中心博士生吕婷同学的工作——A Low-Redundancy Programmable Optical Delay Line Network Based on Dynamic Allocation of Optical Delay Resources（基于动态光延时线资源分配的低冗余可编程光延时线网络架构）的相关成果被IEEE Photonics Conference 2025接收为Oral presentation。
现有的可编程光延时线网络（PODLN）级联架构存在严重的资源冗余问题，具体表现为延时线无法共享、多通道延迟组合受限。
本研究提出一种基于动态资源分配的低冗余可编程光延时线网络架构，通过光开关调度实现延时线跨通道复用，有效降低了延时线需求数量并显著提高了可实现的最大通道间延迟差（ΔDmax）。具体设计了两通道及三通道基础架构，并通过共享延时模块级联扩展实现了更大规模的多通道延时网络。
实验验证表明，所提出的架构相比传统方法显著提高了延迟组合能力（ΔDmax提升超过2倍），并有效减少了延时线资源成本（6通道和15通道分别节省13.7%和24.7%）。该研究成果为光通信、微波光子处理、量子光学等领域提供了一种高效、经济且具有良好扩展性的光延时网络解决方案。
基于上述基础单元架构，本工作进一步实现了多通道扩展，通过将三通道单元级联、串联构建出更大规模的延时网络。例如，通过引入共享延时模块并配合波分复用（WDM）技术构建的九通道架构中，ΔDmax可高达63σ，显著高于传统级联架构。此外，实验对比分析显示，该方法能够在大幅提高ΔDmax（提高2倍以上）的同时显著降低延时线硬件成本。以具体实例验证，该架构在6通道与15通道应用场景下，分别实现了约13%与24.7%的延时线资源节省，验证了本工作提出的低冗余架构在资源效率和延时组合能力上的卓越优势。
摘要:
We propose a programmable optical delay line network architecture based on dynamic allocation of optical delay resources, which provides a greater diversity of programmable delay combinations among multiple channels and reduces the required number of delay lines. By scheduling optical switches, optical delay lines are reused across different channels, reducing redundancy of optical delay lines of previous architectures. Results demonstrate that, compared with state-of-the-art cascade architectures, the proposed one achieves more than 2-fold increase of maximum inter-channel delay difference at lower delay line cost. Our architecture is thus significant advantageous in applications such as optical beamforming, optical computing, programmable microwave photonic processors, and quantum circuits.
Conclusion: We proposed a low-redundancy PODLN architecture featuring dynamic delay allocation. Due to the reuse of delay lines, the proposed PODLN need less delay line resources while obtain better programmability and scalability. The proposed 2/3-channel architectures can be extended to arbitrary channel numbers and delay ranges. Results show lower hardware cost and superior performance versus state‑of‑the‑art designs, making it a strong candidate for next‑generation efficient photonics.