Abstract

In this paper, we identify challenges in developing future optical network infrastructure for new services based on technologies such as 5G, virtual reality, and artificial intelligence, and we suggest approaches to handling these challenges that include a business model, architecture, and diversity. Through activities in multiservice agreement and de facto standard organizations, we have shown how the hardware abstraction layer interfaces of optical transceivers are implemented for multivendor and heterogeneous environments, coherent digital signal processor interoperability, and optical transport whiteboxes. We have driven the effort to define the transponder abstraction interface with partners. The feasibility of such implementation was verified through demonstrations and trials. In addition, we are constructing an open-transport platform by combining existing open-source software and implementing software components that automate and enhance operations. An open architecture maintains a healthy ecosystem for industry and allows for a flexible, operator-driven network.

© 2020 Optical Society of America under the terms of the OSA Open Access Publishing Agreement

1. INTRODUCTION

In recent years, a number of revolutionary technologies have emerged in the optical network and data center (DC) domains to support the explosive growth in traffic generated by services such as social networking and video streaming. The transmission capacity per fiber of optical networks has increased to over 1 Tbps with the introduction of innovative devices such as the erbium-doped optical fiber amplifier (EDFA) and arrayed waveguide grating in the 2000s [1] (see Fig. 1). In the decade that followed, the capacity has expanded to 8 Tbps with the development of digital coherent technology to compensate for waveform distortion caused by physical phenomena such as chromatic dispersion and polarization dispersion. This has enabled beyond 100G transmission with polarization multiplexing and multilevel modulation [24]. The evolution of digital signal processing (DSP) large-scale integrations (LSIs) has also led to significant space and power savings in equipment. Communication service providers (SPs) have used these innovations to build high-capacity, reliable networks based on multistage ring topologies to efficiently accommodate increasing traffic while providing quality services to users.

DC technologies that decouple software from hardware, such as server virtualization [5] and storage virtualization (the process of grouping the physical storage from multiple network storage devices to improve the capacity and fault tolerance), have been implemented since the 2000s (see Fig. 2). In addition to reducing costs by increasing efficiency, these technologies have also enabled rapid system integration and reliable and flexible system expansion. Around 2007, researchers, mainly in the field of computing, proposed and began implementing new types of routers and switches in which the hardware and software are disaggregated for highly flexible networks [6]. The Open Networking Foundation (ONF) was launched in 2011 to develop OpenFlow [7], and the Open Compute Project (OCP) was launched to improve the efficiency, flexibility, and scalability of DC equipment [8]. These activities led to the creation of a business model for DC equipment. The model was based on horizontal segmentation, which is not limited to a specific vendor’s products. The Linux-based platform has facilitated the evolution of open-source software (OSS); operational automation tools such as zero touch provisioning and streaming telemetry are now being used in DCs. As a result, DC operators have been able to reduce their operational costs.

 figure: Fig. 1.

Fig. 1. Trends in optical transport technologies.

Download Full Size | PPT Slide | PDF

 figure: Fig. 2.

Fig. 2. Virtualization technology and OSS.

Download Full Size | PPT Slide | PDF

The development and spread of 5G and virtual-reality-related technologies are expected to lead to the emergence of entirely new services that were not previously possible. Real-time communication between distant devices, as used in automated driving and telemedicine, make realistic, high-quality experiences available to humans [9,10]. To develop these new services, it is necessary to provide services with hyperscale DCs located far from users as well as edge DCs close to users (e.g., within the same city), so the carrier’s central office is also expected to serve as an edge DC [11]. In addition to traffic between large DCs, traffic between edge DCs and hyperscale DCs is expected to grow in the future. In the past, technology for optical networks and DCs were developed independently. However, in the future, we will need a network architecture that combines optical network hardware with DC software to continually expand capacity while meeting new service requirements. In addition, operational automation is needed to efficiently expropriate ever-expanding traffic [12] and provide services in a timely manner. This work extends [13], which presented a high-level view of the open whitebox architecture. This article includes a detailed explanation of their challenges and approaches and includes new terminology for the standardization of vendor-to-vendor interconnection.

2. CHALLENGES IN ARCHITECTURE DEVELOPMENT

Challenges arising from business models, diversification, and operations need to be addressed in order to develop a new network architecture.

A. Business Models

As described in the previous section, the business model of horizontal segmentation has already been established for DC equipment. However, optical transport equipment was developed based on a vertically integrated business model, as its main application was carrier services, where extremely high reliability is required due to the difficulty of ensuring reliability through redundancy. As a result, the management software that controls optical transport equipment is divided into a chassis and transceiver according to each company’s own specifications.

  • 1) Since hardware and software are tightly coupled, operators cannot choose their preferred ones.
  • 2) Each company’s equipment requires the development of an operating system and training for operators.
  • 3) There is a risk of market dominance and vendor lock-in by a few vendors.

B. Diversification

Recent innovations in optical components, such as coherent DSPs and silicon photonics, have shortened the technology development cycle. As a result, the next generation of devices are quickly being brought to market. In addition, there has been a significant increase in the types and generations of form factors on the market [quad small form-factor pluggable (QSFP), octal SFP (OSFP), C form-factor pluggable (CFP)/CFP2/CFP4/CFP8, intensity-modulation direct-detection (IM-DD)/coherent, analog coherent optics (ACO)/digital coherent optics (DCO), etc.]. This is primarily due to an increase in the number of operator types building and operating optical networks (communication SPs, Internet content SPs, DC SPs, etc.) with different transceiver requirements. Therefore, it is important to efficiently accommodate a wide variety of form factors in terms of type and generation to reduce procurement and operation costs.

C. Operation

As the number of devices and components that need to be supported increases as described above, network operations become increasingly complex. In addition, since the design optimization and construction of optical transport networks are carried out by a limited number of specialized engineers, it is difficult to reduce service provision costs and lead times. Replacing equipment remains difficult even when next-generation technologies with high performance and power savings are introduced.

3. APPROACH FOR SMART INTEGRATION

A new network architecture that enables efficient deployment of network, computing and storage resources and automatic operation control requires the following:

  • • defining an open interface to eliminate hardware complexity;
  • • defining vendor-to-vendor interconnection modes between coherent DSPs;
  • • implementing a Linux-based platform that can accelerate the use of OSS;
  • • defining application programmable interfaces (APIs), procedures, and monitoring techniques for automatic operation.

A. Defining an Open Interface to Eliminate Hardware Complexity

SPs have been paying attention to the hardware and software disaggregation of network equipment. This approach has many advantages, such as avoiding vendor lock-in and applying a “pay-as-you-grow” model to provide right-sized equipment to meet customer demand. In optical networks, there are some disaggregated network architecture alternatives: aggregated, partially disaggregated, and fully disaggregated [14]. The partially disaggregated model decouples transponders from open line systems (OLSs), and the fully disaggregated model disaggregates each building block in the OLS. As the level of disaggregation increases, a greater degree of controllability and manageability is required. To realize that, the whitebox approach and defining the open interface will play an important role in exposing the configuration parameters of network elements and controlling them universally. In this paper, we assume the partially disaggregated model with a whitebox as the transponder.

The switch abstraction interface (SAI), a common interface for switch application-specific integrated circuits (ASICs) headed by Microsoft, was formalized into an OCP project in 2015 [15]. It was supported by eight switch ASIC vendors, including Broadcom and Mellanox, accelerating the openness of the network switch. SAI enables the switch manager, a software component that controls the switch ASIC, to be developed independently of the switch ASIC. Since SAI was developed as an OSS, the switch manager can also be developed as an OSS [16]. In fact, Microsoft is leading the development of SONiC, one of the most popular network operation systems (NOSs), as an OSS, with many operators and vendors adding and improving features through open collaboration [17]. On the other hand, because the conventional transponder does not have an open interface like SAI, the software that controls the optical transport components tends to be hardware-dependent. This makes it difficult to decouple the software from the hardware and to develop the software as an OSS in a community-driven manner. The fastest method would be to add transponder functionality on top of the already market-established whitebox switch architecture; however, this would require whiteboxes to be able to implement digital coherent technology developed for long-distance communications beyond 100 Gbps. Digital coherent technology has various advantages, including high receiver sensitivity due to coherent detection, compensation of waveform distortion caused by chromatic dispersion and polarization mode dispersion circuits, and high frequency utilization efficiency due to multilevel coding. However, the complexity and diversity of implementation and configuration have been obstacles in the establishment of a horizontal segmentation-based business model.

Figure 3 shows an example of digital coherent technology implemented in a whitebox switch architecture. Implementing the software development kit (SDK) in user space to hide complex register access to coherent physical layer (PHY) (DSP) and Ethernet PHY chips. This enables the operator to configure coherent modules in a manner similar to how they configure Ethernet modules on the whitebox switch NOS [18]. The example in Fig. 4 is the simplest example of 100 Gbps dual-polarization quadrature phase shift keying (DP-QPSK) communication with a single DSP die using CFP2-ACO as the form factor. However, implementation of digital coherent technology takes many forms due to system characteristics such as multilevel modulation, as well as restrictions arising from implementation, such as power consumption limits, footprint limitations, and the I/O limitation of LSI.

 figure: Fig. 3.

Fig. 3. Whitebox packet transponder architecture.

Download Full Size | PPT Slide | PDF

 figure: Fig. 4.

Fig. 4. Examples of hardware complexity for coherent communications.

Download Full Size | PPT Slide | PDF

The left side of Fig. 4 shows an example of two CFP4-ACOs used as the form factor and a DSP chip in the dual die configuration for communicating at two wavelengths in DP 8 quadrature amplitude modulation (8QAM) format at 150 Gbps. Here, the complexity of the hardware for coherent communication is not limited to differences among vendors but also includes the number of dies implemented in the DSP package, types of control interfaces (IFs), tributary and multiflow control of optical paths [19] associated with polarization multiplexing and multilevel modulation, DSP implementation methods, and module implementation methods.

In 2016, the Open Optical & Packet Transport (OOPT) project [20] was launched within the Telecom Infra Project (TIP), which aims to disaggregate network equipment used in carrier networks. The first specification and architecture of the transponder abstraction interface (TAI) for hardware and software disaggregation of transponders was proposed in December 2017 [21].

Figure 5 shows the architecture of TAI. TAI is defined as a C header file, following SAI, and its implementation (the C shared library) is usually provided by each hardware vendor. The C shared library Libtai.so contains the vendor-supplied SDK and is usually provided by the respective hardware vendor [2224].

 figure: Fig. 5.

Fig. 5. TAI architecture.

Download Full Size | PPT Slide | PDF

 figure: Fig. 6.

Fig. 6. Whitebox packet transponders Cassini and Galileo.

Download Full Size | PPT Slide | PDF

By defining TAI and its architecture, it is possible to eliminate the complexity of the hardware for NOS vendors, as shown in Fig. 6. Operators will have more freedom of choice, such as the ability to select individual hardware and software to take advantage of the open-source NOS. Hardware vendors can remove development redundancies themselves by defining a common set of hardware abstraction interfaces. They can also implement their own differentiating technologies as options in TAI without disclosing them to their competitors. In addition to telecom and cloud operators, NOS vendors, original design manufacturing vendors, transceiver vendors, and component vendors are involved in defactoring the TAI architecture and optimizing TAI through cross-industry discussions. For example, each vendor’s transceiver is compliant with ITU-T G.694.1, but generally the optical frequency range covered by each vendor’s products and the channel 1 frequencies defined by each vendor are different. For this reason, TAI is designed so that each frequency channel can be set by optical frequency value rather than channel number. Additionally, NOS can automatically detect the optical frequency range of each vendor’s module through TAI and display it to the operator (see Section 3.D for details). TAI was implemented in three different whitebox packet transponders and NOSs—Voyager [25], designed by Facebook; Cassini [26], designed by Edgecore Networks; and Galileo, designed by Wistron—open NOSs such as TIP’s Goldstone and ONF’s Stratum, and commercial NOSs such as Cumulus Linux from Cumulus Networks and OcNOS from IPinfusion. TIP collaborated with vendors to implement different NOSs on Cassini and Galileo (Fig. 6). They implemented TAI at the TIP Summit 2019 and conducted an interconnection demonstration (Fig. 7). Various operators have also reported on the feasibility of Cassini; a Japanese cloud operator, Mixi, reported its commercial use at the summit [27], and Telefonica reported the results of a field trial at 100 Gbps over 1200 km using Cassini at OFC 2020 [28].

 figure: Fig. 7.

Fig. 7. Interconnection demonstration at TIP Summit 2019.

Download Full Size | PPT Slide | PDF

TIP OOPT is currently promoting the development of Phoenix, an L0/L1 whitebox transponder. Technical specifications were released in December 2019, and a joint request for information in April 2020, done in cooperation with Vodafone, Telefónica, Telia Company, NTT Communications, Deutsche Telekom, and MTN [19]. Phoenix features 400 Gbps/lambda line IF and hardware/software disaggregation. TIP OOPT aims to define and disseminate open interfaces through cross-industry activities to create a highly flexible network and intends to work towards an open line system type of optical architecture integration/management.

 figure: Fig. 8.

Fig. 8. Multivendor interconnection demonstration at OFC 2019.

Download Full Size | PPT Slide | PDF

Tables Icon

Table 1. List of Interconnection Modes

B. Defining Vendor-to-Vendor Interconnection Modes between Coherent DSPs

Compatibility and freedom of equipment are important indicators for operators to provide services at low cost and low risk. In the previous subsection, we introduced efforts to provide compatibility in the control plane. In this paragraph, we introduce compatibility in the data plane. Communicating transmission methods such as forward error correction (FEC), baud rate, symbol mapping, and frame structure of coherent DSPs among different vendors are necessary for data plane compatibility. In 2013, NTT Electronics and Acacia Communications verified an interoperability mode [100G DP-QPSK generic FEC (GFEC)] [29]. Efforts to interconnect ASICs with other vendors were accelerated, and 100G DP-QPSK Staircase (SC) FEC was accepted in 2018 as the ITU-T standard (G.709.2, G.709.3) through career-driven initiatives. Figure 8 shows the multivendor interconnection demonstration conducted at OFC 2019. We implemented TAI in Cassini and connected four coherent transceivers (Acacia Communications CFP2-DCO, Fujitsu Optical Components CFP2-ACO, and Lumentum CFP2-ACO and CFP2-DCO) over 80 km of single-mode fiber, and error-free operation was verified with 100G DP-QPSK SC-FEC.

Progress is also being made in standardization beyond 100 Gbps. 400 Gbps DP-16QAM up to 120 km with concatenated FEC (CFEC) was selected as a mode for OIF 400 G-ZR [30] in response to a request from cloud operators. For 200 Gbps, CableLabs accepted DP-QPSK open FEC (oFEC) as an interoperable PHY specification for a coherent optical access network-to-cable multisystem operator [31]; for 300 Gbps and 400 Gbps, the Open ROADM Multi-Source Agreement (MSA) selected DP-8QAM oFEC and DP-16QAM oFEC as an interoperable mode, respectively [32]. Interconnection among different vendors is now possible in all bands from 100 to 400 Gbps (Table 1).

The OpenZR+ MSA [33] specification is being developed by system vendors, module vendors, and DSP vendors to extend interoperable coherent modules for DC interconnect (DCI) to multiple applications, leveraging the high gain oFEC standardized in the Open ROADM MSA and CableLabs Optical Project. Specifically, OpenZR+ MSA focuses on Ethernet traffic and multivendor interoperability of pluggable modules in order to provide an operationally efficient solution for DCI, regional, and long-haul reaches.

Key features of OpenZR+ include the following:

  • 1) extended transmission distance using high-gain oFEC;
  • 2) dedicated to Ethernet clients, with smaller overhead (OH) line framing;
  • 3) support for 100 Gbps, 200 Gbps, 300 Gbps, and 400 Gbps multirate and muxponder functions such as ${{4}} \times {{100}}\;{\rm{Gbps}}$.

Supported applications of OpenZR+ include the following:

  • 1) a multiapplication small form factor module covering 400 Gbps DCI and 400 Gbps metro and long-haul applications at lower rates (100–300 Gbps),
  • 2) a simple muxponder system using multiple functions of pluggable modules.

Various standards bodies and MSAs are expected to continue working together to advance compatibility in the data plane.

C. Implementing a Linux-Based Platform That Can Accelerate the Use of OSS

In the field of cloud DC infrastructure, various OSSs have been developed, and daily enhancements and bug fixes from collaboration between organizations have led to a highly automated infrastructure. Meanwhile, most of the solutions for optical transport networks are proprietary and vendor-specific, and the use of OSS has not been considered until now. Integrating OSS technology in optical transport networks enables automated and complex operations. Furthermore, by combining OSS technology with hardware abstraction layers such as TAI, it is possible to build software that is not dependent on hardware to prevent vendor lock-in. Using this idea as our basis, we are developing a NOS, Goldstone, as an OSS in TIP [34]. Goldstone implemented OSS, which is widely used for most software components, and reduced the number of proprietary implementations to improve compatibility with software used in the DC infrastructure and cloud space. The system also supports the management methods used in conventional network operations, allowing for a smooth transition to a DC infrastructure-type operation method. Goldstone has already been operating in a commercial network with Cassini for over a year without incident, demonstrating that a combination of OSS components can make a commercially ready NOS feasible [27,35].

Figure 9 is a description of Goldstone’s overall architecture and its major software components.

 figure: Fig. 9.

Fig. 9. Goldstone architecture.

Download Full Size | PPT Slide | PDF

We use Open Network Linux as the base Linux distribution. Open Network Linux is designed to be the basis for a full-fledged NOS and supports a flexible build system and a wide range of whitebox platforms. Next, we use k3s as the application management layer [36]. k3s is a lightweight distribution of Kubernetes, the de facto standard for container orchestrations, and is an OSS developed by Rancher. By using k3s, Kubernetes can be easily run on the host board of a transmission device that is not necessarily rich in computer resources. Goldstone uses k3s to deploy almost all applications as containers. Containers enable software developers to make each application loosely coupled without having to consider conflicting dependent libraries between applications. In addition, it is easier to control the privileges of each container for increased security. Two types of applications run on k3s: one is a group of components that controls the hardware, and the other is a group of management components for higher-level controllers and operators. Each application is described below.

1. Components for Hardware Control

The optical transceiver, gearbox, and peripheral devices such as temperature sensors and fans are all hardware that must be controlled on the transport system. In the case of a packet transponder, the operator needs to control the Ethernet ASIC instead of the gearbox. There are already a variety of OSS components to control them, and Goldstone has integrated them to work as a whole. Goldstone uses TAI for optical transceivers and ONLP [37], a platform API provided by Open Network Linux, as hardware control components for peripheral devices. µSONiC is used in cases where it is necessary to control an Ethernet ASIC. µSONiC is a lightweight package of SONiC, the Microsoft NOS, in which only the components that control SONiC’s Ethernet ASIC are extracted and containerized for easy deployment on Kubernetes. Goldstone adopted the µSONiC package to control Ethernet ASIC because SONiC is the most popular NOS, and many software engineers participate in its development. Note that the Ethernet ASIC’s control components are replaceable as needed, and other components such as DANOS [38] or Stratum [39] can be used instead.

2. Management Components for Higher-Level Controllers

Various northbound APIs have been proposed in recent years, but none of them have become the de facto standard. In response, Goldstone has implemented an architecture in which each component that provides the northbound API is an independent process and does not require changes to existing parts when adding a component that provides a new northbound API. We adopted Sysrepo, a library for configuration and monitoring based on the YANG model, as a core component of the management layer. Applications that use Sysrepo interact with each other via POSIX shared memory. The management layer is divided into two layers: the north layer provides services to users, and the south layer communicates with the hardware control layer. Only the command-line interface is currently implemented in the north layer, but we are planning to develop daemons to provide NETCONF API, gNMI, and Simple Network Management Protocol (SNMP). The south layer contains daemons that communicate independently with TAI/ONLP/SONiC.

 figure: Fig. 10.

Fig. 10. OPEX reduction with the Open Transport Platform.

Download Full Size | PPT Slide | PDF

D. Defining APIs, Procedures and Monitoring Techniques for Automatic Operation

Implementing centrally controlled operation using software will reduce operational expenditures (OPEX) through automation and make it possible to quickly install equipment, change configurations, and perform proactive maintenance (Fig. 10). For this purpose, it is necessary to define open APIs to obtain information useful for optical path design, resource discovery, optical path provisioning, failure isolation, and recovery from optical hardware. Libraries and procedures must also be developed to enable integration with various applications, including OSS. The following is an example of information that can be exchanged through an open API:

  • 1) Coherent DSPs compensate for waveform distortion caused by physical phenomena such as chromatic dispersion (CD) and polarization mode dispersion in an optical fiber. Through this process, it is possible to output information on various physical parameters of the optical fiber, such as CD, polarization-dependent loss (PDL), and differential group delay.
  • 2) When a modular structure is applied to the optical transceiver for multivendor support, it is possible to output resource information for each optical transceiver, such as frequency range, maximum output power, CD compensation range, multilevel modulation mode, and FEC.
  • 3) Transmission quality parameters such as equivalent optical signal-to-noise ratio (e-OSNR) [40] and generalized SNR (GSNR) [41,42] can be determined from the pre-FEC bit error rate (BER) calculated using the FEC function in the coherent DSP and the receiver’s electrical noise value. These quality parameters can be used to optimize multilevel modulation [43,44].
  • 4) Information on the status of the optical transceiver, such as loss of signal (LoS) and end of life (EoL), is output.

We have implemented and validated a method for automatic resource discovery and optical path provisioning by utilizing 2) and 3) on Cassini and TAI/Goldstone (Fig. 11).

  • • After Goldstone is activated, information such as frequency range, maximum optical output, multilevel modulation type, FEC type, and maximum chromatic dispersion tolerance is read from CFP2 and DSP and sent to the controller.
  • • The controller keeps track of each coherent module’s capability (including vendor name, form factor type, and model number) and network-wide resources in the transponder.
  • • Goldstone sets 100G QPSK SC FEC as the default configuration for CFP2 and DSP and then measures the BER after coherent communication is established to inform the controller of the quality of transmission (QoT).
  • • The optical transmission path is designed based on information such as network resources and QoT, and the optimized mode [200G 16QAM soft-decision (SD-FEC) in Fig. 11] is used to re-establish coherent communication as the operational mode.

As shown in the example in Fig. 11, information from modules or optical components can be obtained directly through open API such as TAI to enable configuration and other controls, enabling multivendor coherent modules.

 figure: Fig. 11.

Fig. 11. Resource discovery, optical path provisioning.

Download Full Size | PPT Slide | PDF

As shown in Fig. 12, coherent DSPs that support 600 Gbps transmission (64QAM 65 Gbaud) and a wide range of applications from DCI to metro to ultralong-haul are now commercially available. In addition to the many multivalued and baud rate combinations, the number of modes supported by coherent DSPs is increasing due to the development of coding techniques such as error correction and probabilistic shaping, making manual optical path optimization more difficult. We connected the boards with the coherent DSP shown in Fig. 12 with 80 km of fiber and confirmed that the multilevel modulation could be automatically optimized on the basis of the noise in the optical fiber. Here, we applied a DSP-to-DSP communication channel for control data exchange (QoT notification, setting up DSPs, etc.). Since the required OSNR for the optical path increased as the multilevel increased, it is necessary to consider the effect of the electrical noise from the receiver in the high multilevel region such as 32QAM and 64QAM [40].

 figure: Fig. 12.

Fig. 12. Auto-optical path optimization testing with 600 Gbps DSP.

Download Full Size | PPT Slide | PDF

In addition to the above, we are also investigating a method to monitor the status of the optical transport system in more detail by analyzing the received signals with new techniques such as deep learning as research for future practical use. Figure 13 shows an example the deep-learning technique applied to signal distortion caused by optical fiber bending. By sampling constellation data from the receiver’s analog–digital converter and training it using a convolutional neural network for pattern recognition, we verified that optical fiber bending could be detected accurately [45]. At the initial training phase, a lot of constellation data needs to be pulled from the whitebox packet transponder to the remote controller, but once the training is finished, the fiber bending recognition can be executed locally using the trained model. Note that the constellation data field has not been specified at TAI yet, but the data can be acquired through a vendor-specific field.

 figure: Fig. 13.

Fig. 13. Optical fiber bending detection using an artificial neural network.

Download Full Size | PPT Slide | PDF

This technique has shown potential for detecting signs of failure using only constellation data sampled from the data signal without the need for an external measurement technique such as an optical-time domain reflectometer (OTDR). We also reported on the results of a proactive maintenance framework that combines this technology with optical transport network (OTN) bit lossless protection switching technology based on ${{1}} + {{1}}$ protection to achieve low latency and high-quality network services [46].

In addition, we studied a technique for optimizing the inverse characteristics of the transmission line estimated by digital backpropagation (DBP), a technique to compensate for fiber impairment in optical transmission systems, using neural network (NN)-based DBP. We then estimated the loss and chromatic dispersion of each transmission line span from the estimated results [47]. Figure 14 shows the principle of DBP optimization. The CD and nonlinear compensation are repeated for N steps for the received signal, and the CD coefficients and the nonlinear phase rotation coefficient (NLPR) are optimized to reproduce the original constellation map. We can use a neural network to estimate the loss profile from the NLPR and the dispersion profile from the dispersion coefficient by optimizing the DBP coefficients, which reflect the actual parameters of the optical link.

 figure: Fig. 14.

Fig. 14. Principle of DBP coefficient optimization and loss profile estimation.

Download Full Size | PPT Slide | PDF

Figure 15 shows the results of the NN-based DBP when used to estimate the level diagram of four spans of a 70-km single-mode fiber. The upper figure shows the level diagram; optical attenuators 0, 2, and 5 dB were installed at 50, 90, 190, and 230 km. The fiber input power at each span was adjusted to ${+}{{5}}\;{\rm{dBm}}$. The OTDR power profile was also provided for reference (dotted line). In cases without any loss insertion (solid black line), the NLPR reflected amplification caused by EDFAs and fiber attenuation. The lower figure shows the anomaly indicator, i.e., the difference between the normal (0 dB attenuation) and abnormal conditions. Successful localization of lossy points and detection of relative attenuation level were observed. Applying this technology to maintenance will enable more advanced operations, such as proactive maintenance.

 figure: Fig. 15.

Fig. 15. Loss profile estimation [standard single-mode fiber (SSMF) only, 280 km].

Download Full Size | PPT Slide | PDF

This technology is also expected to be used for automatic optical path design. Conventionally, engineers have designed optical transmission lines using measurement devices, such as OTDR, to measure the physical parameters of each span, and then inputting the measured data into a dedicated simulator. When the accuracy of the method is improved and the system can be used in a commercial network, it will be possible to monitor and estimate optical transmission line information using only the signals received by the coherent DSP and the computer resources on the transponder without external devices. As an example, we show the use of the open-source optical transmission line design tool, GNPy [41,48]. GNPy is an optical transmission line design and optimization tool developed by the Physical Simulation Environment (PSE) members of TIP OOPT. The tool outputs the SNR degradation caused by amplified spontaneous emission (ASE), fiber optical nonlinear effects with the optical transmission line information (fiber loss, amplifier gain, fiber type, etc.), and the optical input information (number of wavelengths, output power, etc.). If the performance parameters, such as the noise figure of EDFA and TxOSNR of the transmitter, are known beforehand, the transmission line can be designed for wavelength division multiplexing transmission without manual measurement.

By applying these monitoring technologies on an open optical hardware interface and the platform that is highly applicable to the DC technology, shown in Section 3.A to Section 3.C, we have shown the possibility of automating operational procedures such as optical path design, resource discovery, optical path provisioning, and failure isolation and recovery.

4. CONCLUSION

A new network architecture that combines optical network hardware technology and software technology in the DC is needed to provide new services that were not previously possible. Such services involve real-time communication between distant devices to provide people with high-quality experiences. The open architecture also maintains a healthy ecosystem for industry and allows for a flexible, operator-driven network. The challenges arising from business models, diversification, and operations need to be addressed in order to develop a new network architecture. We have partnered with organizations to define an open TAI to simplify complex hardware, define vendor-to-vendor interconnection modes between coherent DSPs from 100 to 400 Gbps, and implement a Linux-based platform Goldstone that can accelerate the use of OSS. We will continue to develop monitoring technologies and define APIs and procedures for automatic operation functions such as optical path design, resource discovery, optical path provisioning, and failure isolation and recovery.

Acknowledgment

We thank the TIP OOPT Disaggregated Optical Systems, Network Operating Systems, and Physical Simulation Environment for their valuable insights and discussions and for working with us to promote TAI and the whitebox approach. We also thank Yuichiro Wada, Shigenari Suzuki, and Xiaosheng Zhang of NTT Communications for their cooperation with application studies and field trials; Seiji Okamoto and Yoshiaki Kisaka for their advice on optimization methods for coherent DSPs; and Tomonori Fujita for his advice on software architecture.

REFERENCES

1. S. Matsuoka, “Ultrahigh-speed ultrahigh-capacity transport network technology for cost-effective core and metro networks,” NTT Tech. Rev.9, 1–7 (2011).

2. K. Kikuchi, “Fundamentals of coherent optical fiber communications,” J. Lightwave Technol.34, 157–179 (2016). [CrossRef]  

3. S. J. Savory, “Digital coherent optical receivers: algorithms and subsystems,” IEEE J. Sel. Top. Quantum Electron. 16, 1164–1179 (2010). [CrossRef]  

4. E. Yamazaki, M. Tomizawa, and Y. Miyamoto, “100-Gb/s optical transport network and beyond employing digital signal processing,” IEEE Commun. Mag.50(2), S43–S49 (2012). [CrossRef]  

5. “Server virtualization,” https://www.vmware.com/topics/glossary/content/server-virtualization.

6. D. Clark, S. Shenker, and A. Falk, “GENI research plan,” 2007, https://groups.geni.net/geni/raw-attachment/wiki/OldGPGDesignDocuments/GDD-06-28.pdf.

7. Open Networking Foundation, https://www.opennetworking.org/.

8. Open Compute Project, https://www.opencompute.org/.

9. “The tactile Internet,” ITU-T Technology Watch Report, 2014, https://www.itu.int/dms_pub/itu-t/oth/23/01/T23010000230001PDFE.pdf.

10. K. Kawazoe, “What’s IOWN?—Change the world,” NTT Tech. Rev.18, 17–26 (2020).

11. Amazon Web Service, “AWS Wavelength,” https://aws.amazon.com/wavelength/?nc1=h_ls.

12. M. Filer, J. Gaudette, Y. Yin, D. Billor, Z. Bakhtiari, and J. L. Cox, “Low-margin optical networking at cloud scale [Invited],” J. Opt. Commun. Netw.11, C94–C108 (2019). [CrossRef]  

13. H. Nishizawa, “Architecting cloud-native optical network with whitebox equipment,” in Optical Fiber Communication Conference (2020), paper 3WC.5.

14. V. Lopez, W. Ishida, A. Mayoral, T. Tanaka, O. G. de Dios, and J. P. Fernandez-Palacios, “Enabling fully programmable transponder white boxes,” J. Opt. Commun. Netw.12, A214–A223 (2020). [CrossRef]  

15. Microsoft, https://azure.microsoft.com/en-us/blog/switch-abstraction-interface-sai-officially-accepted-by-the-open-compute-project-ocp/.

16. GitHub, “SAI (Switch Abstraction Interface),” https://github.com/opencomputeproject/SAI.

17. GitHub, “Software for Open Networking in the Cloud (SONiC),” https://github.com/Azure/SONiC.

18. Y. Sone, H. Nishizawa, S. Yamamoto, M. Fukutoku, and T. Yoshimatsu, “Systems and technologies for high-speed inter-office/datacenter interface,” Proc. SPIE1031, 101310D (2017). [CrossRef]  

19. M. Jinno, H. Takara, Y. Sone, K. Yonenaga, and A. Hirano, “Multiflow optical transponder for efficient multilayer optical networking,” IEEE Commun. Mag.50(5), 56–65 (2012). [CrossRef]  

20. Telecom Infra Project, “OOPT,” https://telecominfraproject.com/oopt/.

21. NTT, “Promoting open standards for inter-datacenter networks in cooperation with global partners,” press release, 2018, https://www.ntt.co.jp/news2018/1810e/181016c.html.

22. GitHub, “TAI (Transponder Abstraction Interface),” https://github.com/Telecominfraproject/oopt-tai.

23. M. Long, “Open optical packet transponder leveraging OCP networking technology,” 2018, https://www.youtube.com/watch?v=krOsB7va9M4.

24. W. Ishida, “IP infusion and NEL: state-of-the-art optical technology,” 2018, https://www.youtube.com/watch?v=W1UvnXz-_ck.

25. I. Lyubomirsky and B. Taylor, “An open approach for switching, routing, and transport,” https://engineering.fb.com/connectivity/an-open-approach-for-switching-routing-and-transport/.

26. Edgecore Networks, “Edgecore Networks announces general availability of Cassini open packet transponder,” https://www.edge-core.com/news-inquiry.php?cls=1&id=352.

27. T. Mabuchi, 2020, https://www.janog.gr.jp/meeting/janog45/application/files/1515/7982/5113/019_whitetrapon_Toshiya_Mabuchi.pdf.

28. G. Francia, R. Nagase, W. Ishida, Y. Sone, L. Kumar, S. Krishnamohan, and V. Lopez, “Disaggregated packet transponder field demonstration exercising multi-format transmission with multi-vendor, open packet optical network elements,” in Optical Fiber Communication Conference (OFC) (2020), paper Th3A.1.

29. NTT Electronics, “NTT Electronics and Acacia Communications announce interoperability mode of operation in next-generation coherent ASICs,” 2013, https://www.ntt-electronics.com/en/news/2013/09/100gdsplsi.html.

30. OIF, “OIF 400ZR,” https://www.oiforum.com/technical-work/hot-topics/400zr-2/.

31. CableLabs, “P2P coherent optics physical layer 2.0 specification,” https://www.cablelabs.com/specifications/P2PCO-SP-PHYv2.0.

32. “Open ROADM optical specification and FlexO oFEC definition,” http://openroadm.org/download.html.

33. OpenZR+, http://openzrplus.org/.

34. GitHub, “Goldstone,” https://github.com/Telecominfraproject/oopt-goldstone.

35. NTT Electronics, “NTT electronics contributes Goldstone—open source network OS for disaggregated coherent transponders to the Telecom Infra Project,” 2019, https://www.ntt-electronics.com/en/news/2019/11/ntt-electronics-contributes-goldstone-open-source-network-os-for-disaggregated-coherent-transponders.html.

36. GitHub, “K3s - Lightweight Kubernetes,” https://github.com/rancher/k3s.

37. GitHub, “Open Network Linux Platform Infrastructure Repository,” https://github.com/opennetworklinux/ONLP.

38. DANOS, https://www.danosproject.org/.

39. Stratum, https://www.opennetworking.org/stratum/.

40. S. Okamoto, F. Hamaoka, and Y. Kisaka, “Field trial of distance-adaptive optical transmission with digital in-band OSNR estimation,” Opt. Express24, 22403–22412 (2016). [CrossRef]  

41. A. Ferrari, M. Filer, K. Balasubramanian, Y. Yin, E. Le Rouzic, J. Kundrat, G. Grammel, G. Galimberti, and V. Curri, “GNPy: an open source application for physical layer aware open optical networks,” J. Opt. Commun. Netw.12, C31–C40 (2020). [CrossRef]  

42. E. R. Hartling, P. Pecci, P. Mehta, D. Evans, V. Kamalov, M. Cantono, E. Mateo, F. Yaman, A. Pilipetskii, C. Mott, P. Lomas, and P. Murphy, “Subsea open cables: a practical perspective on the guidelines and gotchas,” 2019, https://web.asn.com/media/data/files_user/72/SDM1/How_to_Open_Cable_The_Guidelines_and_the_Gotchas_-_04-07-2019_R1.pdf.

43. M. Jinno, B. Kozicki, H. Takara, A. Watanabe, Y. Sone, T. Tanaka, and A. Hirano, “Distance-adaptive spectrum resource allocation in spectrum-sliced elastic optical path network,” IEEE Commun. Mag.48(8), 138–145 (2010). [CrossRef]  

44. M. Filer, J. Gaudette, M. Ghobadi, R. Mahajan, T. Issenhuth, B. Klinkers, and J. Cox, “Elastic optical networking in the Microsoft cloud [Invited],” J. Opt. Commun. Netw.8, A45–A54 (2016). [CrossRef]  

45. T. Tanaka, W. Kawakami, S. Kuwabara, S. Kobayashi, and A. Hirano, “Intelligent monitoring of optical fiber bend using artificial neural networks trained with constellation data,” IEEE Netw. Lett.1, 60–62 (2019). [CrossRef]  

46. F. Inuzuka, T. Oda, T. Tanaka, K. Kitamura, S. Kuwabara, A. Hirano, and M. Tomizawa, “Demonstration of a novel framework for proactive maintenance using failure prediction and bit lossless protection with autonomous network diagnosis system,” J. Lightwave Technol.38, 2695–2702 (2020). [CrossRef]  

47. T. Sasai, M. Nakamura, S. Okamoto, F. Hamaoka, S. Yamamoto, E. Yamazaki, A. Matsushita, H. Nishizawa, and Y. Kisaka, “Simultaneous detection of anomaly points and fiber types in multi-span transmission links only by receiver-side digital signal processing,” in Optical Fiber Communications Conference and Exhibition (OFC) (2020), paper Th1F.1.

48. “GNPy: optical route planning library,” https://gnpy.readthedocs.io/en/master/.

References

  • View by:
  • |
  • |
  • |

  1. S. Matsuoka, “Ultrahigh-speed ultrahigh-capacity transport network technology for cost-effective core and metro networks,” NTT Tech. Rev. 9, 1–7 (2011).
  2. K. Kikuchi, “Fundamentals of coherent optical fiber communications,” J. Lightwave Technol. 34, 157–179 (2016).
    [Crossref]
  3. S. J. Savory, “Digital coherent optical receivers: algorithms and subsystems,” IEEE J. Sel. Top. Quantum Electron. 16, 1164–1179 (2010).
    [Crossref]
  4. E. Yamazaki, M. Tomizawa, and Y. Miyamoto, “100-Gb/s optical transport network and beyond employing digital signal processing,” IEEE Commun. Mag. 50(2), S43–S49 (2012).
    [Crossref]
  5. “Server virtualization,” https://www.vmware.com/topics/glossary/content/server-virtualization .
  6. D. Clark, S. Shenker, and A. Falk, “GENI research plan,” 2007, https://groups.geni.net/geni/raw-attachment/wiki/OldGPGDesignDocuments/GDD-06-28.pdf .
  7. Open Networking Foundation, https://www.opennetworking.org/ .
  8. Open Compute Project, https://www.opencompute.org/ .
  9. “The tactile Internet,” ITU-T Technology Watch Report, 2014, https://www.itu.int/dms_pub/itu-t/oth/23/01/T23010000230001PDFE.pdf .
  10. K. Kawazoe, “What’s IOWN?—Change the world,” NTT Tech. Rev. 18, 17–26 (2020).
  11. Amazon Web Service, “AWS Wavelength,” https://aws.amazon.com/wavelength/?nc1=h_ls .
  12. M. Filer, J. Gaudette, Y. Yin, D. Billor, Z. Bakhtiari, and J. L. Cox, “Low-margin optical networking at cloud scale [Invited],” J. Opt. Commun. Netw. 11, C94–C108 (2019).
    [Crossref]
  13. H. Nishizawa, “Architecting cloud-native optical network with whitebox equipment,” in Optical Fiber Communication Conference (2020), paper 3WC.5.
  14. V. Lopez, W. Ishida, A. Mayoral, T. Tanaka, O. G. de Dios, and J. P. Fernandez-Palacios, “Enabling fully programmable transponder white boxes,” J. Opt. Commun. Netw. 12, A214–A223 (2020).
    [Crossref]
  15. Microsoft, https://azure.microsoft.com/en-us/blog/switch-abstraction-interface-sai-officially-accepted-by-the-open-compute-project-ocp/ .
  16. GitHub, “SAI (Switch Abstraction Interface),” https://github.com/opencomputeproject/SAI .
  17. GitHub, “Software for Open Networking in the Cloud (SONiC),” https://github.com/Azure/SONiC .
  18. Y. Sone, H. Nishizawa, S. Yamamoto, M. Fukutoku, and T. Yoshimatsu, “Systems and technologies for high-speed inter-office/datacenter interface,” Proc. SPIE 1031, 101310D (2017).
    [Crossref]
  19. M. Jinno, H. Takara, Y. Sone, K. Yonenaga, and A. Hirano, “Multiflow optical transponder for efficient multilayer optical networking,” IEEE Commun. Mag. 50(5), 56–65 (2012).
    [Crossref]
  20. Telecom Infra Project, “OOPT,” https://telecominfraproject.com/oopt/ .
  21. NTT, “Promoting open standards for inter-datacenter networks in cooperation with global partners,” press release, 2018, https://www.ntt.co.jp/news2018/1810e/181016c.html .
  22. GitHub, “TAI (Transponder Abstraction Interface),” https://github.com/Telecominfraproject/oopt-tai .
  23. M. Long, “Open optical packet transponder leveraging OCP networking technology,” 2018, https://www.youtube.com/watch?v=krOsB7va9M4 .
  24. W. Ishida, “IP infusion and NEL: state-of-the-art optical technology,” 2018, https://www.youtube.com/watch?v=W1UvnXz-_ck .
  25. I. Lyubomirsky and B. Taylor, “An open approach for switching, routing, and transport,” https://engineering.fb.com/connectivity/an-open-approach-for-switching-routing-and-transport/ .
  26. Edgecore Networks, “Edgecore Networks announces general availability of Cassini open packet transponder,” https://www.edge-core.com/news-inquiry.php?cls=1&id=352 .
  27. T. Mabuchi, 2020, https://www.janog.gr.jp/meeting/janog45/application/files/1515/7982/5113/019_whitetrapon_Toshiya_Mabuchi.pdf .
  28. G. Francia, R. Nagase, W. Ishida, Y. Sone, L. Kumar, S. Krishnamohan, and V. Lopez, “Disaggregated packet transponder field demonstration exercising multi-format transmission with multi-vendor, open packet optical network elements,” in Optical Fiber Communication Conference (OFC) (2020), paper Th3A.1.
  29. NTT Electronics, “NTT Electronics and Acacia Communications announce interoperability mode of operation in next-generation coherent ASICs,” 2013, https://www.ntt-electronics.com/en/news/2013/09/100gdsplsi.html .
  30. OIF, “OIF 400ZR,” https://www.oiforum.com/technical-work/hot-topics/400zr-2/ .
  31. CableLabs, “P2P coherent optics physical layer 2.0 specification,” https://www.cablelabs.com/specifications/P2PCO-SP-PHYv2.0 .
  32. “Open ROADM optical specification and FlexO oFEC definition,” http://openroadm.org/download.html .
  33. OpenZR+, http://openzrplus.org/ .
  34. GitHub, “Goldstone,” https://github.com/Telecominfraproject/oopt-goldstone .
  35. NTT Electronics, “NTT electronics contributes Goldstone—open source network OS for disaggregated coherent transponders to the Telecom Infra Project,” 2019, https://www.ntt-electronics.com/en/news/2019/11/ntt-electronics-contributes-goldstone-open-source-network-os-for-disaggregated-coherent-transponders.html .
  36. GitHub, “K3s - Lightweight Kubernetes,” https://github.com/rancher/k3s .
  37. GitHub, “Open Network Linux Platform Infrastructure Repository,” https://github.com/opennetworklinux/ONLP .
  38. DANOS, https://www.danosproject.org/ .
  39. Stratum, https://www.opennetworking.org/stratum/ .
  40. S. Okamoto, F. Hamaoka, and Y. Kisaka, “Field trial of distance-adaptive optical transmission with digital in-band OSNR estimation,” Opt. Express 24, 22403–22412 (2016).
    [Crossref]
  41. A. Ferrari, M. Filer, K. Balasubramanian, Y. Yin, E. Le Rouzic, J. Kundrat, G. Grammel, G. Galimberti, and V. Curri, “GNPy: an open source application for physical layer aware open optical networks,” J. Opt. Commun. Netw. 12, C31–C40 (2020).
    [Crossref]
  42. E. R. Hartling, P. Pecci, P. Mehta, D. Evans, V. Kamalov, M. Cantono, E. Mateo, F. Yaman, A. Pilipetskii, C. Mott, P. Lomas, and P. Murphy, “Subsea open cables: a practical perspective on the guidelines and gotchas,” 2019, https://web.asn.com/media/data/files_user/72/SDM1/How_to_Open_Cable_The_Guidelines_and_the_Gotchas_-_04-07-2019_R1.pdf .
  43. M. Jinno, B. Kozicki, H. Takara, A. Watanabe, Y. Sone, T. Tanaka, and A. Hirano, “Distance-adaptive spectrum resource allocation in spectrum-sliced elastic optical path network,” IEEE Commun. Mag. 48(8), 138–145 (2010).
    [Crossref]
  44. M. Filer, J. Gaudette, M. Ghobadi, R. Mahajan, T. Issenhuth, B. Klinkers, and J. Cox, “Elastic optical networking in the Microsoft cloud [Invited],” J. Opt. Commun. Netw. 8, A45–A54 (2016).
    [Crossref]
  45. T. Tanaka, W. Kawakami, S. Kuwabara, S. Kobayashi, and A. Hirano, “Intelligent monitoring of optical fiber bend using artificial neural networks trained with constellation data,” IEEE Netw. Lett. 1, 60–62 (2019).
    [Crossref]
  46. F. Inuzuka, T. Oda, T. Tanaka, K. Kitamura, S. Kuwabara, A. Hirano, and M. Tomizawa, “Demonstration of a novel framework for proactive maintenance using failure prediction and bit lossless protection with autonomous network diagnosis system,” J. Lightwave Technol. 38, 2695–2702 (2020).
    [Crossref]
  47. T. Sasai, M. Nakamura, S. Okamoto, F. Hamaoka, S. Yamamoto, E. Yamazaki, A. Matsushita, H. Nishizawa, and Y. Kisaka, “Simultaneous detection of anomaly points and fiber types in multi-span transmission links only by receiver-side digital signal processing,” in Optical Fiber Communications Conference and Exhibition (OFC) (2020), paper Th1F.1.
  48. “GNPy: optical route planning library,” https://gnpy.readthedocs.io/en/master/ .

2020 (4)

2019 (2)

T. Tanaka, W. Kawakami, S. Kuwabara, S. Kobayashi, and A. Hirano, “Intelligent monitoring of optical fiber bend using artificial neural networks trained with constellation data,” IEEE Netw. Lett. 1, 60–62 (2019).
[Crossref]

M. Filer, J. Gaudette, Y. Yin, D. Billor, Z. Bakhtiari, and J. L. Cox, “Low-margin optical networking at cloud scale [Invited],” J. Opt. Commun. Netw. 11, C94–C108 (2019).
[Crossref]

2017 (1)

Y. Sone, H. Nishizawa, S. Yamamoto, M. Fukutoku, and T. Yoshimatsu, “Systems and technologies for high-speed inter-office/datacenter interface,” Proc. SPIE 1031, 101310D (2017).
[Crossref]

2016 (3)

2012 (2)

E. Yamazaki, M. Tomizawa, and Y. Miyamoto, “100-Gb/s optical transport network and beyond employing digital signal processing,” IEEE Commun. Mag. 50(2), S43–S49 (2012).
[Crossref]

M. Jinno, H. Takara, Y. Sone, K. Yonenaga, and A. Hirano, “Multiflow optical transponder for efficient multilayer optical networking,” IEEE Commun. Mag. 50(5), 56–65 (2012).
[Crossref]

2011 (1)

S. Matsuoka, “Ultrahigh-speed ultrahigh-capacity transport network technology for cost-effective core and metro networks,” NTT Tech. Rev. 9, 1–7 (2011).

2010 (2)

S. J. Savory, “Digital coherent optical receivers: algorithms and subsystems,” IEEE J. Sel. Top. Quantum Electron. 16, 1164–1179 (2010).
[Crossref]

M. Jinno, B. Kozicki, H. Takara, A. Watanabe, Y. Sone, T. Tanaka, and A. Hirano, “Distance-adaptive spectrum resource allocation in spectrum-sliced elastic optical path network,” IEEE Commun. Mag. 48(8), 138–145 (2010).
[Crossref]

Bakhtiari, Z.

Balasubramanian, K.

Billor, D.

Cox, J.

Cox, J. L.

Curri, V.

de Dios, O. G.

Fernandez-Palacios, J. P.

Ferrari, A.

Filer, M.

Francia, G.

G. Francia, R. Nagase, W. Ishida, Y. Sone, L. Kumar, S. Krishnamohan, and V. Lopez, “Disaggregated packet transponder field demonstration exercising multi-format transmission with multi-vendor, open packet optical network elements,” in Optical Fiber Communication Conference (OFC) (2020), paper Th3A.1.

Fukutoku, M.

Y. Sone, H. Nishizawa, S. Yamamoto, M. Fukutoku, and T. Yoshimatsu, “Systems and technologies for high-speed inter-office/datacenter interface,” Proc. SPIE 1031, 101310D (2017).
[Crossref]

Galimberti, G.

Gaudette, J.

Ghobadi, M.

Grammel, G.

Hamaoka, F.

S. Okamoto, F. Hamaoka, and Y. Kisaka, “Field trial of distance-adaptive optical transmission with digital in-band OSNR estimation,” Opt. Express 24, 22403–22412 (2016).
[Crossref]

T. Sasai, M. Nakamura, S. Okamoto, F. Hamaoka, S. Yamamoto, E. Yamazaki, A. Matsushita, H. Nishizawa, and Y. Kisaka, “Simultaneous detection of anomaly points and fiber types in multi-span transmission links only by receiver-side digital signal processing,” in Optical Fiber Communications Conference and Exhibition (OFC) (2020), paper Th1F.1.

Hirano, A.

F. Inuzuka, T. Oda, T. Tanaka, K. Kitamura, S. Kuwabara, A. Hirano, and M. Tomizawa, “Demonstration of a novel framework for proactive maintenance using failure prediction and bit lossless protection with autonomous network diagnosis system,” J. Lightwave Technol. 38, 2695–2702 (2020).
[Crossref]

T. Tanaka, W. Kawakami, S. Kuwabara, S. Kobayashi, and A. Hirano, “Intelligent monitoring of optical fiber bend using artificial neural networks trained with constellation data,” IEEE Netw. Lett. 1, 60–62 (2019).
[Crossref]

M. Jinno, H. Takara, Y. Sone, K. Yonenaga, and A. Hirano, “Multiflow optical transponder for efficient multilayer optical networking,” IEEE Commun. Mag. 50(5), 56–65 (2012).
[Crossref]

M. Jinno, B. Kozicki, H. Takara, A. Watanabe, Y. Sone, T. Tanaka, and A. Hirano, “Distance-adaptive spectrum resource allocation in spectrum-sliced elastic optical path network,” IEEE Commun. Mag. 48(8), 138–145 (2010).
[Crossref]

Inuzuka, F.

Ishida, W.

V. Lopez, W. Ishida, A. Mayoral, T. Tanaka, O. G. de Dios, and J. P. Fernandez-Palacios, “Enabling fully programmable transponder white boxes,” J. Opt. Commun. Netw. 12, A214–A223 (2020).
[Crossref]

G. Francia, R. Nagase, W. Ishida, Y. Sone, L. Kumar, S. Krishnamohan, and V. Lopez, “Disaggregated packet transponder field demonstration exercising multi-format transmission with multi-vendor, open packet optical network elements,” in Optical Fiber Communication Conference (OFC) (2020), paper Th3A.1.

Issenhuth, T.

Jinno, M.

M. Jinno, H. Takara, Y. Sone, K. Yonenaga, and A. Hirano, “Multiflow optical transponder for efficient multilayer optical networking,” IEEE Commun. Mag. 50(5), 56–65 (2012).
[Crossref]

M. Jinno, B. Kozicki, H. Takara, A. Watanabe, Y. Sone, T. Tanaka, and A. Hirano, “Distance-adaptive spectrum resource allocation in spectrum-sliced elastic optical path network,” IEEE Commun. Mag. 48(8), 138–145 (2010).
[Crossref]

Kawakami, W.

T. Tanaka, W. Kawakami, S. Kuwabara, S. Kobayashi, and A. Hirano, “Intelligent monitoring of optical fiber bend using artificial neural networks trained with constellation data,” IEEE Netw. Lett. 1, 60–62 (2019).
[Crossref]

Kawazoe, K.

K. Kawazoe, “What’s IOWN?—Change the world,” NTT Tech. Rev. 18, 17–26 (2020).

Kikuchi, K.

Kisaka, Y.

S. Okamoto, F. Hamaoka, and Y. Kisaka, “Field trial of distance-adaptive optical transmission with digital in-band OSNR estimation,” Opt. Express 24, 22403–22412 (2016).
[Crossref]

T. Sasai, M. Nakamura, S. Okamoto, F. Hamaoka, S. Yamamoto, E. Yamazaki, A. Matsushita, H. Nishizawa, and Y. Kisaka, “Simultaneous detection of anomaly points and fiber types in multi-span transmission links only by receiver-side digital signal processing,” in Optical Fiber Communications Conference and Exhibition (OFC) (2020), paper Th1F.1.

Kitamura, K.

Klinkers, B.

Kobayashi, S.

T. Tanaka, W. Kawakami, S. Kuwabara, S. Kobayashi, and A. Hirano, “Intelligent monitoring of optical fiber bend using artificial neural networks trained with constellation data,” IEEE Netw. Lett. 1, 60–62 (2019).
[Crossref]

Kozicki, B.

M. Jinno, B. Kozicki, H. Takara, A. Watanabe, Y. Sone, T. Tanaka, and A. Hirano, “Distance-adaptive spectrum resource allocation in spectrum-sliced elastic optical path network,” IEEE Commun. Mag. 48(8), 138–145 (2010).
[Crossref]

Krishnamohan, S.

G. Francia, R. Nagase, W. Ishida, Y. Sone, L. Kumar, S. Krishnamohan, and V. Lopez, “Disaggregated packet transponder field demonstration exercising multi-format transmission with multi-vendor, open packet optical network elements,” in Optical Fiber Communication Conference (OFC) (2020), paper Th3A.1.

Kumar, L.

G. Francia, R. Nagase, W. Ishida, Y. Sone, L. Kumar, S. Krishnamohan, and V. Lopez, “Disaggregated packet transponder field demonstration exercising multi-format transmission with multi-vendor, open packet optical network elements,” in Optical Fiber Communication Conference (OFC) (2020), paper Th3A.1.

Kundrat, J.

Kuwabara, S.

F. Inuzuka, T. Oda, T. Tanaka, K. Kitamura, S. Kuwabara, A. Hirano, and M. Tomizawa, “Demonstration of a novel framework for proactive maintenance using failure prediction and bit lossless protection with autonomous network diagnosis system,” J. Lightwave Technol. 38, 2695–2702 (2020).
[Crossref]

T. Tanaka, W. Kawakami, S. Kuwabara, S. Kobayashi, and A. Hirano, “Intelligent monitoring of optical fiber bend using artificial neural networks trained with constellation data,” IEEE Netw. Lett. 1, 60–62 (2019).
[Crossref]

Le Rouzic, E.

Lopez, V.

V. Lopez, W. Ishida, A. Mayoral, T. Tanaka, O. G. de Dios, and J. P. Fernandez-Palacios, “Enabling fully programmable transponder white boxes,” J. Opt. Commun. Netw. 12, A214–A223 (2020).
[Crossref]

G. Francia, R. Nagase, W. Ishida, Y. Sone, L. Kumar, S. Krishnamohan, and V. Lopez, “Disaggregated packet transponder field demonstration exercising multi-format transmission with multi-vendor, open packet optical network elements,” in Optical Fiber Communication Conference (OFC) (2020), paper Th3A.1.

Mahajan, R.

Matsuoka, S.

S. Matsuoka, “Ultrahigh-speed ultrahigh-capacity transport network technology for cost-effective core and metro networks,” NTT Tech. Rev. 9, 1–7 (2011).

Matsushita, A.

T. Sasai, M. Nakamura, S. Okamoto, F. Hamaoka, S. Yamamoto, E. Yamazaki, A. Matsushita, H. Nishizawa, and Y. Kisaka, “Simultaneous detection of anomaly points and fiber types in multi-span transmission links only by receiver-side digital signal processing,” in Optical Fiber Communications Conference and Exhibition (OFC) (2020), paper Th1F.1.

Mayoral, A.

Miyamoto, Y.

E. Yamazaki, M. Tomizawa, and Y. Miyamoto, “100-Gb/s optical transport network and beyond employing digital signal processing,” IEEE Commun. Mag. 50(2), S43–S49 (2012).
[Crossref]

Nagase, R.

G. Francia, R. Nagase, W. Ishida, Y. Sone, L. Kumar, S. Krishnamohan, and V. Lopez, “Disaggregated packet transponder field demonstration exercising multi-format transmission with multi-vendor, open packet optical network elements,” in Optical Fiber Communication Conference (OFC) (2020), paper Th3A.1.

Nakamura, M.

T. Sasai, M. Nakamura, S. Okamoto, F. Hamaoka, S. Yamamoto, E. Yamazaki, A. Matsushita, H. Nishizawa, and Y. Kisaka, “Simultaneous detection of anomaly points and fiber types in multi-span transmission links only by receiver-side digital signal processing,” in Optical Fiber Communications Conference and Exhibition (OFC) (2020), paper Th1F.1.

Nishizawa, H.

Y. Sone, H. Nishizawa, S. Yamamoto, M. Fukutoku, and T. Yoshimatsu, “Systems and technologies for high-speed inter-office/datacenter interface,” Proc. SPIE 1031, 101310D (2017).
[Crossref]

H. Nishizawa, “Architecting cloud-native optical network with whitebox equipment,” in Optical Fiber Communication Conference (2020), paper 3WC.5.

T. Sasai, M. Nakamura, S. Okamoto, F. Hamaoka, S. Yamamoto, E. Yamazaki, A. Matsushita, H. Nishizawa, and Y. Kisaka, “Simultaneous detection of anomaly points and fiber types in multi-span transmission links only by receiver-side digital signal processing,” in Optical Fiber Communications Conference and Exhibition (OFC) (2020), paper Th1F.1.

Oda, T.

Okamoto, S.

S. Okamoto, F. Hamaoka, and Y. Kisaka, “Field trial of distance-adaptive optical transmission with digital in-band OSNR estimation,” Opt. Express 24, 22403–22412 (2016).
[Crossref]

T. Sasai, M. Nakamura, S. Okamoto, F. Hamaoka, S. Yamamoto, E. Yamazaki, A. Matsushita, H. Nishizawa, and Y. Kisaka, “Simultaneous detection of anomaly points and fiber types in multi-span transmission links only by receiver-side digital signal processing,” in Optical Fiber Communications Conference and Exhibition (OFC) (2020), paper Th1F.1.

Sasai, T.

T. Sasai, M. Nakamura, S. Okamoto, F. Hamaoka, S. Yamamoto, E. Yamazaki, A. Matsushita, H. Nishizawa, and Y. Kisaka, “Simultaneous detection of anomaly points and fiber types in multi-span transmission links only by receiver-side digital signal processing,” in Optical Fiber Communications Conference and Exhibition (OFC) (2020), paper Th1F.1.

Savory, S. J.

S. J. Savory, “Digital coherent optical receivers: algorithms and subsystems,” IEEE J. Sel. Top. Quantum Electron. 16, 1164–1179 (2010).
[Crossref]

Sone, Y.

Y. Sone, H. Nishizawa, S. Yamamoto, M. Fukutoku, and T. Yoshimatsu, “Systems and technologies for high-speed inter-office/datacenter interface,” Proc. SPIE 1031, 101310D (2017).
[Crossref]

M. Jinno, H. Takara, Y. Sone, K. Yonenaga, and A. Hirano, “Multiflow optical transponder for efficient multilayer optical networking,” IEEE Commun. Mag. 50(5), 56–65 (2012).
[Crossref]

M. Jinno, B. Kozicki, H. Takara, A. Watanabe, Y. Sone, T. Tanaka, and A. Hirano, “Distance-adaptive spectrum resource allocation in spectrum-sliced elastic optical path network,” IEEE Commun. Mag. 48(8), 138–145 (2010).
[Crossref]

G. Francia, R. Nagase, W. Ishida, Y. Sone, L. Kumar, S. Krishnamohan, and V. Lopez, “Disaggregated packet transponder field demonstration exercising multi-format transmission with multi-vendor, open packet optical network elements,” in Optical Fiber Communication Conference (OFC) (2020), paper Th3A.1.

Takara, H.

M. Jinno, H. Takara, Y. Sone, K. Yonenaga, and A. Hirano, “Multiflow optical transponder for efficient multilayer optical networking,” IEEE Commun. Mag. 50(5), 56–65 (2012).
[Crossref]

M. Jinno, B. Kozicki, H. Takara, A. Watanabe, Y. Sone, T. Tanaka, and A. Hirano, “Distance-adaptive spectrum resource allocation in spectrum-sliced elastic optical path network,” IEEE Commun. Mag. 48(8), 138–145 (2010).
[Crossref]

Tanaka, T.

F. Inuzuka, T. Oda, T. Tanaka, K. Kitamura, S. Kuwabara, A. Hirano, and M. Tomizawa, “Demonstration of a novel framework for proactive maintenance using failure prediction and bit lossless protection with autonomous network diagnosis system,” J. Lightwave Technol. 38, 2695–2702 (2020).
[Crossref]

V. Lopez, W. Ishida, A. Mayoral, T. Tanaka, O. G. de Dios, and J. P. Fernandez-Palacios, “Enabling fully programmable transponder white boxes,” J. Opt. Commun. Netw. 12, A214–A223 (2020).
[Crossref]

T. Tanaka, W. Kawakami, S. Kuwabara, S. Kobayashi, and A. Hirano, “Intelligent monitoring of optical fiber bend using artificial neural networks trained with constellation data,” IEEE Netw. Lett. 1, 60–62 (2019).
[Crossref]

M. Jinno, B. Kozicki, H. Takara, A. Watanabe, Y. Sone, T. Tanaka, and A. Hirano, “Distance-adaptive spectrum resource allocation in spectrum-sliced elastic optical path network,” IEEE Commun. Mag. 48(8), 138–145 (2010).
[Crossref]

Tomizawa, M.

Watanabe, A.

M. Jinno, B. Kozicki, H. Takara, A. Watanabe, Y. Sone, T. Tanaka, and A. Hirano, “Distance-adaptive spectrum resource allocation in spectrum-sliced elastic optical path network,” IEEE Commun. Mag. 48(8), 138–145 (2010).
[Crossref]

Yamamoto, S.

Y. Sone, H. Nishizawa, S. Yamamoto, M. Fukutoku, and T. Yoshimatsu, “Systems and technologies for high-speed inter-office/datacenter interface,” Proc. SPIE 1031, 101310D (2017).
[Crossref]

T. Sasai, M. Nakamura, S. Okamoto, F. Hamaoka, S. Yamamoto, E. Yamazaki, A. Matsushita, H. Nishizawa, and Y. Kisaka, “Simultaneous detection of anomaly points and fiber types in multi-span transmission links only by receiver-side digital signal processing,” in Optical Fiber Communications Conference and Exhibition (OFC) (2020), paper Th1F.1.

Yamazaki, E.

E. Yamazaki, M. Tomizawa, and Y. Miyamoto, “100-Gb/s optical transport network and beyond employing digital signal processing,” IEEE Commun. Mag. 50(2), S43–S49 (2012).
[Crossref]

T. Sasai, M. Nakamura, S. Okamoto, F. Hamaoka, S. Yamamoto, E. Yamazaki, A. Matsushita, H. Nishizawa, and Y. Kisaka, “Simultaneous detection of anomaly points and fiber types in multi-span transmission links only by receiver-side digital signal processing,” in Optical Fiber Communications Conference and Exhibition (OFC) (2020), paper Th1F.1.

Yin, Y.

Yonenaga, K.

M. Jinno, H. Takara, Y. Sone, K. Yonenaga, and A. Hirano, “Multiflow optical transponder for efficient multilayer optical networking,” IEEE Commun. Mag. 50(5), 56–65 (2012).
[Crossref]

Yoshimatsu, T.

Y. Sone, H. Nishizawa, S. Yamamoto, M. Fukutoku, and T. Yoshimatsu, “Systems and technologies for high-speed inter-office/datacenter interface,” Proc. SPIE 1031, 101310D (2017).
[Crossref]

IEEE Commun. Mag. (3)

E. Yamazaki, M. Tomizawa, and Y. Miyamoto, “100-Gb/s optical transport network and beyond employing digital signal processing,” IEEE Commun. Mag. 50(2), S43–S49 (2012).
[Crossref]

M. Jinno, H. Takara, Y. Sone, K. Yonenaga, and A. Hirano, “Multiflow optical transponder for efficient multilayer optical networking,” IEEE Commun. Mag. 50(5), 56–65 (2012).
[Crossref]

M. Jinno, B. Kozicki, H. Takara, A. Watanabe, Y. Sone, T. Tanaka, and A. Hirano, “Distance-adaptive spectrum resource allocation in spectrum-sliced elastic optical path network,” IEEE Commun. Mag. 48(8), 138–145 (2010).
[Crossref]

IEEE J. Sel. Top. Quantum Electron (1)

S. J. Savory, “Digital coherent optical receivers: algorithms and subsystems,” IEEE J. Sel. Top. Quantum Electron. 16, 1164–1179 (2010).
[Crossref]

IEEE Netw. Lett. (1)

T. Tanaka, W. Kawakami, S. Kuwabara, S. Kobayashi, and A. Hirano, “Intelligent monitoring of optical fiber bend using artificial neural networks trained with constellation data,” IEEE Netw. Lett. 1, 60–62 (2019).
[Crossref]

J. Lightwave Technol. (2)

J. Opt. Commun. Netw. (4)

NTT Tech. Rev. (2)

S. Matsuoka, “Ultrahigh-speed ultrahigh-capacity transport network technology for cost-effective core and metro networks,” NTT Tech. Rev. 9, 1–7 (2011).

K. Kawazoe, “What’s IOWN?—Change the world,” NTT Tech. Rev. 18, 17–26 (2020).

Opt. Express (1)

Proc. SPIE (1)

Y. Sone, H. Nishizawa, S. Yamamoto, M. Fukutoku, and T. Yoshimatsu, “Systems and technologies for high-speed inter-office/datacenter interface,” Proc. SPIE 1031, 101310D (2017).
[Crossref]

Other (33)

H. Nishizawa, “Architecting cloud-native optical network with whitebox equipment,” in Optical Fiber Communication Conference (2020), paper 3WC.5.

Microsoft, https://azure.microsoft.com/en-us/blog/switch-abstraction-interface-sai-officially-accepted-by-the-open-compute-project-ocp/ .

GitHub, “SAI (Switch Abstraction Interface),” https://github.com/opencomputeproject/SAI .

GitHub, “Software for Open Networking in the Cloud (SONiC),” https://github.com/Azure/SONiC .

Amazon Web Service, “AWS Wavelength,” https://aws.amazon.com/wavelength/?nc1=h_ls .

“Server virtualization,” https://www.vmware.com/topics/glossary/content/server-virtualization .

D. Clark, S. Shenker, and A. Falk, “GENI research plan,” 2007, https://groups.geni.net/geni/raw-attachment/wiki/OldGPGDesignDocuments/GDD-06-28.pdf .

Open Networking Foundation, https://www.opennetworking.org/ .

Open Compute Project, https://www.opencompute.org/ .

“The tactile Internet,” ITU-T Technology Watch Report, 2014, https://www.itu.int/dms_pub/itu-t/oth/23/01/T23010000230001PDFE.pdf .

Telecom Infra Project, “OOPT,” https://telecominfraproject.com/oopt/ .

NTT, “Promoting open standards for inter-datacenter networks in cooperation with global partners,” press release, 2018, https://www.ntt.co.jp/news2018/1810e/181016c.html .

GitHub, “TAI (Transponder Abstraction Interface),” https://github.com/Telecominfraproject/oopt-tai .

M. Long, “Open optical packet transponder leveraging OCP networking technology,” 2018, https://www.youtube.com/watch?v=krOsB7va9M4 .

W. Ishida, “IP infusion and NEL: state-of-the-art optical technology,” 2018, https://www.youtube.com/watch?v=W1UvnXz-_ck .

I. Lyubomirsky and B. Taylor, “An open approach for switching, routing, and transport,” https://engineering.fb.com/connectivity/an-open-approach-for-switching-routing-and-transport/ .

Edgecore Networks, “Edgecore Networks announces general availability of Cassini open packet transponder,” https://www.edge-core.com/news-inquiry.php?cls=1&id=352 .

T. Mabuchi, 2020, https://www.janog.gr.jp/meeting/janog45/application/files/1515/7982/5113/019_whitetrapon_Toshiya_Mabuchi.pdf .

G. Francia, R. Nagase, W. Ishida, Y. Sone, L. Kumar, S. Krishnamohan, and V. Lopez, “Disaggregated packet transponder field demonstration exercising multi-format transmission with multi-vendor, open packet optical network elements,” in Optical Fiber Communication Conference (OFC) (2020), paper Th3A.1.

NTT Electronics, “NTT Electronics and Acacia Communications announce interoperability mode of operation in next-generation coherent ASICs,” 2013, https://www.ntt-electronics.com/en/news/2013/09/100gdsplsi.html .

OIF, “OIF 400ZR,” https://www.oiforum.com/technical-work/hot-topics/400zr-2/ .

CableLabs, “P2P coherent optics physical layer 2.0 specification,” https://www.cablelabs.com/specifications/P2PCO-SP-PHYv2.0 .

“Open ROADM optical specification and FlexO oFEC definition,” http://openroadm.org/download.html .

OpenZR+, http://openzrplus.org/ .

GitHub, “Goldstone,” https://github.com/Telecominfraproject/oopt-goldstone .

NTT Electronics, “NTT electronics contributes Goldstone—open source network OS for disaggregated coherent transponders to the Telecom Infra Project,” 2019, https://www.ntt-electronics.com/en/news/2019/11/ntt-electronics-contributes-goldstone-open-source-network-os-for-disaggregated-coherent-transponders.html .

GitHub, “K3s - Lightweight Kubernetes,” https://github.com/rancher/k3s .

GitHub, “Open Network Linux Platform Infrastructure Repository,” https://github.com/opennetworklinux/ONLP .

DANOS, https://www.danosproject.org/ .

Stratum, https://www.opennetworking.org/stratum/ .

E. R. Hartling, P. Pecci, P. Mehta, D. Evans, V. Kamalov, M. Cantono, E. Mateo, F. Yaman, A. Pilipetskii, C. Mott, P. Lomas, and P. Murphy, “Subsea open cables: a practical perspective on the guidelines and gotchas,” 2019, https://web.asn.com/media/data/files_user/72/SDM1/How_to_Open_Cable_The_Guidelines_and_the_Gotchas_-_04-07-2019_R1.pdf .

T. Sasai, M. Nakamura, S. Okamoto, F. Hamaoka, S. Yamamoto, E. Yamazaki, A. Matsushita, H. Nishizawa, and Y. Kisaka, “Simultaneous detection of anomaly points and fiber types in multi-span transmission links only by receiver-side digital signal processing,” in Optical Fiber Communications Conference and Exhibition (OFC) (2020), paper Th1F.1.

“GNPy: optical route planning library,” https://gnpy.readthedocs.io/en/master/ .

Cited By

OSA participates in Crossref's Cited-By Linking service. Citing articles from OSA journals and other participating publishers are listed here.

Alert me when this article is cited.


Figures (15)

Fig. 1.
Fig. 1. Trends in optical transport technologies.
Fig. 2.
Fig. 2. Virtualization technology and OSS.
Fig. 3.
Fig. 3. Whitebox packet transponder architecture.
Fig. 4.
Fig. 4. Examples of hardware complexity for coherent communications.
Fig. 5.
Fig. 5. TAI architecture.
Fig. 6.
Fig. 6. Whitebox packet transponders Cassini and Galileo.
Fig. 7.
Fig. 7. Interconnection demonstration at TIP Summit 2019.
Fig. 8.
Fig. 8. Multivendor interconnection demonstration at OFC 2019.
Fig. 9.
Fig. 9. Goldstone architecture.
Fig. 10.
Fig. 10. OPEX reduction with the Open Transport Platform.
Fig. 11.
Fig. 11. Resource discovery, optical path provisioning.
Fig. 12.
Fig. 12. Auto-optical path optimization testing with 600 Gbps DSP.
Fig. 13.
Fig. 13. Optical fiber bending detection using an artificial neural network.
Fig. 14.
Fig. 14. Principle of DBP coefficient optimization and loss profile estimation.
Fig. 15.
Fig. 15. Loss profile estimation [standard single-mode fiber (SSMF) only, 280 km].

Tables (1)

Tables Icon

Table 1. List of Interconnection Modes