Current Projects

1. Microwave photonics

This project uses photonic time stretch system to implement RF/microwave processing unit functionalities using photonic components and instruments. This can be used to improve the performance of conventional RF/microwave processing units or to decrease their cost.

2. Highly accurate time stretch vibrometry system  

This project is based on a novel idea to dramatically increase the resolution of conventional time stretch vibrometry system. This has application in biomedical imaging and remote sensing. It is based on using the movement of fringes in time stretch vibrometry system.

3. Amplitude and phase measurement in time stretch systems

It is important for a city to know about the quality of its roads. This is typically done by filming the road and then asking an engineer to watch the film and find and analyze the severity of the cracks. This is very time consuming task specially for long roads. In this project we develop a commercial-level software that can process the captured film and automatically find and analyze the severity of the cracks.

4. Development of a stable super-continuum laser

In this project, we plan to build a stable super-continuum laser for biomedical imaging and remote-sensing applications.

Previous Projects

Ultrafast Optical Coherence Tomography (2016)

Ultrafast techniques are very important to analyze fast transient phenomena as well as fast capturing in high throughput applications. This includes cancer cell detection from blood monitoring as well as finding new rare physical phenomena that happen in sub-ps time scales at the speed of light. In this project, we develop a new biomedical imaging system with unprecedented performance. This method uses optical coherence tomography combined with an ultrafast optical signal processing technique to capture and analyze images at the rate of million frames per second. This project will put a world record for the speed of such techniques.

Using raspberry-pi tower to analyze network-on-chip systems (2016)

Network-on-chip is a practical solution to make multi-core processing units more efficient. These techniques are already used in the recent smart phones. Since these techniques are fabricated using integrated circuits, it is difficult to analyze their performance before fabricating the design. In this project, we plan to develop a system that allows researchers to analyze the performance of Network-On-Chip systems without the need to fabricate the design. It is based on using a network of Raspberry-Pis (a popular low cost mini computer platform) to emulate behavior of a Network-On-Chip system. At the same time, the built Raspberry-Pi tower will act as a super computer.

Automatic crack detection on the road (2016)

It is important for a city to know about the quality of its roads. This is typically done by filming the road and then asking an engineer to watch the film and find and analyze the severity of the cracks. This is very time consuming task specially for long roads. In this project we develop a commercial-level software that can process the captured film and automatically find and analyze the severity of the cracks.

Research Interests

  1.  Photonic/microwave signal processing, optics in biomedical applications (cancer cell detection, ultrafast microscopy, etc), optical telecommunications, photonic integrated circuits, silicon photonics, photonic assisted wideband data conversion, and real-time spectroscopy.

  2. Dealing with big data in real-time high-throughput instruments, image/video processing, transmission and compression, Gigapixel imaging, edge detection, time/space-bandwidth product compression, pattern recognition in big data analysis.


Research contributions (click on each subject to jump):

A. Dealing with the Big Data bottleneck in photonic real-time  instruments

B. Detection of ultrafast transient and dynamical events in real-time and single-shot

C. Microwave photonics signal processing

D. Photonic signal processing


A. Dealing with the Big Data bottleneck in photonic real-time  instruments

The world today is awash in digital information captured by sensors or generated in computer simulations. Coping with the fire hose of digital information in this era requires developing more efficient methods to capture, sample and store data. In applications such as data communication, medicine and scientific research, communication signals and phenomena of interest occur on time scales too rapid and at throughputs too high to be sampled and digitized in real time. To give some examples, the record throughput of recently developed instruments such as MHz-frame-rate brightfield cameras, ultra-high-frame-rate fluorescent cameras for biological imaging, and wideband photonic-assisted analog to digital conversion has enabled the discovery of optical rogue waves, the detection of cancer cells in blood with sensitivity of one cell in a million and demonstration of ADCs with 1 Tera bits per second sampling rate. These instruments produce information in the order of one trillion bits of data per second which overwhelms even the most advanced computers. Detecting rare events such as cancer cells in a flow requires that data to be recorded continuously and for a long time, resulting in vast data sets. Dealing with such data loads requires new approaches to data capture, transfer, compression and analytics.

Anamorphic Stretch Transform (AST) is an analog optical data compression technique, that tackles the big data bottleneck in photonic real-time instruments. Inspired by operation of Fovea centralis in the human eye and by anamorphic transformation in visual arts, AST allows capturing of analog signals that would otherwise be beyond the digitizer’s resolution, and at the same time, it reduces the volume of the generated digital data. This capture and compression technique is achieved by reshaping the signal prior to sampling. Analog optical data compression with compression factors up to 6.2 times have been demonstrated without the loss of information.

Anamorphic Stretch Transform (AST):

Anamorphic Stretch Transform (AST) is a mathematical transform that can be used to engineer the time bandwidth product of the input signal. It is mathematically defined as follow:

where Ei(ω) is the input signal spectrum, and β(ω) is called the phase kernel of the AST.  Following figure shows the block diagram of a system incorporating the Anamorphic transform for time bandwidth product compression.


Fig. 1 The system block diagram for application of Anamorphic Stretch Transform for time-bandwidth compression of optical signals. (a) Real time measurement: In such system optical signal bandwidth is compressed to match the backend digitizer speed, and at the same time the volume of the data is reduced. After capturing and post-processing, input signal is digitally reconstructed by back propagation. 

Stretched Modulation (SM) Distribution:

The Stretched Modulation (SM) Distribution is a complex valued three dimensional plot that allows one to design time bandwidth engineering systems. It describes the information bandwidth and duration of the signal intensity after a transformation that is mathematically described by the kernel of SM. The Distribution can be written as:

where t and ω are time and frequency variables, Ei(ω) is the input complex-field spectrum and  β(ω) is the operation kernel describing the photonic operation performed on the signal.


Fig. 2. Stretched Modulation (SM) Distribution, is a mathematical tool to design and benchmark optical time-bandwidth engineering systems. At Time=0 (horizontal axis) the magnitude of SM function represents the intensity modulation bandwidth and its half-extent along the time axis is the record length. The top and bottom plots are qualitative and show the magnitude of SM Distribution of input and output signals in a system with sub-linear group delay profile. The Distribution shows how the time bandwidth product of the signal intensity can be engineered.

SM distribution can be used to design time-bandwidth engineering systems in both near-field (small dispersion) and far-field (large dispersion) regimes. Following figure shows a comparison of SM distribution for conventional time-stretch transform (i.e. a system with linear group delay) and AST operated on an arbitrary signal. The derivative of the AST phase kernel for this simulation is a sublinear function of frequency.

Comparison of the Stretched Modulation (SM) Distribution  for Time Stretch Transform (TST) and Anamorphic Stretch Transform (AST) systems in near-field and far-field regimes. TST is based on a filter with linear group delay profile whereas AST is based on using a filter with nonlinear group delay profile. Using SM distribution, systems for time-bandwidth product engineering can be designed.

Analysis of SM distribution suggests that for time bandwidth product compression, the derivative of the AST phase Kernel should be a sublinear function of frequency.

Time bandwidth product compression using Anamorphic Stretch Transform (AST) has has been experimentally proved. The experimental setup for proof of the concept experiments is shown bellow.

Experimental setup for time-bandwidth compression using Anamorphic Stretch Transform (AST).

For experimental demonstrations, AST was implemented using a designed nonlinearly chirped fiber Bragg grating. AST was designed to have a phase Kernel with inverse tangent derivative profile (sublinear function). Experimental results are summarized bellow.

Fig. 3 Experimental results for analog data compression using Anamorphic Stretch Transform. In this implementation, the signal bandwidth is compressed for 500 times so it can be captured using a slow digitizer. At the same time the resulting data size is reduced for 2.5 times.


B. Jalali, and M. H. Asghari, “Method for data compression and time-bandwidth product engineering,” International PCT application submitted Dec. 2013, PCT/US13/77969, Supporting provisional application No. 61/867,515; 61/867,519; 61/888,867, Dec. 2012.

M. H. Asghari and B. Jalali, “Anamorphic transformation and its application to time-bandwidth compression,” Applied Optics, Vol. 52, pp. 6735-6743 (2013).

M. H. Asghari and B. Jalali, “Experimental demonstration of optical real-time data compression,” Applied Physics Letters, Vol. 104, 111101, pp. 1-4 (2014).

M. H. Asghari and B. Jalali, “Warped time lens in temporal imaging for optical real-time data compression,” Springer publishing group, Chinese Science Bulletin journal, ISSN: 1861-9541, DOI: 10.1007/s11434-014-0352-0, pp. 1-6 (2014), INVITED.

B. Jalali, J. Chan, and M. H. Asghari, “Time bandwidth engineering,” Inaugural issue of OSA’s high-impact journal Optica, Vol. 1, pp. 23- 31, (2014), INVITED.

B. Jalali and M.H. Asghari, “Anamorphic Stretch Transform; putting the squeeze on big data,” Optics and Photonics News 25, 24 (February 2014).

H. Gao, M. H. Asghari, and B. Jalali, “Time-bandwidth engineering for arbitrary waveform generation,” IEEE Global Signal and Information Processing Symposium (GlobalSIP 2014), Accepted, Atlanta, December 2014.

J. Chan, M. H. Asghari, and B. Jalali, “Performance of time bandwidth engineering systems,” IEEE Photonic Conference (IPC 2014), Accepted, San Diego, October 2014.

M. H. Asghari, J. Chan, and B. Jalali, “Time-frequency manipulation in real-time instru-ments,” Progress In Electromagnetics Research Symposium (35th PIERS 2014), Accepted, August 2014, Guangzhou, China, INVITED.

J. Zhang, M. H. Asghari, J. Yao, and B. Jalali, “Time-bandwidth product expansion of microwave waveforms using Anamorphic Stretch Transform,” Conference on Lasers and Electro-Optics (CLEO 2014), paper: JTh2A.38, June 2014, San Jose, USA.

M. H. Asghari, and B. Jalali, “Anamorphic temporal imaging using a warped time lens,” Conference on Lasers and Electro-Optics (CLEO 2014), paper: STu3E.3, June 2014, San Jose, USA.

B. Jalali, and M. H. Asghari, “Anamorphic Stretch Transform for Analog and Digital Compression of Big Data,” Novel Optical Systems Design and Optimization, SPIE Optics-Photonics 2014, paper: OP14O-63, August 2014, San Diego, USA, INVITED.

M. H. Asghari and B. Jalali, "Demonstration of analog time-bandwidth compression using anamorphic stretch transform," Frontiers in Optics (FIO 2013), Paper: FW6A.2, Orlando, USA, POST-DEADLINE PAPER.

M. H. Asghari and B. Jalali, "Anamorphic time stretch transform and its application to analog bandwidth compression," IEEE Global Signal and Information Processing Symposium (GlobalSIP 2013) paper: NSSIMb.PD.2, Austin, USA.

M. H. Asghari and B. Jalali, "Warped dispersive transform and its application to analog bandwidth compression," IEEE Photonic Conference (IPC 2013), paper TUG 1.1, Seattle, USA.

Jump to top

B. Detection of ultrafast transient and dynamical events in real-time and single-shot

Stereopsis-Inspired Time-Stretched Amplified Real-Time Spectrometer (STARS)

Measurement of ultrafast single-shot events harbor a wealth of fascinating science that is inaccessible to pump-and-probe measurements. Real-time instruments fueling this field include high-throughput time-stretch imaging, real-time spectroscopy, high-throughput optical coherence tomography and wideband analog to digital conversion. One of the critical bottlenecks of the electronics is the speed of current analog-to-digital converters (ADCs). Current ADCs are limited by their speed to tens of GHz. They are also very costly and can produce a lot of heat. In contrast, photonic-assisted ADCs can operate at THz speeds, they are cheaper, produce much lower heat and require much lower power. As one of my contributions to this subject, I have recently introduced a photonic-assisted data conversion method, called STARS, to capture electrical signals using a digitizer with much lower sampling rate than required by Nyquist sampling rate theorem. STARS stands for Stereopsis-Inspired Time-Stretched Amplified Real-Time Spectrometer. To give an example I demonstrated capturing of 40 GHz streaming telecommunication signals using a digitizer with only 1.5 Gbps sampling rate. To use a conventional ADC to capture this signal, an ADC with 80 Gbps sampling rate would be required (based on Nyquist sampling rate theorem).

In this method (i.e. STARS) first the electrical signal to be digitized is modulated on an optical signal. STARS then employs amplified dispersive Fourier transform to slow down the modulated optical signal in time (to measure the signal using a low-speed digitizer). Finally our novel stereopsis reconstruction algorithm inspired by binocular vision in biological eyes is used to recover the signal in both amplitude and phase. The STARS system is shown in Fig. 4.

Fig. 4. Schematic of STARS: photonic-assisted analog-to-digital conversion. Electrical (RF) signal to be digitized is modulated on an optical signal, then amplified dispersive Fourier transform is used to slow  down the signal so it can be captured using a low speed backend digitizer. Our novel stereopsis reconstruction algorithm is finally used to recover the slowed-down signal in both amplitude and phase.

STARS benefits from a novel dynamic time-stretch concept to enhance the phase accuracy and dynamic range of the system more than 30 times. Fig. 5 shows the experimental demonstration of the huge dynamic range of STARS method to measure input signal's phase profile. STARS holds the world record for input signal phase dynamic range.

Fig. 5. Experimental demonstration of the huge dynamic range of STARS method to measure input signal's phase profile. Measured group delay profiles of the input signals with different group velocity dispersions (GVDs) ranging from -5 ps/nm to -4640 ps/nm (solid lines), compared to the corresponding simulated ideal group delay profiles (dotted lines).  Inset shows the reconstructed energy spectrum of the input signals.

We have employed this instrument to capture 40 Gbps electrical signal using a 1.5 Gbps digitizer. The experimental setup is shown in Fig. 6. Experimental results (see Fig. 7) confirms the capability of STARS to capture ultrafast electrical signal suing a low speed backend digitizer.

Fig. 6. Experimental setup to capture 40 Gbps electrical signal using a 1.5 Gbps digitizer. PC: Polarization controller, RF: Radio frequency.

Fig. 7 Experimental results confirming the capability of STARS to capture 40 Gbps electrical signal using a 1.5 Gbps digitizer.


Hossein Asghari and Bahram Jalali, "Stereopsis-inspired time-stretched amplified real-time spectrometer (STARS)",  IEEE Photonics Journal, Vol. 4, pp. 1693-1701 (2012), Invited. (pdf)

Coherent Dispersive Fourier Transform

Future progresses in a wide range of fields essentially depend on capturing ultrafast phenomena as they evolve in time. Also in many biomedical applications (e.g. cancer cell detection) or mass material analysis, high-throughput measurement methods are highly demanded. These demands call for development of measurement methods capable of capturing dynamic phenomenon in a high-throughput fashion. The capability of performing such advanced measurements is specifically important for applications in which random (non-repetitive), rapidly-changing ultrafast waveforms need to be fully characterized and evaluated. These include real-time monitoring in ultrahigh-bit-rate optical telecommunication, computing and information processing systems; testing of electronic and photonic materials, devices and sub-systems; and observation and analysis of a large variety of ultrafast dynamic events in physics, biology, chemistry etc. Photonic technologies can be employed to observe dynamic phenomena as they evolve in time at ultrahigh frame rates and ultrahigh throughputs.

We have demonstrated an optical measurement technique, called real-time spectral interferometry, capable of detecting dynamic optical phenomenon at 20 million frames per second. Optical measurement methods prior to our  technique had update rates of thousands of frames per second range, i.e. more than thousand times slower than our method. Since in our method the captured signals are recorded using a real-time oscilloscope, billions of these frames can be captured and analyzed for ultrahigh throughput detection.

Schematic of real time spectral interferometry is shown in Fig. 8. Our method is based on combining conventional spectral interferometry with a dispersion element having a large group velocity dispersion (GVD), e.g. using a linearly chirped fiber grating (LCFG) device. In conventional spectral interferometry the signal under test is interfered with a reference signal in an optical coupler. The resulted spectral interferogram is then measured using a spectrometer. Spectrometers are very slow equipments (thousand of frames per second). The dispersive element in our method is employed to operate real-time optical Fourier transformation, this way the resulted spectral interferogram after the spectral interferometry is mapped to the time domain so it can be measured using a real time oscilloscope. Oscilloscopes are much faster equipments than spectrometers, they can operate at millions of frames per second.

Fig. 8. Schematic of real time spectral interferometry. In this method spectral interferometry is combined with dispersive real-time Fourier transformation. Using this method THz bandwidth dynamic optical phenomena can be captured as they evolve in time at unprecedented frame rate of 20 million frames per second.

Proof of concept experiments setup is shown in Fig. 9(a). Experimental results are shown in Fig. 9(b) to (d). We demonstrated capturing of dynamical optical waveforms with TH bandwidth as they evolve in time at an unprecedented frame rate of 20 million frames per second.

Fig. 9. (a) Experimental setup for generating rapidly-changing optical phenomenon by intensity modulation of dispersed broadband pulses using an electro optic modulator (EOM) driven by a synchronized train of electronic pulses in which the DC bias level is rapidly swept (the bias is driven by a 1.6-MHz electrical sinusoids). Amplitude (b) and phase (c) time profiles of 30 rapidly-changing ultrafast waveforms as measured at the EOM output with frame rate of 20 million frames per second, expanding over a total duration of ~1.7μs. Results corresponding to the individual characterization of 3 of these ultrafast waveforms at the measurement times of 236.4 ns, 354.6 ns and 827.4 ns are plotted in (d).


Hossein Asghari, Yongwoo Park and Jose Azana, "Complex-field measurement of ultrafast dynamic optical waveforms based on real-time spectral interferometry," Optics Express, Vol. 18, pp. 16526-16538 (2010). (pdf)

Laser Focus World (LFW) magazine review: Click here

Jump to top

C. Microwave photonics signal processing

Microwaves are radio waves with wavelengths ranging from as long as one meter to as short as one millimeter, or equivalently, with frequencies between 300 MHz (0.3 GHz) and 300 GHz. Microwave is basically a field that fills the gap between electronics (<GHz frequencies) and optics (>300 GHz frequencies). Microwave has a wide range of applications including telecommunications, Radar, radio astronomy, navigation, heating and power, spectroscopy etc. Microwave signal processing addresses the devices that can process, measure or manipulate microwave signals for different applications. Microwave signal processing devices at speeds more than GHz become extremely costly, power hungry and challenging to fabricate. Microwave photonics signal processing is a novel field in which photonic technologies are employed to operate microwave signal processing. These photonic-based devices are much cheaper, easier to operate and they require much lower power. Here I provide two examples of my contributions to microwave photonics signal processing, (1) photonic-based microwave reconfigurable differential operators and (2) photonic-based microwave temporal integrator.

Microwave Reconfigurable Differential Operators Using Photonic Technologies

Differential operators are generally defined as systems that can virtually emulate any operation on the input analog signal by combining different derivatives of the input signal with different weights. To emulate time-variant operations a reconfigurable differential operator system is required. To give an example on applications of differential operator systems, they can model the behavior of  resistor-inductor-capacitor (RLC) processing circuits. Another example is to model microwave filter design methods using differential operators. If the differential operator systems is reconfigurable it can be used to implement any kind of filtering or microwave circuit using a single device. Unfortunately making such devices in the microwave domain are very challenging, costly.

In this work, we proposed and experimentally demonstrated the design of a reconfigurable microwave  arbitrary differential operator using photonic technologies. We showed that any differential operator can be implemented using the same programmable platform. To implement an arbitrary differential operation we use well-known discrete-time Euler's approximation so the differential operation can be represented by finite-difference-time-domain (FDTD) equations.

The experimental setup for our proposed method is shown in Fig. 10. Basically in our method the microwave signal is modulated on an optical signal, then it is processed in the photonic domain and finally is converted back to the microwave domain. To understand our method in more details, the microwave signal is first modulated on an incoherent light using an electro optic modulator (MOD). This way photonic methods can be used to operate the same microwave functionality but in the photonic domain. We operate the target differential operator functionality (based on FDTD equations) in the photonic domain using a dispersive element combined with a wavelength division multiplexing (WDM) system. Using the WDM system different frequency bands of the signal are separated and using a multi channel attenuator system different frequency bands are given different weights. This way an arbitrary differential operation can be emulated in the photonic domain. Finally the resulted optical signal is measured using a balanced photo detector (BD) to generate the output microwave signal. Note that using different weights for different frequency bands, arbitrary differential operator system can be emulated using the same platform.

Fig. 10. Experimental setup for the proposed method to implement reconfigurable microwave differential operators using photonic technologies. The spectrum of the incoherent light source is also shown on the left. MOD: electro-optic modulator. SMF: single-mode fiber. DEMUX: wavelength-division-demultiplexing filter. MUX: wavelength-division-multiplexer filter. Multi-ch attn: multi-channel attenuator. BD: differential balanced photoreceiver.

In the first experiment we targeted to use our method to emulate the operation of a microwave RLC circuit in the photonic domain. The schematic of the targeted RLC circuit is shown in Fig. 11. The RLC circuit was described as a linear second-order differential operator.

Fig. 11. Schematic of the equivalent RLC circuit emulated through the use of the proposed photonics platform.

The experimental results are shown in Fig. 17 confirming the proper operation of the designed microwave RLC with 40 GHz bandwidth (limited by the speed of electro-optic modulator).

Fig. 11. (a) Measured input microwave signal. (b) Measured output microwave signal (red curve) after the photonics-based microwave RLC circuit. A numerically calculated output is also shown with black curve.

In another experiment we implemented first to fourth-order microwave temporal differentiators with the same platform using our method. The processing speeds in tens of GHz was proved experimentally, see Fig. 12.

Fig. 12. Third and Fourth-order derivatives of a 70 ps input microwave Gaussian pulse operated using our method. (a) and (c) are spectral density profiles shaped according to the finite difference codes of the third and the fourth-order derivatives; (b) and (d) are numerical (solid dots) and experimental curves (black lines) of the derivative outputs. The inset of (b) shows the measured 70 ps input pulse waveform.


Yongwoo Park, Hossein Asghari, Robin Helsten, and Jose Azana, "Implementation of broadband Microwave arbitrary-order time differential operators using a reconfigurable incoherent photonic processor," IEEE Photonics Journal 2, 1040-1050 (2010). (pdf)

Hossein Asghari and Jose Azana, "Proposal and analysis of a reconfigurable pulse shaping technique based on multi-arm optical differentiators ", Optics Communications, Vol. 281, Issue 18, 15, pp. 4581-4588 (2008). (pdf)

Microwave Temporal Integrator Using Photonic Technologies

A temporal integrator is a block that gets an arbitrary time-domain signal at its input port and generates the cumulative temporal integration of this input waveform at the output port. Temporal integrators are fundamental devices for implementing a wide variety of signal processing operations of interest, e.g., in computing, control, and communication networks. Similar applications can be expected for temporal integrators in the microwave domain. Implementation of  microwave temporal integrators for speeds beyond GHz using microwave devices are very costly and challenging to fabricate. Photonic technologies can be employed to operate microwave temporal integrators with tens of GHz bandwidth with much lower cost and better power efficiency.

Before going to the details of our work I give some information about how temporal integrators work. It has been proved that temporal integration is nothing but convolution of the input signal with a unit-step function. Unit-step function is a signal that its value is 0 for t<0 and 1 for t>0. We have also proved that if the input signal is convolved with a square-like signal  (instead of unit-step function) the output of the convolution is the proper integration of the input signal but in a limited time window given by the duration of the square-like function. Remember that unit-step function is actually a square-like signal with unlimited time duration.

We have proposed and experimentally demonstrated a design for implementing a microwave temporal integrator offering high processing speed using photonic technologies. The experimental setup for our proposed method is shown in Fig. 19(a). Basically in our method the microwave signal is modulated on an optical signal, then it is processed in the photonic domain and finally is converted back to the microwave domain. To understand our method in more details, first the microwave signal is modulated on an incoherent  optical signal with square-like broadband spectrum. The modulated optical signal is then passed through a dispersive element with specific amount of group velocity dispersion (GVD). We have shown that when the modulated signal passes through the dispersive element, its time domain amplitude (which has the information of the microwave signal) is convolved with the envelope of the incoherent light spectrum (a square-like shape). Based on our discussions in previous paragraph, the output of the dispersive element is the proper integration of the input microwave signal in a limited time window given by the bandwidth of the incoherent light source.

To solve the issue of limited operation time window of this microwave integrator, we proposed to cascade this photonic-based time-limited microwave integrator with a discrete-time optical integrator in which its unit-time delay is equal to the operation time window of the time-limited microwave integrator. We have proved mathematically and experimentally that this way the operation time window of the integrator can go to infinity, i.e. ideal temporal integration. The experimental setup for this implementation is shown in Fig. 13(b).

Fig. 13. Conceptual diagram of the proposed ultrafast photonic intensity integrator design through illustration of its temporal impulse response.

We have successfully implemented and tested  this idea. Specifically we implemented a microwave temporal integrator with 36 GHz bandwidth using our photonic method. The measured frequency transfer function of the implemented integrator is shown in Fig. 14. Integration of some microwave signals with tens of GHz bandwidth is also shown in Fig. 15 confirming the accurate operation of this photonic-based microwave integrator.

Fig. 14 Frequency transfer function of the experimentally demonstrated a microwave temporal integrator using photonic technologies (solid line) compared to the ideal transfer function (circles).

Fig. 15 Experimentally measured temporal response (solid lines) of the proposed microwave photonics temporal integrator to two input microwave signals (dotted lines). Ideal outputs are plotted with circles.


Hossein Asghari, Yongwoo Park, and Jose Azana, "Photonic temporal integration of broadband intensity waveforms over long operation time windows", Optics Letters, Vol. 36, pp. 3557-3559 (2011). (pdf)

Jump to top

D. Photonic signal processing

Optical memory unit

High bit-rate telecommunications, switching networks and signal processing units with ever-increasing the required speed of processing and computing calls for novel data computing units operating at THz range speed. Unfortunately electronic solutions cannot operate at such ultrafast speeds. For example, commercially available memory units (see Fig. 5(a)) operate at the speeds of <10 GHz. In scientific laboratories speeds in tens of GHz range have been demonstrated. Photonic solutions in both analog and digital domains are promising candidates for ultrafast signal processing and data computing operations because they have the potential to operate on signals in the THz regime.

To give an example, we have proposed and experimentally demonstrated a photonic-based memory unit (see Fig. 16(b)) which is the fastest memory unit in the world, working at the speed of ~700 GHz. This is well beyond the speed of available electronic memory units.

Fig. 16 Photonics technologies can be employed to enhance data computing and signal processing systems in terms of speed and power efficiency, (a) conventional electronic memory unit operation (tens of GHz speed). (b) Our photonic-based memory unit (THz speed).

Our proposed photonic-assisted memory unit's operation is based on a conceptually novel design for a 1-bit optical memory unit using an ultrafast photonic time integrator. For proof-of-concept implementation we used the setup shown in Fig. 17. The set and reset pulse were emulated using a photonic pulse shaper (in this example using a free space interferometer). The ultrafast memory unit was build based on a novel design for an ultrafast photonic temporal integrator: cascading three-stage interferometers with a designed fiber grating. Experimental  results are shown in Fig. 18 confirming the proper operation of the proposed memory unit with an unprecedented switching time of ~1.4ps, i.e. memory speed of 700 GHz.

Fig. 17. Experimental proof-of-concept setup to implement a photonic-assisted memory unit.


Fig. 18. Experimentally measured normalized intensities of the input and output signals from the optical memory unit. Dashed lines: input Set (S) and Reset (R) pulses with different relative time delays; solid lines: corresponding output signals from the photonic memory unit. Inset in (c) is zoomed plot over the time interval between 64ps to 75ps.


Hossein Asghari and Jose Azana, "Photonic integrator-based optical memory unit," IEEE Photonics Technology Letters, Vol. 23, pp. 209-211 (2011). (pdf)

Hossein Asghari, Yongwoo Park, and Jose Azana, "New design for photonic temporal integration with combined high processing speed and long operation time window," Optics Express, Vol. 19, pp. 425-435 (2011). (pdf)

Laser Focus World (LFW) magazine review: Click here

Analog signal processing at the speed of light  

Photonic analog signal processing offers similar functionalities as in electronic signal processing but with speeds in 1,000 GHz range. To implement general and complex photonic signal processing circuits, basic photonic building blocks are demanded to be designed and demonstrated in photonic platforms. This is a primary step toward the practical realization of all-optical photonic signal processing circuits. These analog basic building blocks include photonic temporal integrator, differentiator and Hilbert transformer. A very relevant example of application of these fundamental devices is that of analog computing systems devoted to solving ordinary differential equations (ODEs). These equations play a central role in virtually any field of science or engineering, e.g., physics, chemistry, biology, economics, medical sciences, and the different branches of engineering. It is well known that linear ODEs can be solved in real time using a suitable combination of first- and higher-order temporal integrators, adders, and multipliers (amplifiers/attenuators) (see Fig. 19). The possibility of realizing these computations all-optically translates into potential processing speeds well beyond the reach of present electronic digital or analog computers.

Fig. 19. Application of photonic signal processing devices to operate analog signal processing at THz speed. In this example, an analog optical integrator is used to solve a first order ordinary differential equation (ODE) at THz speed.

We have pioneered the design and implementation of these photonic signal processing basic building blocks. Here I give more details about two of these devices, namely photonic temporal integrators and photonic Hilbert transformer.

1. Photonic Temporal Integrator

An Nth-order temporal integrator (where N = 1, 2, 3 . . . refers to the integration order) is a device that calculates the Nth cumulative time integral of an input signal. Photonic temporal integrators compared with their electronic counterparts can provide much higher processing speeds. Photonic temporal integrators have already been proposed for various interesting applications, including optical signal characterization, ultrafast pulse shaping and all-optical memories.

In some consecutive works, we have designed the first and higher-order integrators using fiber grating technology. We have also fabricated and tested these devices. The experimental setup to test these devices is shown in Fig. 20.

Fig. 20. Experimental setup for test the fabricated fiber Bragg gratings (FBGs) to implement first and higher-order photonic temporal integrators.

Experimental results are shown in Fig. 21. Accurate and efficient first and second-order temporal integrations of optical signals with 500 GHz bandwidth were successfully demonstrated using the fabricated fiber Bragg grating (FBG) devices.

Fig. 21 Experimental results confirming the operation of the fabricated first and second-order photonic temporal integrators. (a) Measured input signal. Measured output of the first-order (b) and second-order (c) photonic temporal integrators. Blue lines show the amplitude and green dotted lines show the phase profiles.  Numerically calculated integrals are shown with red circles.


Hossein Asghari, Chao Wang, Jianping Yao and Jose Azana, "High-order passive Photonic temporal integrators," Optics Letters, Vol. 35, pp. 1191-1193 (2010). (pdf)

Hossein Asghari and Jose Azana, "On the design of efficient and accurate arbitrary-order temporal optical integrators using fiber Bragg gratings ," IEEE/OSA Journal of Lightwave Technology, Vol. 27, pp. 3888-3895 (2009). (pdf)

Hossein Asghari and Jose Azana, "Proposal and analysis of a reconfigurable pulse shaping technique based on multi-arm optical differentiators ", Optics Communications, Vol. 281, Issue 18, 15, pp. 4581-4588 (2008). (pdf)

Hossein Asghari and Jose Azana, "Design of all-optical high-order temporal integrators based on multiple-phase-shifted Bragg gratings ," Optics Express, Vol. 16, Issue 15, pp. 11459-11469 (2008). (pdf)

Hossein Asghari and Jose Azana, "Proposal for arbitrary-order temporal integration of ultrafast optical signals using a single uniform-period fiber Bragg grating ", Optics Letters, Vol. 33, Issue 13, pp. 1548-1550 (2008). (pdf)

Laser Focus World (LFW) magazine review: Click here

2. Photonic Hilbert Transformer

A time-domain Hilbert transformer also referred to as a quadrature filter or a wide-band π phase shifter, is a device that calculates the Hilbert transform of an arbitrary input temporal signal (see Fig. 22). Hilbert transformers can be routinely implemented in the electronic domain, either as analog or digital filters, and they are fundamental devices for numerous applications, e.g. in communications, computing, information processing, signal analysis and measurement etc.. A similar range of applications could be expected for a photonic implementation of the Hilbert transformer, i.e. photonic Hilbert transformer (PHT), with the essential difference that such a device would enable processing signals directly in the all-optical domain and at speeds (operation bandwidths) well beyond the reach of electronic technologies.

Fig. 22 Schematic diagrams showing the spectral transfer function (a) and temporal impulse response (b) of an ideal Photonic Hilbert Transformer (PHT) (dotted curves). Our  proposed physically realizable PHT is hsown with solid curves.

We have proposed and demonstrated the first design for a photonic Hilbert transformer. We showed that a photonic Hilbert transformer can be implemented using a uniform-period fiber Bragg grating (FBG) with a properly designed amplitude-only grating apodization profile. Photonic Hilbert transformers capable of processing arbitrary optical waveforms with bandwidths in THz range can be implemented using readily feasible FBGs. The numerical results confirming the accurate operation of the designed photonic Hilbert transformer (PHT) are shown in Fig. 23.

Fig. 23. Spectral and temporal responses of a Photonic Hilbert transformer (PHT) based on a 2-cm long mid-strength uniform-period fiber Bragg grating (FBG) with the apodization profile plotted in the inset of (b): (a) Reflectivity as a function of optical frequency deviation (around 193THz); the insets show the phase change of the reflection FBG spectral response and corresponding amplitude in dB around the central frequency; (b) Envelope amplitude of the reflection temporal impulse response (solid curve), shown in normalized units [n.u.]. The impulse response amplitude of an ideal bandwidth-limited PHT is also shown (red circle dots).


Hossein Asghari and Jose Azana, "All-optical Hilbert transformer based on a single phase-shifted fiber Bragg grating: design and analysis ," Optics Letters, Vol. 34, pp. 334-336 (2009). (pdf)

Jump to top

Copyright 2018 Loyola Marymount University