Within the optical access space, we are excited to have a paper from Orange (“Coexistence in future optical access networks from an operator’s perspective,” Saliou et al.) giving an operator’s perspective on the practical evolution of optical access networks in a heterogeneous world in which both point-to-point and PON technologies have a part to play. Additionally, we have a review of optical access as applied to fixed and mobile applications (“High-speed reach-extended IM-DD system with low-complexity DSP for 6G fronthaul,” Zhu et al.), and coupled to this is a paper focused on the radio use case for optical access, showcasing the use of PON for CPRI fronthaul (“Efficient transport of enhanced CPRI fronthaul over PON,” Maes et al.). Then we have a contribution adopting the increasingly popular strategy of borrowing technology from data centers and applying it to broader networks: in this case high-speed PON (“Higher-speed PONs based on data center technology and optics,” Houtsma and van Veen). Finally, here, a paper exploring what is likely to become an area of increased interest, namely, the need for improved timing and synchronization for radio access networks (“Clock synchronizing radio access networks to picosecond precision using optical clock distribution and clock phase caching,” Clark et al.).

This Special Issue sees a substantial focus on an increased role for photonics in data center networks. First, there is an expanded tutorial surveying the potential role of optical switching in this environment (“Optical switching will innovate intra data center networks,” Sato). With the new AI era, heralded by ChatGPT and other language models starting to be used across the world, it is natural to ask what impact this will have on optical networks. Within this context, KDDI (“Modoru: Clos nanosecond optical switching for distributed deep training,” Wang et al.) have contributed a paper describing a specific optical switching architecture, optimized to address the distributed deep training that underlies many of these recent AI developments. There is also a proposal for a silicon photonic architecture to help speed up this deep learning process (“Flexible silicon photonic architecture for accelerating distributed deep learning,” Wu et al.). As well as assisting in the deployment of AI, optical networks will significantly benefit from it, and one such area is optical network automation, discussed in detail by Chalmers (“AI/ML-as-a-service for optical network automation: use cases and challenges,” Natalino et al.). Another area of benefit is in network optimization, and recent progress is ably summarized in a paper from NTT (“Reinforcement-learning-based path planning in multilayer elastic optical networks,” Tanaka) with wider attention to the multi-layer aspects as well as to cost and energy consumption and network throughput.

There are multiple other papers addressing a wider range of current hot topics within our industry so I am proud to see this Special Issue published as I am sure it will be the catalyst by which optical network technologies continue to make a difference to the world.