investigating the relationship between BER, stealth signal frequency, and the dispersion offset, as shown in Fig. 6(b). To successfully demodulate a 16QAM-OFDM signal, a forward error correction (FEC) threshold of 1×10−3 is needed, as shown by the dark blue horizontal plane in Fig. 6(b). As observed, the dispersion needs to be within ± 10 ps/nm dispersion mismatch to retrieval the hidden stealth signal at 5 GHz correctly. Stealth signal with higher center frequency could increase the robustness of stealth transmission due to the tighter requirement of dispersion matching. The broadband characteristic of the optical comb makes the use of tunable dispersive device to monitor the stealth channel impossible. Commercially available dispersion-tunable compensation module usually either has very small bandwidth and dispersion, or has specific channel spacing, i.e. 0.8 nm. It would also be impossible for the eavesdropper to physically switch out wideband fixed dispersive medium to apply brute force attack to guess the dispersion. Furthermore, the estimation of dispersion using coherent detection could easily fail due to the wideband and incoherent nature of the broadband optical carrier that carries the stealth signal [27]. Therefore, the proposed bio-inspired steganography scheme is effective in terms of both prevention of stealth signal detection and inhibiting signal recovery.
While the concealing of stealth signal is important, the transmission performance of the stealth signal and the public signal are equally important. BER measurements and constellation plots of the public and stealth signals are investigated. Figure 7(a) shows the measured BER of the public channel with and without stealth signal and system noise, which indicates that only 0.2 to 0.4 dB power penalty is resulted when the stealth signal is added. The addition of system noise will add another 0.3 dB of power penalty to the public signal. Figure 7(b) shows the BER measurement of the stealth signal when different amount of system noise is added to the transmission for better concealment. In our experiment, it is observed that a power penalty is increased while still below FEC limit as the system noise is increased. Comparing the orange and blue curves, the presence of public signal has insignificant effect on the stealth signal BER. Constellation diagrams of the stealth signal at various received power are shown in Fig. 7(b-i)-(b-iii), indicating that the stealth can be retrieval successfully if BER is below the FEC limit.
Fig. 7. (a) BER measurements and constellation diagram of the public channel with and without stealth signal and system noise (ASE); (b) BER measurements and constellation diagrams of the stealth channel signal under different system noise power.
4. Summary
In summary, we presented a bio-inspired optical steganography scheme for enhancing the security of RF signal transmitting in radio-over-fiber and optical networks. Our approach borrows the silvering and counterilluminate camouflage strategies in marine hatchetfish, mimic it with photonics, and experimentally apply it in the stealth transmission of 16QAM OFDM signal. Silvering is achieved by the generation of destructive interference condition at the stealth signal frequency for the concealing of signal using photonic RF FIR, while counterilluminate is achieved by using a broadband optical comb source to blend in the wideband system noise. The proposed bio-inspired steganography scheme is an important part in cryptography when using with optical encryption. In our experiment, a 200MBaud/s 16QAM OFDM stealth signal at 5 GHz is successfully concealed in temporal domain, RF spectral domain, and optical spectral domain. Moreover, the proposed stealth transmission design could potentially simplify the remote node function in future 5G network and beyond for secure communication since the reveal of stealth is done physically without the need of high-speed digital signal process.
Funding
Directorate for Engineering (1653525, 1917043).
Disclosures
The authors declare no conflicts of interest.
References
1. N. Ghazisaidi, M. Maier, and C. M. Assi, “Fiber-wireless (FiWi) access networks: A survey,” IEEE Commun. Mag. 47(2), 160–167 (2009). [CrossRef]
2. B. Wu and Y. Ergaibi, “Optical stealth communication for smart sensor networks,” in 2017 IEEE Sensors Applications Symposium (SAS), (IEEE, 2017), 1–5.
3. N. Skorin-Kapov, M. Furdek, S. Zsigmond, and L. Wosinska, “Physical-layer security in evolving optical networks,” IEEE Commun. Mag. 54(8), 110–117 (2016). [CrossRef]
4. R. Waterhouse and D. Novack, “Realizing 5G: Microwave photonics for 5G mobile wireless systems,” IEEE Microwave Magazine 16(8), 84–92 (2015). [CrossRef]
5. G.-K. Chang and C. Liu, “1–100 GHz microwave photonics link technologies for next-generation WiFi and 5G wireless communications,” in 2013 IEEE International Topical Meeting on Microwave Photonics (MWP), (IEEE, 2013), 5–8.
6. B. Liu, L. Zhang, X. Xin, and J. Yu, “Constellation-masked secure communication technique for OFDM-PON,” Opt. Express 20(22), 25161–25168 (2012). [CrossRef]
7. M. P. Fok, Z. Wang, Y. Deng, and P. R. Prucnal, “Optical layer security in fiber-optic networks,” IEEE Transactions on Information Forensics and Security 6(3), 725–736 (2011). [CrossRef]
8. Z. Wang and P. R. Prucnal, “Optical steganography over a public DPSK channel with asynchronous detection,” IEEE Photonics Technol. Lett. 23(1), 48–50 (2011). [CrossRef]
9. B. Wu, Z. Wang, Y. Tian, M. P. Fok, B. J. Shastri, D. R. Kanoff, and P. R. Prucnal, “Optical steganography based on amplified spontaneous emission noise,” Opt. Express 21(2), 2065–2071 (2013). [CrossRef]
10. J. Lubacz, W. Mazurczyk, and K. Szczypiorski, “Principles and overview of network steganography,” IEEE Commun. Mag. 52(5), 225–229 (2014). [CrossRef]
11. L. Deng, M. Cheng, X. Wang, H. Li, M. Tang, S. Fu, P. Shum, and D. Liu, “Secure OFDM-PON system based on chaos and fractional Fourier transform techniques,” J. Lightwave Technol. 32(15), 2629–2635 (2014). [CrossRef]
12. G. D. Vanwiggeren and R. Roy, “Communication with chaotic lasers,” Science 279(5354), 1198–1200 (1998). [CrossRef]
13. A. Argyris, D. Syvridis, L. Larger, V. Annovazzi-Lodi, P. Colet, I. Fischer, J. Garcia-Ojalvo, C. R. Mirasso, L. Pesquera, and K. A. Shore, “Chaos-based communications at high bit rates using commercial fibre-optic links,” Nature 438(7066), 343–346 (2005). [CrossRef]
14. A. Uchida, Optical communication with chaotic lasers: applications of nonlinear dynamics and synchronization (John Wiley & Sons, 2012).
15. N. Li, H. Susanto, B. Cemlyn, I. Henning, and M. Adams, “Secure communication systems based on chaos in optically pumped spin-VCSELs,” Opt. Lett. 42(17), 3494–3497 (2017). [CrossRef]
16. D. Wang, L. Wang, Y. Guo, Y. Wang, and A. Wang, “Key space enhancement of optical chaos secure communication: chirped FBG feedback semiconductor laser,” Opt. Express 27(3), 3065–3073 (2019). [CrossRef]
17. G. Unnikrishnan, J. Joseph, and K. Singh, “Optical encryption by double-random phase encoding in the fractional Fourier domain,” Opt. Lett. 25(12), 887–889 (2000). [CrossRef]
18. Z. Zhang, M. J. Nowak, M. Wicks, and Z. Wu, “Bio-inspired RF steganography via linear chirp radar signals,” IEEE Commun. Mag. 54(6), 82–86 (2016). [CrossRef]
19. H. Zhu, R. Wang, T. Pu, T. Fang, P. Xiang, J. Zheng, Y. Tang, and D. Chen, “Experimental demonstration of optical stealth transmission over wavelength-division multiplexing network,” Appl. Opt. 55(23), 6394–6398 (2016). [CrossRef]
20. E. I. Rosenthal, A. L. Holt, and A. M. Sweeney, “Three-dimensional midwater camouflage from a novel two-component photonic structure in hatchetfish skin,” J. R. Soc. Interface. 14(130), 20161034 (2017). [CrossRef]
21. M. F. Land, “The physics and biology of animal reflectors,” Prog. Biophys. Mol. Biol. 24, 75–106 (1972). [CrossRef]
22. E. J. Denton, “Review lecture: on the organization of reflecting surfaces in some marine animals,” Phil. Trans. R. Soc. Lond. B 258(824), 285–313 (1970). [CrossRef]
23. Q. Liu and M. P. Fok, “Adaptive photonic RF spectral shaper,” Opt. Express 28(17), 24789–24798 (2020). [CrossRef]
24. Q. Liu, J. Ge, and M. P. Fok, “Microwave photonic multiband filter with independently tunable passband spectral properties,” Opt. Lett. 43(22), 5685–5688 (2018). [CrossRef]
25. X. Xue, X. Zheng, H. Zhang, and B. Zhou, “Spectrum-Sliced Microwave Photonic Filter With an Improved Dynamic Range Based on a LiNbO3 Phase Modulator and Balanced Detection,” IEEE Photonics Technol. Lett. 24(9), 775–777 (2012). [CrossRef]
26. P. Bower and I. Dedic, “High speed converters and DSP for 100G and beyond,” Opt. Fiber Technol. 17(5), 464–471 (2011). [CrossRef]
27. C. Huang, P. Y. Ma, B. J. Shastri, P. Mittal, and P. R. Prucnal, “Robustness of optical steganographic communication under coherent detection attack,” IEEE Photonics Technol. Lett. 31(4), 327–330 (2019). [CrossRef]
