Routing, spectrum and core assignment for space division multiplexing elastic optical networks

Abstract

Elastic Optical Network (EON) is expected as one of future networks in terms of spectrum flexibility. While Routing and Wavelength Assignment (RWA) problem is one of the key issues in traditional Wavelength Division Multiplexing (WDM) networks, Routing and Spectrum Assignment (RSA) problem has much impact on network performance in EONs. On the other hand, Data-Center (DC) traffic and/or mobile back-haul traffic keep on increasing. To deal with such forthcoming huge capacity of applications, Space Division Multiplexing (SDM) technologies, such as Multi-Core Fiber (MCF) and few-mode fiber, are intensively researched. From network perspective, this paper focuses on Routing, Spectrum and Core Assignment (RSCA) problem for future SDM-EON.Introducing multi-core fibers makes RSA problem more complex because fiber-core dimension is newly expanded. In addition, physical impairment caused by MCF must be taken into account. In this paper, first, the target RSCA problem is divided into the routing and SCA problems, and Kshortest path based pre-computation method is introduced as the routing solution. Next, according to whether MCF has intercore crosstalk or not, we propose SCA methods with crosstalk awareness and with prioritized area concept, respectively. Finally, the paper evaluates the effectiveness of the proposed algorithms compared with representative ones.

Survey and Evaluation of Space Division Multiplexing: From Technologies to Optical Networks

“…Connection AE presents elastic wavelength/band and sub-lamda granularity, whereas CG shows fiber/core granularity instead [128]. [136] RSSMA problem has been tackled by researchers by proposing several algorithms and by simulations [119], [132]- [136] usually regarding MCFs. These methods, as seen in the flowchart of Fig.…”

Resource Allocation for Space-Division Multiplexing: Optical White Box Versus Optical Black Box Networking

“…For the SDM network planning and operation, there exist limited work on resource allocation strategies that use the additional flexibility offered by the spatial domain [15], [16], [18], [19], [38], [39]. For example, the optimal planning for MCF based SDM networks with the objective to minimize the required spectral resources is investigated in [15].…”

“…The study in [16] extends the planning problem to networks implemented through programmable AoDs with the aim to reduce the network switching resources. References [18], [19] solve the resource allocation problem for dynamic SDM networks with the objective to minimize the network blocking probability. Similarly, the routing, wavelength and mode assignment problem for dynamic networks with FMF is examined in [38].…”

“…A more flexible and scalable architecture for multi-dimensional switching node can be accomplished using optical white box (AoD) [14]. In the context of network planning and operation, the advent of spatial multiplexing introduces an additional degree of freedom and new issue, such as inter-core crosstalk, which is incorporated in the proposed resource allocation strategies [15]- [19]. These schemes target specific objectives for static [15], [16] and dynamic [17]- [19] traffic, without analyzing the impact of network parameters (e.g., traffic growth, adaptive modulation, optical node solutions) on the network capital and operational expenditures.…”

“…In the context of network planning and operation, the advent of spatial multiplexing introduces an additional degree of freedom and new issue, such as inter-core crosstalk, which is incorporated in the proposed resource allocation strategies [15]- [19]. These schemes target specific objectives for static [15], [16] and dynamic [17]- [19] traffic, without analyzing the impact of network parameters (e.g., traffic growth, adaptive modulation, optical node solutions) on the network capital and operational expenditures. This paper investigates the planning and dimensioning of MCF-based SDM networks considering the quasi-static traffic model (which is the norm on current and future operational metro/core networks).…”