Beschreibung:
<jats:p>The sheer traffic increase stimulated by new 5G-and-beyond services requires continuous innovation in designing optical networks that scale in terms of capacity while minimizing equipment cost and ensuring lightpath quality-of-transmission (QoT). Multiband transmission is among the most prominent solutions for addressing capacity needs and consists of extending the transmission spectrum, e.g., to the (C+L)-band. Multiband transmission poses new challenges, mainly due to physical layer impairments such as interchannel stimulated Raman scattering, which causes power transfer among transmission bands with a consequent degradation in lightpaths’ QoT. To mitigate such degradation, we investigate how to introduce hybrid erbium-doped fiber amplifier (EDFA)/Raman amplification (HFA) by selectively upgrading an initial EDFA placement in the network. We propose a genetic algorithm approach to optimize the HFA upgrade process and compare its performance to baseline upgrade strategies. Based on when the HFA upgrade is performed, we consider two HFA upgrade strategies: (i) bulk HFA upgrade (bulk-HFA) and (ii) sequential HFA upgrade (seq-HFA). Bulk-HFA assumes a one-shot optimization that upgrades HFA only once, when all links are operating in the (C+L)-band, whereas seq-HFA assumes a sequential optimization considering an HFA upgrade in multiple steps, triggered by incremental traffic. Numerical experiments run through a national and continental network topology show that an optimized HFA upgrade allows us to avoid lightpath degradation while deploying up to 64% fewer amplifiers compared with baseline strategies that upgrade EDFAs to HFA amplifiers in all eligible spans, i.e., spans longer than 70 km. Moreover, we show that seq-HFA planning achieves higher spectral efficiency compared with bulk-HFA while upgrading a comparable number of EDFAs to HFA.</jats:p>