The global data traffic has experienced an exponential growth with the rapid development of the Internet of Things (IoT) and the cloud computing. As the key elements of the worldwide communications infrastructure, optical communication systems face great challenges to tackle the upcoming “capacity crunch”. The photonic research community has expected the mode-division multiplexing (MDM) technology to bring the next big jump in spectral efficiency. In the emerging era, it’s increasingly demanded to control as many spatial mode channels as possible. However, the conventional on-chip mode manipulation techniques typically adopt mode-order oriented design strategy, inevitably leading to long development time and huge trail-and-test cost. Additionally, their possibly available mode orders are inherently limited by either the working principle itself or the fabrication technologies. Hence, they are far from efficient in many points of view, from design complexity to device scalability.
In a new paper published in Light Science & Application , a team of scientists, led by Professor Xuhan Guo and Professor Yikai Su from Shanghai Jiao Tong University, have developed a universal design framework for arbitrary on-chip spatial mode control based on the concept of metamaterial building blocks (BBs). By simply programming the topological layout of BBs with predefined mathematical formulas, arbitrary mode conversion and mode exchange can be implemented with uniform good performance. As such, their metamaterial BBs allow us to experimentally manipulate record high-order mode up to the 20 th for the first time. As an experimental demonstration of principle, its huge potential in high-capacity on-chip data communication has been verified with an 8-mode photonic integrated circuit, reaching an aggregate data rate of 813 Gb/s. The reported method and technique can support high-efficiency integrated photonic communication systems, and may boost the development of various information processing fields from optical sensing to nonlinear and quantum photonics.
The metamaterial BB supporting the TE 0 -TE 2 mode conversion and mode exchange is realized by exploiting fully-etched dielectric slots on a multimode silicon waveguide. By programming a set of BBs in a parallel arrangement with pre-defined formulas, the metamaterial waveguide can manipulate arbitrary high-order guided modes efficiently without any further optimization. These scientists summarize the operational principle of their BBs-based design framework:
“The beam shaping principle provides a straightforward route to control arbitrary high-order modes, however, its great potential has been previously constrained by the complicated implementations of the traditional Mach-Zehnder interferometer architecture or other trail-and-test approaches for different modes. We innovatively engineer dielectric slots, which can function as a power splitter and a phase shifter simultaneously, to realize the process of beam splitting, phase shifting and beam combining within an ultra-compact conversion region. So the metamaterial BBs are placed to the regions where “peaks” or “valleys” are located in the transverse field profile of the high-order mode to maximize the mode overlap between involved modes. By simple topological arrangement of the BBs using predefined mathematical formulas, arbitrary high-order mode can be effectively generated.”
“We believe the presented generic mode manipulation approach represents critical progress towards advanced control of more physical dimensions of optic carriers. More importantly, all necessary components for coherent communications have been demonstrated on integrated platforms so far including both coherent transmitters and receivers. The metamaterial BBs make it possible to efficiently manipulate record high-order modes in practice, therefore, it becomes extremely promising to obtain fully-integrated MDM communication systems of ultra-high spectral efficiency” they added.
“Hopefully, the novel concept of building blocks itself can be flexibly transferred to other waveguide platforms (InP, Si3N4, etc.) as well as other wavelength bands (O band, mid-infrared band, etc.). Besides, the fundamentals gained from our on-chip arbitrary spatial mode manipulation may provide inspirations to more versatile metamaterial-assisted building block designs and could open up fascinating opportunities for complex photonic functionalities previously inaccessible.
Light Science & Applications