Optical-computing-enabled Network: A New Dawn for Optical-layer Intelligence?

Dao Thanh Hai,Minh Nguyen,Isaac Woungang

DOI: https://doi.org/10.1145/3663408.366582

2024-06-01

Abstract:Inspired by the renaissance of optical computing recently, this poster presents a disruptive outlook on the possibility of seamless integration between optical communications and optical computing infrastructures, paving the way for achieving optical-layer intelligence and consequently boosting the capacity efficiency. This entails a paradigm shift in optical node architecture from the currently used optical-bypass to a novel one, entitled, optical-computing-enabled mode, where in addition to the traditional add-drop and cross-connect functionalities, optical nodes are upgraded to account for optical-computing capabilities between the lightpath entities directly at the optical layer. A preliminary study focusing on the optical aggregation operation is examined and early simulation results indicate a promising spectral saving enabled by the optical-computing-enabled mode compared with the optical-bypass one.

Networking and Internet Architecture

What problem does this paper attempt to address?

The paper primarily explores an innovative network architecture concept, namely the Optical-computing-enabled Network, which aims to achieve optical layer intelligence by integrating optical computing functions at the optical layer, thereby improving network capacity efficiency. Specifically, the study attempts to address some key issues present in current optical networks:

1. **Introduction of Optical Layer Intelligence**: Existing optical network architectures have remained largely unchanged since the 2000s, primarily based on the optical bypass mode. In this mode, to prevent interference between different optical paths, they need to be separated in time, frequency, or space. This limits the flexibility and scalability of the network. Therefore, the authors propose a new architecture, the Optical-computing-enabled Network, where optical nodes not only have traditional add/drop and cross-connect functions but also possess additional optical computing capabilities.

2. **Improving Spectral Efficiency**: Through techniques such as optical aggregation, computational operations can be performed on multiple optical paths at the optical layer to generate new, more efficient optical paths. For example, two 400Gbps demands can be integrated into a higher-order modulation format optical path through optical aggregation technology, thereby reducing the number of wavelengths required and improving spectral utilization.

3. **Challenges of Network Design Algorithms**: To fully exploit the potential of Optical-computing-enabled Networks, new network design algorithms need to be developed. These algorithms must consider not only routing and wavelength assignment but also the allocation of computational tasks. The study proposes an integer linear programming model to address this issue, and preliminary evaluations show that in real-world network topologies, Optical-computing-enabled Networks can significantly improve spectral efficiency compared to traditional optical bypass networks.

In summary, the goal of this study is to explore how to achieve more flexible, efficient, and intelligent optical layer communication networks by integrating optical computing functions into optical networks.