Scaling Optical Interconnects for Hyperscale Data Center Networks

Elastic optical network (EON) is a promising candidate for the next generation optical transport network because of its high spectrum efficiency and flexible spectrum allocation. However, to migrate from today’s dense wavelength division multiplexing (DWDM) network to EON, optical nodes such as reconfigurable optical add/drop multiplexers (ROADMs) should be replaced with the ones with expensive flexi-grid wavelength selective switches (WSSs). Thus, whether and how today’s optical network will be evolved to EON is still under debate. Meanwhile, with the technical maturity of optical super-channels that transmit bit rates at the level of 1 Tb/s, super-channels are expected to dominate the future optical transport network because of their attractiveness of high spectrum efficiency and superior cost effectiveness. In this paper, we propose a new quasi-coarse wavelength division multiplexing (Quasi-CWDM) optical network architecture that is compatible to today’s WDM network in spectrum operation, while still maintaining the flexibility of adaptation between the bit rate and the transparent reach of an optical super-channel, but not requiring expensive flexi-grid WSSs. Specifically, the Quasi-CWDM optical network employs large fixed frequency grids, whose spacing falls between the traditional CWDM and DWDM, e.g., 200 GHz or 400 GHz. To maintain the flexibility between the bit rate and the transparent reach of an optical super-channel, different modulation formats are chosen according to channel physical distances. To minimize network hardware cost and maximize carried user traffic demand, we develop a mixed integer linear programming (MILP) model and a heuristic algorithm that can efficiently groom upper-layer IP traffic onto the optical layer by optical super-channels. We compare the design based on the Quasi-CWDM network to those based on the DWDM network and the advanced EON. It is found that the proposed Quasi-CWDM network has a much lower hardware cost than the other two types of networks, while achieving spectrum efficiency close to EON.

To accommodate big data, new technologies and features have started to be used in core network. Consequently, optical transport network has evolved towards Elastic Optical Network (EON). Despite the effort made towards efficient and reliable network management, interoperability problem in multi-vendor setting of the EON remains unsolved and it hinders the network management and monitoring. Hence, Network operators need to rely on the proprietary management solutions. This phenomenon poses complexity in network configuration, control, monitoring and management and also in turn results in higher Capital Expenditure (CAPEX) and Operational Expenditure (OPEX) for the service providers. In this paper, standard data definition model of network elements of EON is presented for configuration, management and monitoring purpose. Standard YANG data models for the network elements, namely, Reconfigurable Optical Add/drop Multiplexer (ROADM), Bandwidth Variable Transponder (BVT) and Optical amplifier (OA) are presented. The implementation scenario for the developed data definition models are demonstrated for managing state of the art EON. It is contended that the adoption of these standard YANG data models within the NETCONF protocol not only facilitates the network configuration and management, but also improves the multi-vendor interoperability in the control and management plane of the EON.

This book presents an in-depth deliberation on optical networks in four parts, capturing the past, present, and ensuing developments in the field. Part I has two chapters presenting an overview of optical networks and the enabling technologies. Part II has three chapters dealing with the single-wavelength optical networks: optical LANs/MANs, optical access networks using passive optical network architecture, SONET/SDH, optical transport network and resilient packet ring. Part III consists of four chapters on WDM-based optical networks, including WDM-based local/metropolitan networks (LANs/MANs) using single and multihop architectures over passive-star couplers, WDM/TWDM access networks as an extension of PONs with WDM transmission, WDM metro ring networks covering circuit-switched (using point-to-point WDM and wavelength-routed transmission) plus packet-switched architectures and WDM long-haul backbone networks presenting the offline and online design methodologies using wavelength-routed transmission. Part IV deals with some selected topics in six chapters. The first deals with transmission impairments and power-consumption issues in optical networks, while the next three chapters deal with the survivable optical networks, network control and management techniques, including GMPLS, ASON, and SDN/SDON, and datacenter networks using electrical, optical, and hybrid switching techniques. The final two chapters present elastic optical networks using flexible grid for better utilization of the optical-fiber spectrum and optical packet and burst-switched networks. The three appendices present the basics of the linear programming techniques, noise processes encountered in the optical communication systems, and the fundamentals of queuing theory and its applications in telecommunication networks. (238 words)

The cost of an optical network in wavelength division multiplexing (WDM) networks can be reduced using optical reconfigurable optical add/drop multiplexers (ROADMs), which allow traffic to pass through without the need for an expensive optical-electro-optical (O-E-O) conversion. Waveband switching (WBS) is another technique to reduce the network cost by grouping consecutive wavelengths and switching them together using a single port per waveband. WBS has attracted the attention of researchers for its efficiency in reducing switching complexity and therefore cost in WDM optical networks. In this paper, we consider the problem of switching wavelengths as non-overlapping uniform wavebands, per link, for mesh networks using the minimum number of wavebands. Given a fixed band size b s , we give integer linear programming formulations and present a heuristic solution to minimize the number of ROADMs (number of wavebands) in mesh networks that support a given traffic pattern. We show that the number of ROADMs (or number of ports in band-switching cross-connects) can be reduced significantly in mesh networks with WBS compared to wavelength switching using either the ILP or the heuristic algorithm. We also examine the performance of our band assignment algorithms under dynamic traffic.

Growing concerns regarding the scalability of current optical networks as well as IP-based networks are driving two important trends. One trend is a shift from the current rigid optical networks to spectrally efficient elastic optical networks with a flexible bandwidth and adaptive channel spacing. The other trend is IP traffic offloading to a lower layer yielding benefits that are potentially cost-effective and power-efficient. This article presents a novel multiflow optical transponder (OTP) that enables more efficient IP optical networking. A multiflow OTP allows client data flows that arrive from a single client interface to be mapped to multiple optical flows. In cooperation with the emerging spectrally efficient elastic optical path networking technology, multiflow OTPs can provide multiple optical connections from a single OTP to multiple OTPs. IP traffic offloading to an elastic optical path layer architecture and the effect of introducing multiflow OTPs (i.e., the potential reduction in the number of router interfaces and considerable potential for cost savings) are discussed. These benefits are brought about by increasing the number of directly connected router pairs while keeping router-to-optical-node interconnections simple. A novel optical virtual private line service based on multiflow OTPs that supports multiple optical connections from a single customer site to multiple customer sites with capacity adjustment is also discussed.

DWDM optical transport network is the hotspot of the optical communications, the protection and restoration of the multiple wavelength optical transport network is the key technique in the DWDM optical transport network. In this paper, at first we introduce the concept of separate optical layer- the concept of separate optical layer has been demonstrated in some important experimental optical networks- and some advantages of the multiple wavelength optical transport networks, we also compare the DWDM optical transport networks with the SDH networks, put forward some analogies of them. The second, we analyzes the protection and restoration of the multiple wavelength optical networks, we analyze multiple fiber ring networks in detail, we analyze the span protection, line protection and path protection, discuss the advantage and the shortcoming of these different protection mode.

Network automation is the current challenge that operators face to accelerate end-to-end service delivery and improve network operations in the context of rapidly growing capacity needs. Optical networks have long been built using vendors’ turn-key solutions, resulting in complexity for the operator to automate its multi-vendor networks in a consistent way. Questioning the vendors’ lock-in in disaggregating the optical transmission systems (i.e., splitting the system into parts, where each part can be provided by a different vendor) provides advantages (e.g., cost) but also additional complexity. Openness and standardization appear as key for these multi-vendor scenarios. The path is not easy, but it is not a matter of choice: opening and interoperability are obligations that we face because of the need for automation. The journey that we detail in this paper is the one that we think will be sustainable from our side. The opposite, i.e., continue building vendors’ silos and staking per-vendor specialized applications, is not sustainable. We propose a step-by-step approach, starting with the non-disaggregated situation, followed by partial and full disaggregation architectures to the last evolution towards data centric networking. We present and discuss implementations that Orange has been contributing to and identify some gaps the industry should address. We show that current works in the communities of open source, open initiatives, and standardization bodies are addressing all these steps, and in this respect can accelerate the deployment of automation solutions in current and future optical transport networks.

This paper presents a policy-based management framework for GMPLS-enabled optical networks. First, the ASON (Automatic Switched Optical Networks) and GMPLS (Generalized Multi-Protocol Label Switching) architecture are presented. We will explain how the policy plane relates to the control and management plane, and the role it plays in the dynamic configuration of GMPLS optical paths. We will present a framwork based on an admission control protocol COPS) Common Open Policy Service) defined by the IETF (Internet Engineering Task Force) and on an Information Model based on the CIM (Common Information Model) model defined by DMTF (Distributed Management Task Force). This framework allows interworking between the policy and control plane, and provides a comprehensive policy provisioning setup for optical networks. An information model extended from CIM for managing dynamically configured optical networks is presented.

OXC (Optical Cross-Connect) system is on indispensable network element to realize the all-optical mesh networks, because that supports the increase of link channels and the expansion of an OTN (optical transport network) capacity at low expense. This paper describes architecture of APRS (Automated Connection Provisioning and Restoration Subsystem), which can be applicable various network topologies. Since APRS manages network connections and optical circuits fOr transparent OXC that can be classified by survivability class level, it provides an efficient mechanism for optical connection management. APRS coordinates with SMS (Service Management System) or network services, etc, VPN (Virtual Private Network), VLAN (Virtual Local Area Network), and QoS (Quality of Service), so it can support optical level CoS (Class of Service). And APRS supports an automated connection provisioning that can provide PC (Permanent Connection), SPC (Soft Permanent Connection), and SC (Switched Connection) and supports a dynamic restoration mechanism according to the policy and survivability class level. APRS consists of distributed objects, and then it can be applied in not only the layered-centralized network connection management without control plane but also the distributed network connection management with control plane.

With the fast development of various applications, the demand for network bandwidth has been increasing significantly during the past decade. Although the WDM (wavelength-division multiplexing) technology has already provided high bandwidth using parallel wavelength channels, the overall network spectrum efficiency is however severely discounted by the fixed grid definition and standard rate transmission. The short-term solution to increase capacity is by introducing hardware enabling elastic optical networking (EON), which is intrinsically related to flexible-grid channel allocation; and the longer-term solution involves extending the fiber capacity by space division multiplexing (SDM) based on new types of fibers supporting multiple cores, multiple elements, few LP modes or even modes carrying Orbital Angular Momentum (OAM). Moreover, for short-term evolution, it may not be economically viable to make a one-time complete upgrade to full flexible-grid technology on the entire network, which means, before the current optical transport network capacity is exhausted, the current fixed-grid network could be maximally kept operational during the migration to preserve the already-made investment. In this talk, a series of topics will be discussed, such as the network architecture with co-existing fixed-grid and flexible-grid technologies, migration strategies evolving from fixed-grid to flexible-grid, SDM based Reconfigurable Optical Add/Drop Multiplexers (ROADMs), as well as networking aspects of routing and spectrum resource allocation when considering spectral/spatial switching.

Radio access network (RAN) cost savings are expected in future cloud RAN (C-RAN). In contrast to traditional distributed RAN architectures, in C-RAN, remote radio heads (RRHs) from different sites can share baseband processing resources from virtualized baseband unit pools placed in a few central locations (COs). Due to the stringent requirements of the several interfaces needed in C-RAN, optical networks have been proposed to support C-RAN. One of the key elements that needs to be considered are optical transponders. Specifically, sliceable bandwidthvariable transponders (SBVTs) have recently shown many advantages for core optical transport networks. In this paper, we study the connectivity requirements of C-RAN applications and conclude that dynamicity, fine granularity, and elasticity are needed. However, there is no SBVT implementation that supports those requirements, and thus, we propose and assess an SBVTarchitecture based on dynamic optical arbitrary generation/measurement. We consider different long-term evolution-advanced configurations and study the impact of the centralization level in terms of the capital expense and operating expense. An optimization problem is modeled to decide which COs should be equipped and which equipment, including transponders, needs to be installed. The results show noticeable cost savings from installing the proposed SBVTs compared to installing fixed transponders. Finally, compared to the maximum centralization level, remarkable cost savings are shown when a lower level of centralization is considered.

This paper describes the broadband situation in Japan and the recent development of optical metro networks, their future prospects and technologies. First, the situation as regards the broadband infrastructure and internet traffic in Japan is overviewed. Then, the recent development of reconfigurable optical add drop multiplexers (ROADMs) in optical metro networks is reviewed. Finally, the future prospects for optical metro networks are described with particular emphasis on 100Gbit/s standardization, sub-λ function, and colorless/directionless ROADMs.

The chapter discusses the security of the client signals over the optical network from any wiretapping or loosing. The physical layer of the optical transport network (OTN) is the weakest layer in the network; anyone can access the optical cables from any location and states his attack. A security layer is proposed to be added in the mapping of OTN frames. The detection of any intrusion is done by monitoring the variations in the optical signal to noise ratio (OSNR) by using intelligent software defined network. The signal cryptographic is done at the source and the destination only. The chapter shows how the multi-failure restorations in the multi-domains could be done. A new model is introduced by slicing the multi-domains to three layers to fit the needs of 5G. The results show that the multi-failure restoration improved from 25% to 100%, the revenue from some OTN domains increased by 50%, the switching time enhanced by 50%, the latency reduced from 27 msec to 742 usec, and it will take many years to figure out the right keys to perform the decryption process.

As power grids and optical interconnection networks are interdependent, the reliabilities of the optical networks are critical issues in power systems. The optical networks hold prominent performance including wide bandwidth, low loss, strong anti‐interference capability, high fidelity, and reliable performance. They are regarded as promising alternatives to electrical networks for parallel processing. This paper is aimed at taking the first step in understanding the communication efficiencies of optical networks. For that purpose, on optical networks, we propose a series of novel notions including communication pattern, r‐communication graph, reduced diameter, enhanced connectivity, r‐diameter, and r‐connectivity. Using these notions, we determine that the r‐diameter and r‐connectivity of the optical n‐dimensional hypercube network are ⌈n/r⌉ and , respectively. Since the parameter r is variable, we can adjust different values of r on the basis of the wavelength resources and load of the optical networks, achieving enhanced communication efficiencies of these networks. Compared with the electric n‐dimensional hypercube network, the proposed communication pattern on the optical hypercube network not only reduces the maximum communication delay of the conventional electrical hypercube significantly but also improves its fault tolerance remarkably.

The paper gives an overview of the automatically switched optical network (ASON), concentrating on its control plane. Drivers for automatically switched optical networks are briefly presented. The ASON architecture, based on transport, control, and management planes, is discussed. Types of connections supported by ASON are described, including permanent, switched, and soft permanent connections. Different kinds of internal and external interfaces are presented. Various issues related to the ASON control plane are enumerated.