🤖 AI Summary
To address bottlenecks in reconfigurable optical networks—including degraded worst-case throughput, high congestion-control overhead, and deployment complexity—caused by multi-hop routing, this paper proposes a hop-free, periodically reconfigurable optical network architecture. Our method introduces (1) a novel short-period traffic-aware scheduling mechanism, coupled with matrix rounding to generate near-optimal periodic lightpath tables; and (2) a seamless periodic table-switching strategy that handles residual traffic, eliminating reconfiguration gaps and obviating dependency on congestion control. We theoretically prove a ≥33% improvement in worst-case throughput. Experiments under typical datacenter workloads demonstrate microsecond-scale end-to-end latency, strong determinism, and significant performance gains over baselines such as RotorNet. The architecture fully circumvents the performance degradation and operational complexity inherent to multi-hop routing.
📝 Abstract
The increasing gap between datacenter traffic volume and the capacity of electrical switches has driven the development of reconfigurable network designs utilizing optical circuit switching. Recent advancements, particularly those featuring periodic fixed-duration reconfigurations, have achieved practical end-to-end delays of just a few microseconds. However, current designs rely on multi-hop routing to enhance utilization, which can lead to a significant reduction in worst-case throughput and added overhead from congestion control and routing complexity. These factors pose significant operational challenges for the large-scale deployment of these technologies. We present Vermilion, a reconfigurable optical interconnect that breaks the throughput barrier of existing periodic reconfigurable networks, without the need for multi-hop routing -- thus eliminating congestion control and simplifying routing to direct communication. Vermilion adopts a traffic-aware approach while retaining the simplicity of periodic fixed-duration reconfigurations, similar to RotorNet. We formally establish throughput bounds for Vermilion, demonstrating that it achieves at least $33%$ more throughput in the worst-case compared to existing designs. The key innovation of Vermilion is its short traffic-aware periodic schedule, derived using a matrix rounding technique. This schedule is then combined with a traffic-oblivious periodic schedule to efficiently manage any residual traffic. Our evaluation results support our theoretical findings, revealing significant performance gains for datacenter workloads.