July 15, 2022

Fiber-optic networks have now widely penetrated the society. At NEC, we became the first in the world to develop a prototype of “multi-core fiber”, which achieves even larger capacity for fiber-optic networks. First of all, why do networks need larger capacity? And what exactly is multi-core technology? Here, we interviewed researchers on the details.

Today’s optical fiber cannot support tomorrow’s society

― What is the reason for the urgency to increase the capacity of optical transmission systems?

de Gabory: Internet traffic has been doubling about every three years and is still rapidly increasing. When I was a student, the bandwidth of optical fiber was said to be infinite―but that’s an old story now. Now, we are already seeing the cap.
As 5G mobile traffic increases, naturally more optical fibers will be needed to accommodate that traffic. As autonomous driving and metaverses penetrate society in the near future, they will certainly drive up traffic even further. Larger capacity for optical transmission systems is now a pressing and proximate issue.
Then, how do we achieve larger capacity in optical transmission systems? As I mentioned before, we are about to see the limit in the transmission capacity of current optical fiber. The maximum transmission capacity of a single fiber is said to be around 100 Tbps at most—which means that current fiber optics cannot support the petabit-class traffic that will be necessary for the near-future society. A breakthrough is essential.
And that is where the idea of “multi-core” came in. Conventional optical fiber has a single “core" that goes through the center for refracting and propagating light. If multiple cores can be bundled into a single optical fiber, we can drastically increase traffic while maintaining the outer diameter of the optical fiber.
This multi-core fiber particularly has tremendous demand in the area of submarine cables. Submarine cables are loaded onto a ship to be laid out on the seafloor. Therefore, the larger the outer diameter of the cable is, the smaller the amount of cable that can be loaded on the ship, which means more round trips between ports and the open sea are required. This is indeed inefficient. NEC is among the world’s top three enterprises when it comes to submarine cables, so we were quick to take advantage of our unique expertise to conduct this research and lead the world.

Creating multiple “cores” in a single optical fiber

― Specifically, what kind of multi-core fiber research is being conducted in NEC?

Takeshita: There are two types of multi-core fibers: the “uncoupled type” and its evolved mode, the “coupled type”. At NEC, we study both.
The uncoupled multi-core fiber has the cores spaced so that no crosstalk occurs within the optical fiber. Currently, the world is working to standardize multi-core fibers with four cores. Delicate processing is required: passing four cores inside an optical fiber with a diameter of 125 microns. In a collaborative research project under the auspices of the Japan’s Ministry of Internal Affairs and Communications, NEC worked with OCC Corporation and Sumitomo Electric Industries, Ltd. and succeeded in developing a prototype submarine cable that houses uncoupled multi-core fibers for the first time in the world. This result has garnered attention and headlines across the media, including being reported on the NHK terrestrial news at the time. Our publication summarizing these results was also selected as the featured research paper at the 2022 Optical Fiber Communication Conference (OFC), which is a top international conference in the area of optical communications. We also received an invitation from the European Conference and Exhibition on Optical Communication (ECOC), which is on par with OFC, to give a lecture.
The advantage of uncoupled fiber is that it can be used with existing systems. If you break down the four cores embedded inside one fiber into four single cores, it is possible to reuse the transmitters, receivers, repeaters, and other systems. In this sense, we believe that this is a very effective approach as a step toward the coupled multi-core fiber system.

This cable was developed as part of the "Multi-core high-capacity optical transmission system technology" (JPMI00316), a technical issue of the "Research and development of innovative optical network technology for new social infrastructure" in the "Research and development project of ICT priority technology" of the Japan’s Ministry of Internal Affairs and Communications.

Tateno: There are still problems even when multi-core fibers can be used with existing systems. Since the number of cores will be quadrupled from one to four, the power required to operate the system will also increase correspondingly. In particular, submarine cables laid out across the ocean require extremely high voltages because the power supply to the repeaters must be continuously transmitted from land. Therefore, power consumption has been an issue even for conventional single-core fiber.
Therefore, we at NEC provide a total system design that reduces power consumption by adjusting the core layout inside the fiber and by using original technologies to recover the energy discharged from repeaters. Our strength is the capability to design the entire system with optimization in mind, reducing power consumption while maximizing performance.
In a research paper recently submitted to the top international conference ECOC and accepted for publication, we proposed a bidirectional optical submarine cable system by reversing the direction of transmission in each core of a multi-core fiber. We believe that this will enable a more efficient transmission even with uncoupled fiber.

Takeshita: Currently, Japan is leading the world in research on multi-core optical transmission technology. I think this is due to the significant contribution of the government taking the initiative in launching projects for the early formation of optical fiber related ecosystems in Japan. NEC will continue to conduct research in collaboration with various enterprises and research institutes.

Aiming for even higher capacity by leveraging our knowledge of wireless communication

― What kind of research is ongoing for coupled fiber?

Hosokawa: We are aiming for the commercialization of the coupled fiber around 2030, so we are currently working on slightly forward-looking research. Coupled fiber is a method in which a larger number of cores are bundled in a single fiber than the uncoupled multi-core fiber, and therefore, crosstalk must be taken into account. In the case of submarine cables, signals transmitted and intricately complexed over the long distance of about 10,000 kilometers must be decoded at the receiver side. This requires a more sophisticated signal processing than what is used for the current single-core fibers. Designing a circuit that can be implemented is another challenge: we are faced with limitations of the chip scale in semiconductor fabrication, power supply, and installation space. We are currently working on our research in collaboration with various partners.
One thing we are focusing on is the “MIMO (Multi Input Multi Output)” technology used in wireless communication. We are conducting research with the idea that this wireless technology, which split transmission among multiple antennas, could be applied to the decoding of signals transmitted in optical fiber. NEC has researchers who are involved in not only fiber-optic but also wireless communications such as Beyond 5G, so we are now gathering such expertise to work on this issue.

― What about future prospects and goals?

de Gabory: First, we aim to implement uncoupled multi-core optical transmission systems in society. Based on the world’s first prototype we created last year, we are moving forward with commercialization and implementation in combination with NEC’s previously implemented systems
As a medium-to-long term goal, we will work on research of coupled fiber in parallel. In particular, as Hosokawa mentioned, we will focus on research of advanced signal processing that can be included in LSIs, which is currently a bottleneck.

Go back to Featured Technologies