Identified Traffic Partition Grooming in Multi-Granularity Optical Network

Abstract

Currently, the required bandwidth of users tends to be diversity, i.e., coarse-granularity demands (e.g., terabit/circuit routing) and small-granularity demands (e.g., sub-rate traffic streams). Only employing traffic grooming or waveband switching may result in several traffic-diversity problems. Therefore, the traffic partition grooming technique with considering traffic transferring is proposed in this paper to solve the traffic-diversity problems, especially for achieving blocking probability reduction and port savings. Correspondingly, we devise the heuristic and Identified MG-IGAG (IMG-IGAG) algorithm based on Granularity-Layered Graph (GLG) since the problem of grooming is the NP-hard. The simulation results show that our IMG-IGAG is more effective and practical than other state-of-the-art methods.