Time-shifted multilayer graph: A routing framework for bulk data transfer in optical circuit-switched networks with assistive storage

Increasing bulk data transfers incur peak-hour bandwidth contention between bulky and short flows as well as between bulky flows. To mitigate this contention, storage is introduced into the forwarding path so that bulk data that are delay tolerant can be temporarily stored and forwarded at a later time. However, the storage introduces an additional complexity into the conventional routing process. What was a spatial resource allocation problem becomes a scheduling problem, in which both bandwidth and storage constraints must be considered and both spatial assignments and temporal arrangements must be performed. In this paper, we propose a routing framework for bulk data transfer in optical circuit-switched networks with assistive storage. This framework is based on a multilayer graph built from a set of snapshots (i.e., layers) of the dynamics in a network. By performing shortest path routing on the multilayer graph, "end-to-end" paths over time and space are found for requests, thus greatly simplifying the provisioning process. We study how the number of layers used for routing affects the network blocking performance and how traffic characteristics and link capacity affect the number of layers. We find that the majority of requests can be served with only one additional layer. A trade-off between computational complexity and blocking performance can be reached by limiting the number of layers used for routing. In our simulations, the request blocking probability is reduced from 13.5% to 0.9% by limiting routing to 11 layers. Request blocking is avoided when routing is allowed within 29 layers.