With its agile spectrum management in the optical layer, the flexible-grid elastic optical network can become a promising physical infrastructure to efficiently support the highly dynamic traffic in future datacenter interconnections (DCIs). While the resulting elastic optical DCIs (EO-DCIs) need to serve requests that not only require bandwidth resources on fiber links but also require multidimensional IT resources in the DCs, multidimensional resource fragmentation can occur during dynamic network operations and deteriorate the network performance. To address this issue, this paper investigates the problem of joint defragmentation (DF) for the spectrum and IT resources in EO-DCIs. Specifically, we reoptimize the allocations of multidimensional resources jointly with complexitycontrolled network reconfigurations. For the DF operation, we first study the request selection process and propose a joint selection strategy that can perform the spectrum-and IT-oriented selections adaptively according to the network status. Then, we formulate a mixed integer linear programming model and design several heuristics to tackle the problem of network reconfiguration in the joint DF. The proposed algorithms are evaluated with extensive simulations. Simulation results demonstrate that the proposed joint DF algorithms can significantly reduce the blocking probability in EO-DCIs by consolidating the spectrum and IT resource usages effectively.Index Terms-Elastic optical datacenter interconnections (EO-DCIs); Multi-dimensional resource defragmentation.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.