Integrated Photonics and Millimeter-Wave Systems for Next-Generation Communication Networks 5 views

Next-generation Terabit-per-second (Tbps) wireless systems operating at millimeter-wave (mm-wave) frequencies beyond 100 GHz are essential to support the increasing demand for ultra-wideband communications. Pure electronic solutions for carrier generation, modulation, filtering, and multiplexing at these high frequencies face significant challenges. Photonics offers substantial advantages by leveraging wide optical bandwidth devices for high-speed modulation and multiplexing. Techniques such as optical heterodyning enable stable mm-wave and sub-THz carrier generation. Despite these benefits, most demonstrated systems rely on discrete optical components, resulting in bulky and non-scalable implementations. A photonically driven mm-wave platform leveraging recent advancements in integrated photonics can address these challenges by combining multiple functions onto a single photonic integrated circuit (PIC) chip. This approach enables compact, low-cost, and scalable Tbps wireless transmitters for next-generation communication networks. In this work, I demonstrate a compact, highly integrated, multi-channel, photonically driven mm-wave transmitter operating from the upper K to W bands (23.7-110 GHz). The system achieves one of the highest integration densities among wireless transmitters by integrating multiple functionalities onto a single silicon photonics chip. The transmitter supports on-chip modulation up to 12 Gbps and on-chip detection up to 40 GHz, as well as off-chip detection and wireless transmission up to 110 GHz. The architecture is flexible and supports higher frequency bands in the sub-THz and THz regimes, enabling compatibility with current 5G, emerging 6G, and future wireless standards. Beyond communications, PICs are increasingly being explored for artificial intelligence (AI) and machine learning (ML) accelerators, neuromorphic computing, and high-performance computing (HPC) platforms. Low-loss interconnects and ultra-wideband photonic components, combined with the inherent massive parallelism of photonic architectures, enable highly dense computational frameworks such as wavelength-division multiplexed neural networks and interferometric matrix-vector multipliers. In this work, an algorithmic framework for demonstrating a Mach-Zehnder interferometer (MZI) mesh-based optical neural network (ONN), supported by initial experimental results, is presented.

PHD (Doctor of Philosophy)

Photonic Integrated Circuits; mm-Wave; Wireless transmitters; Silicon photonics; Optical Neural networks; Heterogeneous integration

English

All rights reserved by the author (no additional license for public reuse)

2026-04-23