Beyond traditional electronics engineering

Beyond traditional electronics engineering: Microwave photonics for faster speed, wider bandwidth and lower-loss signal handling

In the current digital era, the flow of information and data—be it wireless communication, computing or sensing—essentially hinges on signals. Scientists are continually finding ways to make signal transmission and processing faster, more efficient, and with lower loss. The surge of artificial intelligence (AI) has driven unprecedented demand for specially designed microchips, as these devices play a key role in achieving super-speed signal transmission and processing.

Microwave photonics (MWP), an interdisciplinary field that has gained significant traction in the past three decades, is opening new avenues in super signal technology. In the field of MWP, a novel kind of microchip known as the “photonic integrated circuit” (PIC) has enabled drastic improvement in communication and sensing capacities. Like the electronic integrated circuits (EICs) embedded in smart phones, computers and electronic devices, PICs integrate multiple components onto a single chip. However, the fundamental principles on which the two chips operate are significantly different. PICs use photons (light) and photonic components to generate, transmit, control and process microwaves. These capabilities position PICs as an alternative to EICs, which rely on electrical signals (electrons) and electronic components.

With the significant progress made in chip integration, the next major question facing photonics scientists and companies is the implementation of PICs in a wide array of MWP systems.

In this Issue, we chat with PAIR Senior Fellow Prof. YAO Jianping, one of the early pioneers in MWP research, about the development trends and challenges in the field, the attractive functions provided by MWP technologies, and the implications they hold for AI development. Prof. Yao is a Distinguished University Professor and the University Research Chair in Microwave Photonics at the University of Ottawa, Canada.

 

Microwave photonics: Where microwave engineering and photonic technology meet

“Microwave photonics (MWP), an interdisciplinary field that has gained significant traction in the past three decades, is opening new avenues in super signal technology”

Microwave photonics (MWP) combines microwave engineering and photonics engineering to bring forth unique properties that cannot be achieved in traditional electrical engineering. What are some of the major competitive advantages provided by MWP?

Wide bandwidth and low loss are crucial for fast and uninterrupted data transfer, which in turn is extremely important for digital electronics, broadband wireless communication and radar. In microwave engineering, bandwidth is a critical factor in the speed of signal transmission and processing. MWP employs photonics technologies to enable a much wider bandwidth for microwave signal transmission and processing. Other powerful advantages of MWP are small weight and low loss. For example, a 10 km copper cable is so heavy that it needs to be transported by trucks, and its signal loss is a few hundred per kilometer. Optical cable, by contrast, is very thin and light, to the extent that it can be carried by humans, and its loss is only 0.2 dB per kilometer. With optical fibre, long-haul underwater communication across continents has become possible.

Given the very attractive functional properties that MWP offers, are these technologies very expensive? How are they being applied in real-world settings?

Photonic devices were not so integrated in the past. Thirty years ago, before the concept of photonic integration, circuits were composed of separate optical devices like lasers, modulators and photodetectors. These “discrete components” were very expensive. For example, a high-speed modulator, a kind of optical component for converting electrical signals into optical signals, may cost a few thousand dollars or more. At that time, scientists needed to design and develop different architectures to implement different specific functions, like microwave signal generation or filtering, but their applications were limited due to the high cost. In addition, systems based on discrete components are heavy and have poor stability.

With the fast progress of photonic integration technology, MWP systems can now be implemented based on photonic integrated circuits (PICs). PICs contain multiple individual components assembled within a chip, thus significantly reducing the cost and improving the stability. Currently, the major task is to implement the MWP systems using PICs.

 

Transforming random, noisy signals into clearer ones: When signals become tunable and frequency-consistent

MWP emerged in the 1990s, and the Microwave Photonics Research Laboratory (MWPLab) you established in 2002 is among the early research organisations in the field. Your group was the first to demonstrate novel devices, including the Fully Reconfigurable Photonic Integrated Signal Processor and the Parity-Time Symmetry Optoelectronic Oscillator, and has developed other devices such as the Integrated Parity-Time Symmetric Wavelength-Tunable Single-Mode Microring Laser and the Photonic Integrated Field-Programmable Disk Array Signal Processor. These innovations are built on the concepts of “tunability” and “parity-time symmetry”. In what ways are these capabilities important for optimised signal processing?

Just like the wave spectrum, a signal contains multiple frequencies. A major role of signal processors is to filter out certain frequencies to produce a signal containing a preferred and desired frequency range. When we make these devices or processors “tunable”, it becomes possible to handle different types of signals and remove interferences or noise that compromise signal quality.

“For applications like wireless communications and radar, microwave signals with low phase noise are essential to overall system performance.”

Conventional electronic oscillators produce relatively high phase noise because they have a smaller quality factor (Q factor). A smaller Q factor suggests that a system has higher energy loss, resulting in a generated signal with higher phase noise. For applications like wireless communications and radar, microwave signals with low phase noise are essential to overall system performance.

 “Parity-time symmetry” is an important technical concept for optoelectronic oscillators, which are a kind of MWP device for generating microwave signals at a high and single frequency. Through “parity-time symmetry”, a single oscillation frequency can be selected without using an ultra-narrow band filter, thus greatly simplifying the implementation.

 

Accelerating and leveraging supercomputing: Bilateral interactions between AI and microwave photonics

AI technologies heighten the need for hardware that supports real-time processing and large-scale data analysis. PICs stand as a powerful solution for AI-related tasks since they enable ultra-fast processing speed and low power consumption. It is clear that MWP can support AI development. Conversely, how does AI benefit the development of MWP?

Microwave photonic systems can accelerate the computational speed of neural networks—a foundational architecture powering modern AI systems. For example, the computational speed based on MWP can be three orders of magnitude faster than electronics, and thus MWP could complement traditional digital processors, such as graphic processing units (GPUs), in specialised AI tasks. On the other hand, MWP systems can also benefit from AI. For example, microwave photonic signal transmission can be optimised with AI. Signal processing requires the use of filters. With AI algorithms, filter design can be optimised to make the identification of signal deviations and subsequent processing more rapid and efficient.

 

Going beyond computing and communication: Microwave photonics for sensing technology applications

MWP systems find diverse practical applications. In addition to telecommunication and computing, what other fields or sectors can benefit from MWP innovations?

In the past, MWP developments were mainly for radar and wireless communication. In modern times, when MWP innovations are applied in the context of sensing and networking, the technology can find useful applications in diverse fields. Distributed sensing is one powerful example. It is a sensor network system which includes multiple sensors that are spatially scattered to obtain measurements from various locations and transmit data to a central processing unit through connected thin, flexible optical fibre.

This distributed sensor system can be particularly useful for health monitoring, especially today when cities are striving to build smart-ageing environments. For instance, the system can be installed in care homes for monitoring the health of senior residents. A central computer collects health data such as body temperature and heart rate from the sensors worn by elders living in different rooms, and it will notify healthcare workers if the system detects anyone who is feeling unwell.

This MWP-enabled sensing technology can also be used for infrastructure monitoring, such as assessing the structural health of bridges and buildings. It can also be employed in advanced manufacturing: sensors installed along a production line alert manufacturers if there are any machine malfunctions or quality problems.

 

A closer look at the global photonics race

Photonics is a burgeoning industry. Countries around the world are competing to develop silicon photonics to meet the growing demand for high-speed data transmission and advanced sensing, and to drive new application areas such as smart healthcare and quantum computing. China is now leading the photonics market, followed by Europe and the United States. In your view, what are some factors that account for a country’s competitiveness in the photonics race?

I think the efforts of the people working in the field are a major reason for the country’s photonics development. This is evidenced by the sizable number of Chinese photonics entrepreneurs who are recognised by Forbes; the groundbreaking work on optical fibre by Nobel Prize-winning scientist Charles K. KAO, who was ethnically Chinese; and the cutting-edge research conducted by Chinese researchers in the photonics field.

Education, government support, and industry are also important. According to my experience as a university professor, graduate students from China are very strong in fundamental mathematics and physics, and both disciplines provide a solid foundation for understanding and applying concepts in photonics. The government is very supportive of entrepreneurship through policies and incentives, which have helped the establishment of photonics start-ups.

 

Facing the talent shortage in photonics and interdisciplinary research

Lasers and photonics are very specialised fields. The industry struggles to find suitable talents because photonics is a relatively young field that has not attracted much interest from students as compared to other subjects. The Hong Kong government is now proactively developing the city’s photonics industry, noting the need to step up efforts to attract and retain talented workers. At PolyU, PAIR is reviewing the Academy’s strategies aimed at attracting young talents for interdisciplinary research. In your view, what can be done by local authorities and universities to address the talent shortage?

“Hong Kong universities can set up schemes such as fellowships and subsidies to attract students from overseas institutions to join them and conduct interdisciplinary research.”

If the government can attract some international companies to set up branches and offices in the city, the new job opportunities will become a significant draw for talents. Hong Kong universities can set up schemes such as fellowships and subsidies to attract students from overseas institutions to join them and conduct interdisciplinary research. However, an evaluation mechanism is needed to ensure that they are working on research ideas that are different from their PhD disciplines.