Optimizing Light-Tree and Edge Resource Placement for Diverse Latency Needs

Coherent Point-to-MultiPoint (P2MP) transceivers are gaining attention for their efficacy in aggregating diverse workloads through optical light-trees, particularly in extended access scenarios where P2MP hub nodes integrate edge computing capabilities. This work addresses the challenge of optimizing the placement of both computing and networking resources to serve various workload quality classes, which can be suboptimal when treated independently. This is done taking into account the latency and network capacity requirements of the different classes, the optical connections physical layer performance, as well as computational costs.