Optical-layer Intelligence Enables More Capacity from Less Spectrum in Networks

The increasing demand for data transmission drives the search for more efficient network architectures, and a team led by Dao Thanh Hai from RMIT University Vietnam, Shuo Li, and Isaac Woungang from Toronto Metropolitan University now proposes a fundamentally new approach to network design. Their work introduces the concept of an optical-communication integrated network, a system capable of both transmitting data and performing computations within the network itself. This innovative architecture exploits the potential for processing information as light signals travel through the network, effectively adding computational power at no extra cost, and achieving greater efficiency in spectrum usage. By demonstrating the benefits of this ‘optical-layer intelligence’ through a detailed case study and performance analysis on a realistic network model, the researchers pave the way for a future where networks are not just conduits for data, but active participants in information processing.

Optical Computing Enables Future Network Connectivity Scientists are pioneering a new approach to optical networking, moving beyond simply transmitting data to actively computing with it within the network itself. This innovation, termed optical-computing-enabled networks, integrates computing functions directly into the optical layer, promising significant improvements in speed and efficiency for demanding applications like artificial intelligence and machine learning. By performing computations at the speed of light, networks can process data more effectively and reduce overall traffic load, paving the way for sustained connectivity and new capabilities. This new paradigm shifts the focus from traditional optical networks, which act as passive conduits for data, to intelligent networks capable of processing information as it travels. A key element of this approach is optical aggregation, a technique where multiple optical signals are combined into a single, higher-order signal, effectively reducing the amount of data that needs to be transmitted.

The team conducted a thorough review of existing technologies, including silicon photonics and optical computing, to establish a foundation for this new approach. They identified key components, such as optical transceivers and processors, and developed a mathematical framework to optimize network design. Simulations using realistic network topologies demonstrate that aggregation-aware routing achieves significant improvements in spectral efficiency, highlighting the potential to address the rapidly growing demands of internet traffic. This work builds upon recent advancements in dual-service optical networks and lays the groundwork for future research into more sophisticated optical computing operations. Scientists envision networks that can not only transmit data but also perform complex computations, unlocking new possibilities for data processing and analysis. The research identifies several key areas for future investigation, including the development of new optical devices and intelligent control algorithms.

Optical Computing Within Communication Networks Scientists have developed a novel optical network architecture that simultaneously delivers both computing and communication services at the optical layer. This approach, termed optical computing-communication integrated network, fundamentally reshapes the boundaries between these two functions by enabling optical computing between lightpaths as they traverse intermediate nodes, thereby reducing effective network traffic. The core innovation lies in introducing optical-layer intelligence, the capability of optical nodes to perform computations at the speed of light, achieving greater spectral and/or computing efficiency. To demonstrate this concept, researchers focused on optical aggregation as a specific computing operation. Unlike traditional optical networks where lightpaths are isolated, this new architecture harnesses the potential for operations among lightpaths sharing a common node. The study pioneers a method where nodes actively process signals from multiple incoming lightpaths, combining them into fewer outgoing lightpaths, effectively reducing the overall traffic load. This aggregation process, performed entirely in the optical domain, eliminates the need for energy-intensive optical-electrical-optical conversions.

The team formulated a mathematical framework to optimize the design of networks incorporating optical aggregation. This involved developing an integer linear programming model to determine the optimal placement of aggregation points and the corresponding routing and wavelength assignment strategies. Simulations using the realistic NSFNET topology demonstrate that aggregation-aware routing achieves significant improvements in spectral efficiency compared to conventional routing methods, highlighting the potential of this architecture to address the growing demands of internet traffic.

Optical Computing Boosts Network Efficiency This research presents a new architectural paradigm for transport networks, termed optical computing-communication integrated network, capable of providing both computing and communication services at the optical layer. The core innovation lies in performing operations between lightpaths that share intermediate nodes, introducing “optical-layer intelligence” to achieve greater spectral and/or efficiency. A key example of this is optical aggregation, where multiple lower bit-rate channels are combined into a single higher-capacity stream within the optical domain. Consider two traffic demands, each requesting 400 Gbps from nodes A and B to a common destination, node C. In a traditional optical network, provisioning these demands requires independent routes and wavelengths for each lightpath. However, with the new approach, optical aggregation at an intermediate node X combines the two 400G signals into a single 800G signal on the same wavelength, demonstrating significant savings. This effectively increases the bit rate by combining signals, reducing the overall spectral cost. Experiments reveal that this optical computing-communication integrated network model introduces architectural flexibility by allowing controlled interference between lightpaths, unlocking potential for improved spectral efficiency.

The team formulated a mathematical model to minimize wavelength link cost, or equivalently, maximize aggregation benefits, given a network topology and demand matrix. This model identifies optimal routing paths, suitable demand pairs for aggregation, and corresponding aggregation node locations.

Lightpath Computation Boosts Network Efficiency This research presents a new architectural paradigm for future optical transport networks, termed communication-integrated networks, capable of simultaneously providing communication and computation services at the optical layer. The core innovation lies in performing operations between lightpaths that share the same intermediate nodes, introducing the concept of layer intelligence, processing data at the scale of individual lightpaths to improve spectral and overall efficiency. A detailed case study focusing on optical aggregation demonstrates how this approach can reduce wavelength usage compared to traditional optical networks. The study highlights the potential for fiber-optic networks to evolve beyond simple data conduits and become active platforms for in-network computation, addressing the growing demand for efficient processing of large datasets. By strategically deviating from the shortest path, the new approach achieves spectral savings without compromising network performance. While the current work focuses on optical aggregation, the authors acknowledge that further research is needed to develop integrated optimization frameworks capable of jointly managing both computation and communication resources within the optical domain. Future work could extend the applicability of this architecture from long-haul networks to access and data center environments, leveraging the maturing capabilities of optical computing platforms.

This research represents an initial step towards realizing the full potential of optical-layer intelligence and its impact on future network design. Scientists envision networks that can not only transmit data but also perform complex computations, unlocking new possibilities for data processing and analysis. 👉 More information 🗞 More Capacity from Less Spectrum: Tapping into Optical-layer Intelligence in Optical Computing-Communication Integrated Network 🧠 ArXiv: https://arxiv.org/abs/2512.15190 Tags: