We compare IP core networks in three scenarios—those using point-to-point optical links between adjacent nodes with ZR/ZR + based pluggables (called hop-by-hop routing), leveraging reconfigurable optical add–drop multiplexers (ROADMs) with ZR + pluggables, and ROADMs using performance-optimized flex-coherent muxponders—from hardware count, wavelength utilization, and latency perspectives. Specifically, we compare the number of components/router ports used to build an IP-over-WDM core network and attempt to understand if off-the-shelf pluggables are indeed suitable candidates for replacing custom-designed coherent muxponders (which we conveniently term as flexponders) in core networks. To this end, we build a constrained optimization model whose goal is to minimize the number of components used by optimally placed regenerators—the single biggest variable metric in core IP-over-WDM network design that eventually impacts total hardware count. Results show a 75% decrease in transceiver count for flexponders developed from coherent transponders/muxponders when compared to ZR + pluggables in routers with hop-by-hop routing and a 51% decrease in flexponders when compared to using ZR + pluggables in routers across ROADMs in a U.S. core network and a pan-India core network with 5 and 3 years of increasing demands, respectively.

This paper compares the incremental cost of CDC features in super-channel based and wavelength based ROADM architectures. A real world network model analysis shows fewer network components and less incremental cost with superchannel CDC ROADM.

No abstract

No abstract

No abstract