Abstract

Service provider edge and access networks are continuing to evolve. With bandwidth demands increasing exponentially due to exciting new applications being offered to consumers and enterprise customers, their legacy network infrastructures, based on older optical transmission technology, can’t scale to the bandwidth that will be needed to handle these demands in the future. New technologies and strategies are needed to upgrade both the wired and wireless network infrastructure in the segments currently supporting 10G links so that it can support optical links at 100G and beyond, while also accommodating the many different fiber types that exist such as point to point, DWDM, and BiDi links. This paper discusses how advancements in coherent optical technology are paving the way to meet these challenges, providing service providers with the path they need to meet growing bandwidth demands today and in the future.

Service Provider Edge and Access Demand Continues to Grow

Applications are driving the need for increased bandwidth towards the edges of the network that are closer to the end user. New and next-generation applications serving the consumer and enterprise customers are primarily responsible for the growth in edge and access bandwidth requirements. According to Cisco’s Annual Internet Report¹, ultra-high-definition internet-connected 4K TVs are globally forecasted to account for 66% of flat-panel TVs (891 million) by 2023. Connected home applications such as home automation, home security, and video surveillance are forecasted to represent nearly half of all machine-to-machine connections by 2023 (estimated 14.7 billion total machine-to-machine connections). In addition, according to CableLabs², by December 2019, cable gigabit had reached 93% of all US housing units passed by cable broadband providers, with available downstream speeds growing by approximately 50% per year (CAGR).

Applications outside of the home, such as vehicle navigation/diagnostics/entertainment and fleet management, are forecasted to represent the fastest growing machine-to-machine segment with growth at 30% CAGR through 2023, according to the Annual Internet Report. By 2023, Internet of Things (IoT) devices used in multiple industry segments are forecasted to account for 50% (14.7 billion) of all global networked devices.

In addition to the aforementioned bandwidth growth drivers, other applications such as gaming, telemedicine, and autonomous vehicles, as well as high-capacity enterprise services for hybrid cloud connections to centralized cloud networks are also expected to drive up the demand for bandwidth in service provider edge and access networks.

The wired and wireless network infrastructure being deployed to support aggregated residential customer traffic and enterprise business services is driving bandwidth capacity higher than legacy infrastructure can support based on traditional optical transmission technology. For example, 5G wireless connections are forecasted to generate approximately 3 times more traffic than 4G connections. All of these bandwidth growth drivers are creating a challenge for legacy optical service provider edge and access infrastructure to support this traffic.

Wide Range of Access Challenges with a Coherent Theme

A common theme among access network architectures is a desire by the network operator to stay ahead of the curve when it comes to ensuring sufficient network capacity to accommodate rising bandwidth demand. To achieve this, many access architectures require capacity upgrades and they may include at least one of the following link designs, each with its own set of challenges:

• Dedicated point-to-point (P2P) fiber links,
• Higher-capacity dense wave-division-multiplexed (DWDM) links, or
• Fiber-constrained routes requiring single-fiber bidirectional (BiDi) P2P or DWDM links.

Figure 1. Examples of different connectivity solutions in the service provider edge/access portion of the network.

This paper examines the different challenges of increasing bandwidth in these types of access network topologies, and discusses how coherent optical technology can provide a scalable solution to address high-bandwidth demands, and operational and scalability benefits in these networks when increasing bandwidth to 100Gbps and beyond.

Dedicated Point-to-Point Fiber

Optical fiber deployments that do not require optical amplification or DWDM to reach the service provider edge/access terminal equipment rely on economically optimized dedicated P2P fiber links in which a fiber pair is used between the metro core and edge/access aggregation terminal equipment or enterprise site. Here we use the term dedicated P2P fiber to denote a data path between endpoints of a fiber link for a single application or customer service. As described later in this paper, in some optical implementations the optical link may utilize multiple widely spaced wavelengths constituting a single data path over a dedicated P2P fiber link.

Figure 2.  Typical dedicated point-to-point fiber link.

As bandwidth demand in these types of networks increases, limitations are encountered because legacy optical transmission technology have distance limitations when scaling to 100Gbps and beyond. For service provider edge and access networks, there is an emerging requirement to evolve directly from legacy 10Gbps to 100Gbps optical links as a preferred means to increase bandwidth. An attractive solution would be one that can provide a 10x increase in bandwidth capacity without requiring a change to the network architecture, thus minimizing the total cost of ownership.

Higher data rates pushing the limits of legacy dedicated P2P fiber solutions

Traditional direct-detect optical transmission technology used in pluggable optical transceiver module solutions over single mode fiber have served the industry well in providing reliable dedicated P2P fiber at 10Gbps over service provider edge/access networks. However, beyond 10Gbps, the ease of use becomes more challenging. Increasing the bandwidth of a dedicated P2P fiber link from 10Gbps to 100Gbps using direct-detect relies on transmitting over multiple optical lanes at sub-multiples of 100Gbps (see Direct-Detect and Coherent Comparison for 100Gbps Edge/Access inset), such as four lanes at 25Gbps. Using the same technology to address edge/access reaches beyond 10km is challenging. To minimize impairments due to fiber chromatic dispersion, these four wavelengths need to remain in the 1310nm range. Extending to further reaches in this wavelength range would require additional power budget to overcome optical fiber loss. For example, extending to a longer reach of 80km would not only require an additional 28dB^* power budget compared to 10km, but also there would be additional loss from the effects of optical transmission impairments such as chromatic dispersion and polarization mode dispersion (PMD)

Unique requirements in the edge/access

In addition to the technological limitations of pluggable direct-detect solutions scaling to higher bandwidths and reaches, the landscape of the service provider edge/access network provides its own unique challenges. Different fiber types with a range of loss and dispersion characteristics have been deployed over the years in various segments of the network to optimize transmission for different generations of transmission technology. The range of fiber types include ITU-T G.652A/B/C/D, G.653, G.654, G.655, G.656, and G.657. Because a large capital investment is required to install fiber, it does not make sense to rip out the fiber whenever a new technology advancement is made. Rather, accommodations on the terminal equipment optical transceivers/transponders may be required to operate on legacy non-optimized fiber.

It is not unusual to have multiple bulkhead patch-panel fiberoptic connectors and splice points along a fiber route within the edge/access network, as shown in Figure 3. The cumulative effect of these is the accumulation of loss and back reflections (aggravated by unclean connectors), which can be detrimental to the optical transmission performance in a direct-detect link.

^*Assuming 0.4 dB/km attenuation.

Figure 3.  Additional margin in an access link would provide operational flexibility in deployments from increasing the number of addressable reaches to potentially avoiding the need for truck rolls to “shoot fiber” (OTDR measurements), especially for routes deemed marginal using direct-detect transmission.

To account for numerous potential impairments depending on the fiber-plant condition (fiber types, connector reflections and losses), truck rolls may be required to characterize each fiber route (chromatic dispersion, PMD, fiber loss, and reflections) before certifying a link as operational to ensure an optical link can be closed if there is uncertainty about link margin.

In addition to fiber types and connector/splice induced impairments, environmental conditions of the service provider edge/access network must also be considered. Edge/access equipment terminals may be located in uncontrolled outdoor cabinets requiring optical modules to endure temperature ranges beyond what is typically found in an indoor temperature-controlled environment. Managing the performance of a multiple-laser direct-detect solution to meet overall transmission requirements, including longer reach links, may make outdoor temperature resilience a challenge.

DWDM Access Aggregation Links

It is common for multiple dedicated P2P fiber links to aggregate to a network node (as shown in Figure 1), such as the service-provider edge or access aggregation site where traffic is combined into larger multi-transmission bandwidth pipes, transported via a DWDM link to the core of the network. These DWDM links can be either amplified or unamplified (Figure 4).

Figure 4.  Typical DWDM link which may or may not be amplified.

These links are more complex than dedicated P2P fiber links since they require additional optical mux/demuxing components and may also include optical amplifiers for link extension without electrical regeneration along a fiber path.

Each individual DWDM wavelength experiences the similar impairments as in a single-wavelength dedicated P2P fiber link. With tunable laser capabilities, a common module can be used to cover multiple DWDM channels which lowers deployment and sparing costs.

Single-Fiber BiDi Links

Gaining right-of-way access and digging up streets to install fiber optic cables are major hurdles in deploying service provider edge and access optical infrastructure to address the continuing growth in bandwidth demand. This predicament leads to constrained fiber situations in a variety of environments such as urban, suburban, rural, and metropolitan. Both previously discussed dedicated P2P and DWDM links in the service provider edge and access networks typically utilize single-mode duplex fiber in which data transmission from site A to Z travels on a fiber that is different from the data transmission traveling from site Z to A (Figures 2 and 4).

It is not unusual for a service provider to rely on a single strand of fiber optic cable in order to deliver services over this infrastructure, especially if the service provider is sharing a cable bundle or duct space with others. In these single-fiber routes, optical transmission is bi-directionally transmitted and received on the same fiber, as opposed to different fibers in a more typical duplex fiber route. A compounding challenge is how to perform upgrades over a single-fiber BiDi route from legacy bandwidth/distance-limited optical direct-detection technology, given the previously discussed fiber impairments which also affect a BiDi route.

Rather than digging up the streets, an alternative method to increasing bandwidth over service provider edge/access single-fiber routes is to upgrade the BiDi transmission technology used at the terminal equipment endpoints.

Figures 5a and 5b illustrate examples of BiDi deployments for single optical link and DWDM links. Figure 5a illustrates an example of a single-fiber path available in which the A-to-Z transmission and Z-to-A transmission travel bi-directionally along the same fiber route over different wavelengths. Figure 5b illustrates a similar A-to-Z scenario except that DWDM transmission is used.

(a)(b)

Figure 5.  Optical transmission in a service provider edge/access network over (a) a single fiber BiDi link and (b) a DWDM single fiber BiDi link—in both cases, the transmit wavelength is different from receive wavelength for each data transmission stream. The optical implementation to combine signals into a single fiber depends on the link requirements.

As Figures 5 illustrates, the optical transceiver modules used for single-fiber BiDi deployments must have the ability to transmit and receive on independent wavelengths, a capability dependent on the module design.

Coherent Technology Solves Service Provider Edge and Access Challenges

Optical coherent technology has come a long way since its early days when a full line-card of electronics and optics was required. Today, this technology can be housed in a small, compact pluggable module, made possible through advancements in silicon photonics, opto-electronic integration, and CMOS nodes with lower power consumption. These continued innovations have positioned coherent solutions to advance into applications with shorter reaches (Figure 6) such as service provider edge and access networks.

Figure 6.  Coherent solutions are evolving towards shorter reaches.

Unlike bandwidth/distance limited direct-detect solutions for service provider edge and access links, coherent technology can easily bridge the gap to higher bandwidth and longer distances on any deployed fiber type. Coherent also provides an operationally simple solution, which played an important role in driving its adoption in longer-reach environments. Let’s explore some of these advantages in more detail.

Figure 7.  Error-free coherent transmission of 100Gbps QPSK modulation example, tolerant to multiple impairments (only one transmission direction shown).

Coherent transmission tolerant to various service provider edge/access route impairments

Coherent solutions have the ability to electronically overcome both chromatic and PMD transmission impairments, which allows the transmission to adapt over different edge/access fiber types and conditions in a plug-and-play fashion. It is also tolerant to the detrimental effects from loss and back reflections from multiple fiber connector/splice interfaces.

Unlike in intensity-modulated direct-detect transmission where reflections encountered over the fiber route can create noise in the transmission link, coherent modulation formats such as QPSK are inherently much more tolerant to optical reflections. Due to the single-laser coherent transmitter operating in the lowest loss 1550nm window in single mode fiber, and the coherent receiver having extremely high sensitivity due to its coherent detection technology, coherent pluggable modules have ample power budgets. This enables them to not only compensate for losses due to multiple fiber connectors/splices, but also address long transmission links.

Figure 8 illustrates how the effects of dispersion and losses along a fiber route result in a reach limitation for 100Gbps direct-detect solutions. In contrast, the coherent solution with its higher tolerance to impairments provides improved performance in the form of additional margin and longer reach capabilities.

Figure 8.  Direct-detect solutions hit a reach limit at a certain distance while the coherent solution will continue to operate beyond this reach limit.

Some of the other key benefits of coherent technology include the following:

Monitoring, Diagnostics and Troubleshooting.  Built into pluggable coherent transceivers are monitoring and diagnostic capabilities to ensure robust data transmission. As previously stated, coherent 100Gbps solutions have a very wide receiver dynamic range compared to an equivalent direct-detect link, which enables a coherent link to accommodate optical loopback for troubleshooting. In comparison, some direct-detect solutions include internal optical amplification at the receiver to close longer links, resulting in direct optical loopback troubleshooting not being possible due to receiver overload.

Reliability.  In the case of a dedicated P2P fiber link, a pluggable direct-detect 100Gbps solution relies on splitting the single-transmission traffic onto four transmitter lasers which may require operating at the higher end of their transmitter optical power range, especially for the longer reach links of an edge/access network. For these reaches, active optical amplification at the receiver end may also be required to close the link. Thus, a total of eight active elements must be taken into account when determining the reliability of these types of modules. In comparison, a pluggable coherent 100Gbps solution for duplex operation utilizes only one active optical element, the transmission laser, making it more reliable than the direct-detect solution. All of these benefits lead to operational simplicity and shorter provisioning times, which can result in operational savings in the service provider edge/access network.

Scaling to higher data rates

Pluggable coherent solutions enable higher data rates over the same, or greater, distances. Compared to today’s access data rates, higher rate coherent options are already available in small form factor pluggable modules, providing a ready-made path to meeting the demands of service provider edge/access bandwidth growth. Coherent transmission solutions beyond 100Gbps are quite mature, and thus, there is no fundamental near-term impediment in coherent technology for scaling service provider edge and access to higher bandwidths.

Acacia’s Solution

Acacia’s service provider edge and access coherent pluggable solutions are designed for service provider edge and access applications, with a range of modules to address different network applications such as single-transmission P2P links, DWDM links, and single-fiber BiDi links. Acacia’s 100Gbps coherent pluggable, offered in a quad small form-factor double density (QSFP-DD) that is widely used for client-optics solutions, are specifically designed for optimization in service provider edge and access applications with unamplified links including 80km reaches and beyond, as well as amplified or unamplified DWDM links. They were designed to provide network operators the ability to scale to higher data rates to meet growing bandwidth demands over some of the most challenging optical links, while also providing operational simplicity that may lead to overall network savings.

Figure 9. Integration in silicon enables a path to coherent transceivers for service provider edge/access in compact pluggable QSFP-DD and CFP2 modules.

Acacia’s coherent BiDi CFP2 form-factor module delivers an operationally efficient and cost-effective way for network operators to increase capacity to 100Gbps and beyond over both single-fiber as well as diverse transmit/receive network architectures. For more details on how a coherent BiDi solution differs from a duplex fiber P2P or DWDM solution, see the Design optimization for coherent BiDi inset for a detailed explanation of how to enable this capability in a coherent transceiver.

For single-fiber BiDi transmission, there are various implementation options to convert from a dual-port configuration to a single-port configuration such as passive optical mux/demuxes, splitter/combiners, or circulators, depending on the application and link requirements. The module supports both Ethernet and OTN standardized client protocols.

Acacia’s 100Gbps P2P and DWDM QSFP-DD modules and coherent BiDi CFP2 module all feature Acacia’s 3D Siliconization approach, which utilizes high-volume manufacturing processes and benefits from the maturity of Acacia’s silicon photonics technology. Figure 10 illustrates how advances in optical/electrical component consolidation has resulted in size reductions of coherent modules. The use of silicon photonics to take discrete bulky optical components and integrate their functions into a CMOS-based silicon chip has been a key factor in module footprint reduction.

3D Siliconization follows the example of the electronics world and applies integration techniques such as 3D stacking to electronics and silicon photonic integrated circuit (PIC) co-packaging. This approach makes it possible to integrate crucial components into a compact package and reduce the number of electrical inter-connects while preserving robust signal integrity. Integration involving silicon photonics follows progress already made in the semiconductor fabrication process suitable for volume production and high yields.

Figure 10.  Evolution of coherent module size over time.

Conclusion

Applications such as 5G, gaming, telemedicine, autonomous vehicles, and cloud computing are driving increased network traffic at the edges of the network, which is placing enormous strain on traditional service provider edge and access networks. The amount of bandwidth required to deliver these new services and applications are expected to be higher than the legacy infrastructure can support based on traditional optical transmission technology. This is driving service providers to not only look for ways to deliver optical links at 100Gbps, which pose challenges to legacy direct-detect solutions, but also to ensure these new solutions can be supported over existing dedicated P2P fiber, DWDM, and BiDi infrastructure.

Based on recent advancements in silicon photonics, opto-electronic integration, and CMOS nodes with lower power consumption, coherent technology has emerged as an effective way of meeting this challenge. These advancements have enabled Acacia to develop and bring to market a full suite of edge and access coherent pluggable modules designed to address different network applications such as dedicated P2P fiber links, DWDM links, and single-fiber BiDi links. Leveraging these solutions can enable network operators to overcome the many challenges they face today when trying to scale to higher data rates in the future in a cost effective and operationally simplistic way.

References

¹ Cisco website:  https://www.cisco.com/c/en/us/solutions/collateral/executive-perspectives/annual-internet-report/white-paper-c11-741490.html.

² CableLabs website:  https://www.cablelabs.com/gigabit-internet-speeds.

Download the Whitepaper

This paper describes how silicon photonic (SiPh) opto-electronic integration and packaging, with its improved RF performance, is designed to: (1) enable next-generation coherent transmission beyond 100Gbaud; and (2) provide network operators with the ability to avoid “half-step” capacity upgrades and allow them the full step upgrades they need without sacrificing reach or stranding network bandwidth when migrating from current-generation solutions.

Introduction

Similar to consumer electronic devices, high-capacity optical transceiver modules have followed a path towards size reduction with increased performance and functionality. This has been driven by growing bandwidth demands in data center and service provider networks as a result of applications such as cloud services, content streaming, and social media. To meet these growing demands, network operators are seeking high-performance, high-density and low-power coherent optical interconnect solutions. Coherent optics suppliers are responding with large-scale investments and advancements in silicon photonic (SiPh) opto-electronic integration and packaging to introduce compact, high-performance, volume-manufacturable designs that can enable them to meet this growing demand.

Advancements in coherent optics are important for network operators to efficiently evolve their networks as bandwidth demand continues to grow. However, as the gap to Shannon’s Limit (the theoretical limit of channel capacity) decreases, it is becoming more difficult to increase capacity and reduce cost per bit by primarily focusing on increasing the modulation order. Beyond the coherent modulation order of 64QAM, fewer transport applications can be supported due to the resulting reduction of achievable link distances, which leaves higher baud rates and advances in packaging technology as preferred ways to increase capacity and decrease cost per bit. As a result, industry investments are shifting toward increasing the transmission aggregate baud rate to increase capacity while maintaining usable reaches. A direct benefit of these investments and advancements includes electrical high-frequency radio frequency (RF) performance improvements inside the module.

This paper describes how SiPh opto-electronic integration and packaging, with its improved RF performance, is designed to: (1) enable next-generation coherent transmission beyond 100Gbaud; and (2) provide network operators with the ability to avoid “half-step” capacity upgrades and allow them the full step upgrades they need without sacrificing reach or stranding network bandwidth when migrating from current-generation solutions.

Higher Baud Rates for Smooth Network Evolution

The requirements for higher baud rates of optical transceivers to increase the transmission distance and to reduce the cost-per-transmitted bit in turn drives the need for wider optical channels, i.e., wider pass bands or channel spacing. Due to long upgrade intervals as a result of high CapEx and OpEx, careful consideration must be taken in the channel/passband requirements for next-generation optical line system deployments. Even though today’s reconfigurable optical add-drop multiplexers (ROADMs) can support a flexible range of channel widths, network operators often find that adhering to standard intervals of channel widths is a preferable option. This enables them to utilize channel spacing based on regular multiples that can be combined without stranding spectrum due to fragmentation¹. Earlier generations of dense wavelength division multiplexing networks have deployed 50, 75, and 100GHz channels.

To accommodate future transmission schemes while minimizing stranded bandwidth using half-step upgrades, network operators may seek to ensure wider channel plans are integer multiples of recently established plans, such as two times 75GHz as illustrated in Figure 1.

Figure 1.  Channel plan evolution to accommodate coherent transmission trends; migrating to wider channels enables scalability.

150GHz-wide channels are now seen as the next evolutionary step for future-proofing the line system for higher-capacity network architectures, doubling the channel bandwidth (thereby accommodating a doubling of baud rate) of current 75GHz implementations. A wide-channel-capable line system can also provide the network operator with the ability to accommodate multiple coherent transmission techniques.

By dialing down the modulation order and doubling the baud rate, solutions can expand current capacities over a much greater distance. A 150GHz channel plan can accommodate these higher aggregate baud rates. Avoiding intermediate steps of baud rate increases can remedy the issue of stranded bandwidth when a 150GHz channel plan is used¹. In addition to minimizing stranded bandwidth, a wide 150GHz channel also allows for higher capacity per transponder which in turn reduces the cost-per-bit as well as the number of channels required to maximize fiber capacity. This can lower operational cost/complexity and result in a reduction in required ROADM ports.

Current coherent transmission solutions operate in the range of 60Gbaud through 70Gbaud, and doubling the rate obviously also introduces a range to more efficiently utilize a 150GHz channel. For brevity, in the remainder of this article we refer to this doubling range as greater-than 100Gbaud. It is also important to emphasize that baud rates below 100Gbaud do not fully utilize a 150GHz channel passband.

High-speed modulators and RF interconnects for next-gen optics

The two material systems primarily used for high-speed coherent photonic integrated circuit (PIC) designs are silicon (Si) and indium phosphide (InP). Supporting baud rates greater than 100Gbaud requires advancements in modulator design with higher frequency responses compared to the designs currently being used.

Figure 2 illustrates how high bandwidth SiPh optical modulator designs can support >100Gbaud coherent transmission capabilities to efficiently fill 150GHz channel width line-system architectures.

Figure 2. Illustration of simulated frequency response of a high-bandwidth SiPh optical modulator capable of supporting >100Gbaud coherent transmission speeds.

Even with an optical modulator having ample frequency response to support >100Gbaud transmission, there is another key factor that could limit the optical transmission speed. The electrical interconnects of the PIC and surrounding RF components such as electrical drivers can become a limiting factor in transmission performance and design complexity. Additionally, the RF performance at these speeds is sensitive to impairments along the electrical path between the coherent digital signal processor (DSP) and the PIC as well as its surrounding RF components. In traditional designs (Figure 3), the high-speed RF electrical signal may traverse between these components over multiple physical interfaces such as solder ball-grid array (BGA) bumps/balls, substrate, printed circuit board traces, wire bonds, and gold-box lead-frame pins, with each element introducing parasitics and losses that degrade signal integrity. This applies to next-generation optical solutions in general, regardless of whether the transmission technique is coherent or not. Fortunately, techniques to mitigate opto-electronic RF signal impairments already exist in the electronics industry.

Figure 3.  Illustration of a conventional high-speed RF electrical interconnect between coherent DSP and PIC.

Opto-electronic RF interconnect evolution for higher speeds

An advantage of SiPh is its ability to leverage mature, volume-manufacturable processes from the electronics industry, such as component stacking (Figure 4). In component stacking, electrical impairments are reduced due to direct electrical connections between key RF components, creating a robust signal path for extremely high frequency/baud rate operation. In this co-packaged/stacked design, the gold-box packaging is eliminated, the DSP and PIC are tightly co-packaged, and the high-speed Si modulator driver and TIA components are stacked on the PIC.

Figure 4.  Illustration of a component stacking configuration

Figure 5 shows the stacked design has a higher (better) frequency response than the traditional gold-box design. Advanced stacking designs to further eliminate interconnect impairments can result in pushing the frequency response even higher.

Figure 5.  Illustration of example electrical interconnect frequency response comparing traditional gold-box and stacking integration.

It is noteworthy to mention that the conversion of the analog RF high-speed signals into digital signals for coherent DSP processing are performed by high-speed, high-resolution digital-to-analog convertors (DACs) and analog-to-digital convertors (ADCs). These Si-based devices are able to support >100Gbaud operation, especially when utilizing next generation CMOS node sizes such as 5nm material and are therefore not a limiting element at these speeds.

Architectural comparisons of SiPh and InP

As previously stated, coherent optics designs are generally based on either InP or Si. InP is a direct bandgap semiconductor material well suited for use in optical components such as lasers and modulators. Using InP, these functions can be fabricated on the same chip (see figure 6). In contrast, Si is an indirect-bandgap semiconductor material, commonly used to fabricate various electronic components including DSPs and other ASICs. Si is also used to fabricate SiPh devices such as modulators and coherent receivers. However, silicon does not allow for the straightforward fabrication of lasers. As a result, in a SiPh coherent design, the laser is typically separate from the SiPh PIC and made from a different material such as InP. This separation can be an advantage in thermally challenging environments because the temperature-sensitive InP laser can be designed to be further away from heat-producing ASICs, while the temperature insensitive silicon photonics can be placed closer to the DSP. The co-packaging of the Si/SiPh components enable efficient heat dissipation designs, which can lower overall power consumption. Although the InP laser needs to be located separately, the optical connection is easily achieved using well-known design and manufacturing processes.

Figure 6.  Comparison of InP-based and Si-based coherent elements that require high-speed RF electrical interconnects. Red outlined boxes have greater thermal tolerance compared to blue outlined boxes.

Co-packaging of all the Si/SiPh components, as shown in Figure 4, allows stacked high-speed electrical paths that provide high-performance RF interconnects, compared to an InP-based design that may further separate components due to material or thermal mismatches and hermeticity requirements. As previously mentioned, SiPh leverages high-volume manufacturing processes already established in the electronics industry. Acacia’s 3D Siliconization is an example of leveraging these processes to manufacture compact, coherent modules with high-performance electrical RF interconnects. The 3D Siliconization approach takes advantage of mature silicon photonics technology, in-house expertise, and vertical integration, while utilizing high-volume electronics manufacturing processes.

Evolving Networks for Future Growth

Next generation coherent interconnects are designed to offer network operators the ability to evolve their networks to meet growing bandwidth demands. Since this must be done without sacrificing reach or stranding network bandwidth when migrating from current-generation solutions, coherent suppliers have been focusing on increasing the transmission aggregate baud rate to increase capacity while maintaining usable reaches. They have been able to achieve this through continued advancements in SiPh opto-electronic integration and packaging, which has proven to deliver more compact, high-performance, and volume-manufacturable designs.

In the future, SiPh opto-electronic integration and packaging can continue to be important for developing the increased performance and functionality needed for data centers and service providers to meet growing bandwidth demands. Instead of taking half a step to upgrade their network, these advancements can enable them to take the full step with a new generation of coherent optics designed to run beyond 100Gbaud.

References

¹ Acacia Communications (2020), “How to Build a Lasting Optical Network,” Retrieved from: https://acacia-inc.com/blog/how-to-build-a-lasting-optical-network/

Download the Whitepaper

Abstract

In fiber constrained networks, network operators often have a single fiber route, with the ability to transmit and receive data in both directions. Historically, these networks have been served by 10Gbps bi-directional (BiDi) optical modules. With bandwidth demands growing, coherent bi-directional pluggable optical modules can provide these networks with an upgrade path to 100Gbps and beyond. This paper covers the ability to scale to higher bandwidths and address BiDi transmission, with operational and scalability benefits in these networks.

Demand growing for single-fiber BiDi communications

Gaining right-of-way access and digging up streets to install fiber optic cables are major hurdles in deploying service provider edge and access optical infrastructure to address the continuing growth in bandwidth demand. This predicament leads to constrained fiber situations in a variety of environments such as urban, suburban, rural, and metropolitan. It is not unusual for a service provider to rely on a single strand of fiber optic cable in order to deliver services over this infrastructure, especially if the service provider is sharing a cable bundle or duct space with others. In these single-fiber routes, optical transmission is bi-directional (BiDi)–transmitted and received signals co-exist on the same fiber, as opposed to different fibers in a more typical duplex fiber route.

Today’s applications such as uploading 4K-resolution videos into the cloud for social-network sharing and the increased use of two-way video chats and file sharing tied to work-from-home requirements are examples of applications beyond existing streaming entertainment that are driving the need for bandwidth increases, especially when traffic from a large number of end users is aggregated. In addition to demand growth from these applications, the traffic patterns are evolving. Legacy access networks have typically dedicated larger bandwidth to downstream traffic. The aforementioned applications have driven the need for higher bandwidth capabilities to upstream traffic. Network infrastructure being deployed to support 5G wireless and enterprise business customers is also driving bandwidths higher than legacy infrastructure can support based on traditional optical transmission technology. 5G networks are expected to have traffic from a higher number of endpoints to aggregate, compared to legacy 4G LTE. All of these bandwidth growth drivers are creating a challenge for legacy optical service provider edge and access infrastructure to support this traffic. For example, cable network operators (aka, multiple system operators, MSOs) currently utilizing 10Gbps optical links are seeking to upgrade to 100Gbps links that operate over edge and access fiber routes. A compounding challenge is how to perform these upgrades over a single-fiber route.

Rather than digging up the streets, an alternative method to increasing bandwidth over service provider edge/access single-fiber routes is to upgrade the transmission technology used at the terminal equipment endpoints. The challenge is that legacy optical direct-detection technology is not capable of keeping up with increasing data rates in supporting BiDi transmission.

Figure 1.  Examples of different connectivity solutions in the service provider edge/access portion of the network.

This paper examines how coherent BiDi pluggable edge/access solutions address both the ability to scale to higher bandwidths and address BiDi transmission, with operational and scalability benefits in these networks when increasing bandwidth to 100Gbps and beyond.

Limitations of legacy BiDi edge/access solutions

Traditional 10Gbps direct-detect optical transmission technology used in pluggable optical transceiver module solutions over single mode fiber has provided reliable 10Gbps BiDi links over service provider edge/access networks. Implementing direct-detect solutions at higher data rates is more challenging due to the need to utilize parallel optical lanes at lower data rates, e.g. 4 x 25G. Coherent transmission technology can be used to transmit serial 100Gbps with support for BiDi architectures. Thus, coherent provides a more effective path to migrate today’s BiDi routes to 100Gbps.

Figures 2 and 3 illustrate examples of BiDi deployments compared to traditional duplex fiber deployments. Figure 2a illustrates a typical duplex fiber route in which transmission from “A-to-Z” and transmission from “Z-to-A” utilize the same wavelength. However, in fiber constrained environments there may be only a single-fiber path available in which the A-to-Z transmission and Z-to-A transmission travel bi-directionally along the same fiber route over different wavelengths, as shown in Figure 2b.

(a)

(b)

Figure 2.  Optical transmission in a service provider edge/access network over (a) duplex optical fiber; and (b) a single fiber BiDi link in which transmit wavelength is different from receive wavelength and combined onto the same fiber.

Using dense wavelength division multiplexing (DWDM), the number of links can be increased in both duplex and single-fiber deployments, as shown in Figure 3.

(a)

(b)

Figure 3.  Illustrations showing wavelength-multiplexed transmission over (a) duplex optical fiber; and (b) a single-fiber BiDi link in which transmit wavelength is different from receive wavelength—the optical implementation to combine signals into a single fiber depends on the link requirements.

As Figures 2 and 3 illustrate, the optical transceiver modules used for single-fiber BiDi deployments must have the ability to transmit and receive on independent wavelengths, a capability dependent on the module design. This will be covered in more detail in a later section of this white paper.

Operational simplicity with coherent BiDi edge/access solutions

Optical coherent technology has come a long way since its early days when a full line-card of electronics and optics was required. Today, this technology can be housed in a small, compact pluggable module¹, made possible through advancements in silicon photonics, opto-electronic integration, and CMOS nodes with lower power consumption. These continued innovations have positioned coherent solutions to advance into applications with shorter reaches (Figure 4) such as service provider edge and access networks.

Figure 4.  Coherent solutions are evolving towards shorter reaches.

As previously stated, direct-detect solutions for service provider edge and access links are reaching bandwidth/distance limitations, compared to coherent solutions that can easily bridge the gap to higher bandwidth and longer distances on any deployed fiber type. Coherent also addresses the unique requirements of service provider BiDi edge/access applications (to be discussed in a later section), while providing an operationally simple solution that leverages the many capabilities that make coherent a successful solution in longer-reach environments. Coherent solutions are user friendly in deployment and provisioning due to wide tolerance range, increased optical margin, and rich monitoring and diagnostic features.

Figure 5.  Error-free coherent transmission of 100Gbps QPSK modulation example, tolerant to multiple impairments (only one transmission direction shown).

Coherent solutions have the ability to electronically overcome both chromatic and PMD transmission impairments, which allows the transmission to adapt over different edge/access fiber types and conditions in a plug-and-play fashion. It is also tolerant to the detrimental effects from loss and back reflections from multiple fiber connector/splice interfaces. All these capabilities result in coherent solutions providing additional margin compared to direct-detect solutions.

Pluggable coherent solutions enable the capability to achieve higher data rates in the future over the same, or greater, distances. Compared to today’s access data rates, higher rate coherent options are already available in small form factor pluggable modules, providing a ready-made path to meeting the demands of service provider edge/access bandwidth growth. Coherent transmission solutions beyond 100Gbps are quite mature, and thus, there is no fundamental near-term impediment in coherent technology for scaling service provider edge and access to higher bandwidths.

Coherent solutions are user friendly in deployment and provisioning due to wide tolerance range over a variety of fiber types and long fiber links, increased optical margin, plug-and-play implementation, and rich monitoring and diagnostic features. Coherent solutions also provide a roadmap to addressing higher data rates and longer reaches over the same edge/access infrastructure. This all leads to operational simplicity and shorter provisioning times, which can result in savings in the service provider edge/access network.

Design optimization for coherent BiDi

A required element of any coherent optical transmission implementation is the local oscillator (LO), a stable continuous wave (CW) laser used as a clean reference source to optically mix with the received optical signal. It is typical for the transmitter and receiver wavelength in a coherent link to be the same, which allows for the sharing of the same CW laser source for transmission as well as the receiver LO (Figure 6a). However, for BiDi applications in which the transmit and received wavelengths are different, the LO should not be derived from the transmission laser. A separate second laser is required as the LO. Thus, a coherent BiDi module contains dual lasers, one used to transmit the data, while the other is used as the LO. This allows the transmit wavelength to be independent from the receive wavelength.

(a)

(b)

Figure 6.  (a) High-level block diagram of optical coherent module transmission in which the received wavelength is the same as the transmitted wavelength; CW laser is shared as the optical modulator source and the LO source. (b) A coherent bi-directional module in which the received signal is a different wavelength than the transmitted wavelength, and the LO source cannot be shared with the transmitter laser source.

This coherent BiDi module is designed to support service provider edge and access optical network links with single-fiber routes, as illustrated in Figures 2b and 3b. Wavelengths propagating in opposite directions have to be different in order to avoid in-band crosstalk due to back reflections. In addition, having the Tx laser and receiving module’s LO laser be fully tunable enables the management of a single product code rather than managing an inventory of multiple fixed-wavelength pluggable module product codes, thus simplifying network deployment.

Acacia’s Solution

Acacia’s coherent BiDi pluggable solutions are designed for service provider edge and access applications that require independent transmit/receive wavelengths. Acacia’s coherent BiDi module is housed in a CFP2 form-factor module and delivers an operationally efficient and cost-effective way for network operators to increase capacity to 100Gbps and beyond over both single-fiber as well as diverse transmit/receive network architectures.

Figure 7: Integration in silicon enables a path to coherent transceivers for service provider edge/access in a compact pluggable CFP2 module.

For single-fiber BiDi transmission, there are various implementation options to convert from a dual-port configuration to a single-port configuration such as passive optical mux/demuxes, splitter/combiners, or circulators, depending on the application and link requirements. The module supports both Ethernet and OTN client protocols. Acacia’s new coherent BiDi service provider edge and access solutions were designed to provide network operators the ability to scale to higher data rates to meet growing bandwidth demands over challenging BiDi optical links, while also providing operational simplicity that may lead to overall network savings.

Acacia’s 3D Siliconization approach, which utilizes high-volume manufacturing processes and benefits from the maturity of Acacia’s silicon photonics technology, is used in the coherent BiDi CFP2 module. The use of silicon photonics to take discrete bulky optical components and integrate their functions into a CMOS-based silicon chip has been a key factor in module footprint reduction.

3D Siliconization follows the example of the electronics world and applies integration techniques such as 3D stacking to electronics and silicon photonic integrated circuit (PIC) co-packaging. This approach makes it possible to integrate crucial components into a compact package and reduce the number of electrical inter-connects while preserving robust signal integrity. Integration involving silicon photonics follows progress already made in the semiconductor fabrication process suitable for volume production and high yields.

Conclusion

Constrained fiber environments in service provider edge and access networks pose challenges to address increasing bandwidth demand, especially over single-fiber routes. Legacy technology is not capable of supporting data rates at 100Gbps in BiDi applications in these networks. This has driven a need for optical links at 100Gbps and beyond that can address network architectures such as BiDi links that are operationally simple to deploy, along with a path to easily and cost-effectively scale to higher speeds in the future.

Acacia has a proven history of bringing coherent technology to new markets, delivering the low power, performance, operational simplicity, and scalability customers have come to rely on. Since 2011, Acacia has leveraged the benefits of integration to evolve from 100Gbps in a 5”x7” form factor to pluggable form factors such as CFP2 and QSFP-DD supporting up to 400Gbps. This has been enabled by Acacia’s mature and highly integrated silicon photonics designs. As a trusted partner for service providers and network equipment manufacturers (NEMs), Acacia is now leveraging its technology to develop coherent BiDi 100Gbps and beyond transceivers for the service provider edge/access market.

References

¹Acacia Communications (2020), “Coherent Technology for Point-to-Point Edge and Access Network Applications,” Retrieved from: https://acacia-inc.com/acacia-resources/coherent-technology-for-point-to-point-edge-and-access-network-applications/

Applications such as gaming, telemedicine, and autonomous vehicles are driving the need for increased bandwidth towards the edges of the network that are closer to the end user. To meet these needs, there’s an emerging requirement for service provider edge and access networks to evolve directly from legacy 10Gbps to 100Gbps optical links as the way to increase bandwidth. This paper examines the constraints that prohibit legacy Point-to-Point edge/access solutions from scaling to higher bandwidths, and discusses how coherent optical solutions can provide operational and scalability benefits in these networks when increasing bandwidth to 100Gbps and beyond.

Bandwidth demands in edge and access

Applications are clearly driving the need for increased bandwidth towards the edges of the network that are closer to the end user. For example, applications such as gaming, telemedicine, and autonomous vehicles need larger pipes from centralized cloud networks to connect to edge computing sites that are closer to the end user.  In addition, data from these applications may also be transported via a wireless infrastructure which would require large wireline pipes to carry data from cell towers back to the core of the network. And for service providers offering high-capacity enterprise services via a private line fiber link to the customer premise, applications such as hybrid cloud services can drive increased bandwidth demands. Optical fiber deployments that do not require optical amplification or multiplexing using dense wavelength division multiplexing (DWDM) to reach the edge/access terminal equipment rely on economically optimized point-to-point (P2P) links where a dedicated fiber pair is used between the metro core and edge/access aggregation terminal equipment.

As bandwidth demand in these P2P network links increases, limitations are encountered because legacy optical transmission technology is not capable of supporting data rates beyond a certain speed for a given distance. For service provider edge and access networks, there is an emerging requirement to evolve directly from legacy 10Gbps to 100Gbps optical links as a preferred means to increase bandwidth. The alternative of utilizing multiple 10Gbps links (e.g., for enterprise access) may not be the most efficient from a cost-per-gigabit perspective as 100Gbps solutions become more ubiquitous. The challenge is that traditional technologies typically used in this portion of the network have distance limitations when scaling to 100Gbps and beyond.

Figure 1.  Examples of point-to-point links in the service provider edge/access portion of the network.

This paper examines the constraints that prohibit legacy P2P edge/access solutions from scaling to higher bandwidths, and discusses how coherent optical solutions can provide operational and scalability benefits in these networks when increasing bandwidth to 100Gbps and beyond.

Higher data rates pushing the limits of legacy P2P solutions

Traditional direct-detect optical transmission technology used in pluggable optical transceiver module solutions over single mode fiber have served the industry well in providing reliable P2P 10Gbps links over service provider edge/access networks. However, beyond 10Gbps, the ease of use becomes more challenging. Increasing the bandwidth of an optical P2P link from 10Gbps to 100Gbps using direct-detect relies on transmitting over multiple optical lanes at sub-multiples of 100Gbps (see Direct-Detect and Coherent Comparison for 100Gbps Edge/Access inset), such as four lanes at 25Gbps. The widely used 100GBASE-LR4 is an example of a direct-detect solution that supports 100Gbps links up to 10km using four transmitter lasers (operating in the 1310nm wavelength range) and receiver PIN detectors. Using the same technology to address edge/access reaches beyond 10km is challenging. To minimize impairments due to fiber chromatic dispersion, these four wavelengths need to remain in the 1310nm range. Extending to further reaches in this wavelength range requires additional power budget to overcome optical fiber loss. For example, extending to a longer reach of 80km would require an additional 28dB¹ power budget in comparison with 100GBASE-LR4. In addition to loss, the effects of optical transmission impairments such as chromatic dispersion slope and polarization mode dispersion (PMD) add to the loss penalty.

In an effort to stretch the achievable 100Gbps direct-detect reach beyond 10km towards 40km, 100GBASE-ER4 was introduced to address power budget by increasing both the transmitter power and receiver sensitivity. The challenge with this approach is that the maximum transmitter power is limited by laser eye-safety regulations, while improvements to receiver sensitivity are accomplished by using avalanche photo diode (APD) receivers or semiconductor optical amplifiers (SOAs). Even with these approaches, addressing the longer reaches (e.g. ≥80km) remain challenging.

¹Assuming 0.4 dB/km attenuation.

Unique requirements in the edge/access

In addition to the technological limitations of pluggable direct-detect solutions scaling to higher bandwidths and reaches, the landscape of the service provider edge/access network provides some unique challenges. Different fiber types with a range of loss and dispersion characteristics have been deployed over the years in various segments of the network to optimize transmission for different generations of transmission technology. The range of fiber types include ITU-T G.652A/B/C/D, G.653, G.654, G.655, G.656, and G.657.  Because a large capital investment is required to install fiber, it does not make sense to rip out the fiber whenever a new technology advancement is made. Rather, accommodations on the terminal equipment optical transceivers/transponders may be required to operate on legacy non-optimized fiber.

It is not unusual to have multiple bulkhead patch-panel fiberoptic connectors and splice points along a fiber route within the edge/access network. The cumulative effect of multiple connectors and splice points is the accumulation of loss and back reflections (aggravated by unclean connectors), which can be detrimental to the optical transmission performance in a direct-detect link.

To account for numerous potential impairments depending on the fiber-plant condition (fiber types, connector reflections and losses), truck rolls may be required in order to characterize each fiber route —using an optical time-domain reflectometer (OTDR)— before certifying a link as operational to ensure an optical link can be closed if there is uncertainty about link margin.

In addition to fiber types and connector/splice induced impairments, consideration of environmental conditions of the service provider edge/access network must also be taken into account. Edge/access equipment terminals may be located in uncontrolled outdoor cabinets requiring optical modules to endure temperature ranges beyond what is typically found in an indoor temperature-controlled environment. Managing the performance of a multiple-laser direct-detect solution to meet overall transmission requirements, including longer reach links, may make outdoor temperature resilience a challenge.

Operational simplicity with coherent edge/access solutions

Coherent technology has existed in long-haul networks for approximately a decade and is also widely deployed in metro networks. Once requiring a full line-card of electronics and optics, coherent transmission technology can now be housed in a small compact pluggable module about the size of a pack of stick gum.

Figure 2.  Coherent solutions are evolving towards shorter reaches.

This has been made possible through advancements in CMOS nodes with lower power consumption, opto-electronic integration, and silicon photonics technology. These continued innovations have positioned coherent solutions to advance into applications with shorter reaches (Figure 2) such as service provider edge and access networks.

While direct-detect solutions for service provider edge and access links are reaching bandwidth/distance limitations, coherent solutions on the other hand can easily bridge the gap to higher bandwidth and longer distances on any deployed fiber type. In addition to providing a path to higher bandwidth and longer distance capabilities, coherent also addresses the unique requirements of service provider edge/access applications while providing an operationally simple solution that leverages the many capabilities that made coherent a successful solution in longer reach environments. Coherent P2P solutions are user friendly in deployment and provisioning due to wide tolerance range, increased optical margin, loop-back capabilities, and rich monitoring and diagnostic features. Let’s explore some of these advantages in more detail.

Coherent transmission tolerant to various service provider edge/access route impairments.  A feature of coherent technology is its ability to electronically overcome both chromatic and PMD transmission impairments, which allows the transmission to adapt over different fiber types and conditions in a plug-and-play fashion. Figure 3 illustrates how 100Gbps coherent transmission can overcome multiple fiber transmission impairments over a fiber route.

Figure 3.  Error-free coherent transmission of 100Gbps using QPSK modulation tolerant to multiple impairments.

Coherent technology is also tolerant to the detrimental effects of loss and back reflections from multiple fiber connector/splice interfaces. Unlike in intensity-modulated direct-detect transmission where reflections encountered over the fiber route can create noise in the transmission link, coherent modulation formats such as QPSK are inherently much more tolerant to optical reflections. Due to the single-laser coherent transmitter operating in the lowest loss 1550nm window in single mode fiber, and the coherent receiver having extremely high sensitivity due to its coherent detection technology, coherent pluggable modules have ample power budgets to not only compensate for losses due to multiple fiber connectors/splices but also address long transmission links.

Figure 4.  Coherent solutions provide additional margin resulting in operational flexibility in deployments from increasing the number of addressable reaches to potentially avoiding the need for truck rolls to “shoot fiber” (OTDR measurements), especially for routes deemed marginal using direct-detect transmission.

Figure 5 illustrates how the effects of dispersion and losses along a fiber route result in a reach limitation for 100Gbps direct-detect solutions. In contrast, the coherent solution with its higher tolerance to impairments provides improved performance in the form of additional margin and longer reach capabilities.

Figure 5.  Direct-detect solutions hit a reach limit at a certain distance while the coherent solution will continue to operate beyond this reach limit.

Monitoring, Diagnostics and Troubleshooting.  Built into pluggable coherent transceivers are monitoring and diagnostic capabilities to ensure robust data transmission. In addition, as previously stated, coherent 100Gbps P2P solutions have a very wide receiver dynamic range compared to an equivalent direct-detect P2P link, thus enabling a coherent link to accommodate optical loopback for troubleshooting. In comparison, direct-detect solutions include internal optical amplification at the receiver to close longer links, resulting in direct optical loopback troubleshooting not being possible due to receiver overload.

Reliability.  As previously stated, a pluggable direct-detect 100Gbps solution relies on four transmitter lasers which may require operating at the higher end of their transmitter optical power range, especially for the longer reach links of an edge/access network. For these reaches, active optical amplification at the receiver end may also be required to close the link. Thus, a total of eight active elements must be taken into account when determining the reliability of these types of modules. In comparison, a pluggable coherent 100Gbps solution utilizes only one active optical element, the transmission laser, making it more reliable than the direct-detect solution.

In summary, coherent technology provides:

• Ample optical margin to potentially avoid truck rolls for fiber route characterization.
• High-tolerance plug-and-play implementation over a variety of fiber types and longer fiber links.
• The capability to perform optical loopback if necessary for troubleshooting.
• Monitoring of the transmission performance over these links.

This all leads to operational simplicity and shorter provisioning times, which can result in operational savings in the service provider edge/access network.

Scaling to higher data rates

Pluggable coherent solutions offer the capability to achieve higher data rates in the future over the same, or greater, distances. Compared to today’s access data rates, higher-rate coherent options are already available in small form factor pluggable modules, providing a ready-made path to meeting the demands of service provider edge/access bandwidth growth. Coherent transmission solutions beyond 100Gbps are quite mature, and thus, there is no fundamental near-term impediment in coherent technology for scaling service provider edge and access to higher bandwidths.

Acacia’s Solution

Figure 6.  Integration in silicon enables a path to coherent transceivers for service provider edge/access in a compact QSFP-DD package.

Acacia’s 100Gbps coherent pluggable solutions are specifically designed for optimization in service provider edge and access applications with unamplified links including 80km reaches and beyond. Offered in a quad small form-factor double density (QSFP-DD) that is widely used for client-optics, Acacia’s new 100Gbps coherent P2P service provider edge and access solutions were designed to provide network operators the ability to scale to higher data rates to meet growing bandwidth demands over some of the most challenging optical links, while also providing operational simplicity that may lead to overall network savings.

Acacia’s 3D Siliconization approach, which utilizes high-volume manufacturing processes and benefits from the maturity of Acacia’s silicon photonics technology, is used in the 100Gbps P2P QSFP-DD module. Figure 7 illustrates how advances in optical/electrical component consolidation has resulted in size reductions of coherent modules. The use of silicon photonics to take discrete bulky optical components and integrate their functions into a CMOS-based silicon chip has been a key factor in module footprint reduction.

By following the example of the electronics world and applying integration techniques such as 3D stacking, electronic circuits and a silicon photonic integrated circuit (PIC) can be co-packaged. This approach makes it possible to integrate crucial components into a compact package and reduces the number of electrical

Figure 7.  Evolution of coherent module size over time.

inter-connects while preserving robust signal integrity. Integration involving silicon photonics follows progress already made in the semiconductor fabrication process suitable for volume production and high yields.

Conclusion

Bandwidth demand in P2P service provider edge and access networks is increasing at a rapid rate. This is placing enormous pressure on providers because legacy technology is not capable of supporting data rates beyond a certain speed for a given distance. This has driven a need for 100Gbps optical links that are operationally simple to deploy, along with a path to easily and cost-effectively scale to higher speeds in the future.

Acacia has a proven history of bringing coherent technology to new markets, delivering the low power, performance, operational simplicity, and scalability customers have come to rely on. Since 2011, Acacia has leveraged the benefits of integration to evolve from 100Gbps in a 5”x7” form factor to pluggable form factors such as CFP2 and QSFP-DD supporting up to 400Gbps. This has been enabled by the integration of multiple discrete optical components into a single package using Acacia’s mature and highly integrated silicon single-chip PICs that minimize both size and power consumption; as well as innovations in packaging/integration to achieve high densities. As a trusted partner for service providers and network equipment manufacturers (NEMs), Acacia is now leveraging its technology to develop 100Gbps and beyond small form-factor pluggable transceivers for the service provider edge/access market.

Download the Whitepaper

Standardization and interoperability of optical networking technology provide benefits to both network operators as well as suppliers. History has shown that adoption of optical transceiver/transponder solutions accelerates when implementable standardization and interoperability are established. Evidence of this can be seen in the widespread use of client optics based on optical interfaces defined in IEEE and pluggable form factors defined by Multi-Source Agreements (MSAs). Coherent optical technology, traditionally a closed-system technology with proprietary implementations, is evolving towards a similar path. This migration can help carriers and hyperscale data center network operators by ensuring multiple suppliers for their critical components, which reduces the risk of adopting new technology.

Initially adopted in long-haul applications, coherent optical interfaces have traditionally been deployed in bookended configurations, in which equipment from the same supplier are used on both sides of the link. In this model, vendors were differentiated from one another based on small performance differences that could result in cost savings by reducing the need for costly regeneration nodes.

In recent years, coherent technology has been utilized for shorter distance interfaces to support increasing bandwidth demands on the network edge. Some hyperscale data center networks require large numbers of regional interconnections, which has been driving requirements for coherent technology optimized for cost and power. Several hyperscale network operators have also expressed a need for coherent module solutions in the same form factors that support client optics in the emerging 400 Gigabit per second (Gbps) I/O generation.

Carriers and multiple-system operators (MSOs) have also been driving requirements for coherent optical transport interoperability in adjacent applications, such as access aggregation and metro core networks. These applications may require higher performance, additional data rates, OTN functionality and client multiplexing onto line-side transport channels for optimizing network efficiency.

This white paper looks at several of these industry standardization efforts, as well as recent industry efforts to leverage existing standards and introduces a new 400G interoperability solution called OpenZR+. Already publicly supported by numerous industry leaders, OpenZR+ provides an open, flexible and interoperable coherent solution in a small form factor pluggable module that addresses hyperscale data center applications for higher-performance edge and regional interconnects, as well as potential future carrier applications.

Historical Background

Coherent Interoperability Still Nascent

Historically, optical transport solutions were comprised of an end-to-end solution that provided differentiated technology to achieve high-capacity, long-reach links. Although the data being transmitted followed standardized protocols (e.g., SONET/SDH), the underlying optical layer consisted of proprietary solutions typically resulted in bookended transport links. This approach continued as vendors began introducing coherent transport solutions. Coherent interconnect technology required digital signal processors, utilizing unique framing, proprietary signal processing, and Forward Error Correction (FEC) algorithms, increasing the justification for proprietary solutions.

When coherent technology moved into pluggable form-factor solutions (initially in the CFP followed by the CFP2 form factor), there were early carrier efforts, particularly by AT&T and Deutsche Telekom, to demonstrate interoperable 100G coherent transmission to promote multi-supplier interoperability. Through industry organizations such as OpenROADM MSA as well as the ITU, service providers drove standards which could be supported by multiple vendors, while leaving opportunity for higher performance differentiated proprietary implementations. In particular, a hard-decision FEC and differential encoding were defined in the standard. Hard-decision FEC offered less coding gain than widely used proprietary soft-decision FEC algorithms. Differential encoding simplified interoperability, but introduced a performance penalty compared to non-differential encoding.

Though multiple equipment vendors supported this 100G standard, network operator adoption to date has been limited, primarily due to the fact that higher performance offered by proprietary solutions outweighed the benefit of interoperability. Also, efforts to define a standard took several years to materialize despite 100G solutions being already available. During this time, a large performance gap between the standard and proprietary implementations limited the adoption of the standardized solution.

Lessons Learned: OIF 400ZR for DCI Edge 400G and the Power Consumption Target

Some network operators chose to take a different approach at 400G upon recognizing the lessons learned from the efforts at 100G. In late 2016, these network operators and a few vendors identified 400G as an intersection point for the industry to support coherent optics in the same form factors as emerging high-volume client optics, such as QSFP-DD and OSFP. At that time, it was anticipated that these form factors would support approximately 15W power dissipation. Hyperscale network operators proposed to the Optical Internetworking Forum (OIF) a new project targeting DCI edge applications up to 120km (See Figure 1), with an objective that anticipated module power dissipations not exceed 15W. The resulting OIF implementation agreement was known as 400ZR¹. The OIF hoped that by targeting a very specific application and starting their effort before vendors were well down the development path, the OIF standardization activity could be driven expeditiously.

Figure 1. Target application space of DCI edge for 400ZR.

OIF was quite successful in expediting this effort, with power optimization taking the highest priority in each decision. 16QAM and ~60Gbaud were adopted for the modulation format and baud rate, respectively, after a trade-off comparison of these parameters was performed. In addition, a concatenated (hard-decision + soft-decision) FEC was selected after several FEC proposals were debated. Low-power support and sufficient coding gain for edge applications were driving factors towards this decision. In less than one year, the OIF defined most of the 400ZR interface technical details, which helped motivate increased industry investment in pluggable, interoperable coherent interfaces.

The Carrier/MSO Perspective: OpenROADM & CableLabs

In a separate effort, the OpenROADM MSA group embarked on a coherent standardization effort focused on carrier applications. This effort included carrier-centric features that were not within the scope of 400ZR (e.g., the 400ZR host-board electrical interface scope was limited to 400GbE). Carrier applications needed the flexibility to support multiple rates, multiplexing functionality and additional protocols such as OTN. Additionally, compared to 400ZR, these carrier applications had more stringent performance requirements and were not restricted to client optic form factors. This would permit form factors such as CFP2 to be developed.

Based on different priorities compared to OIF, the OpenROADM MSA chose a high performance soft-decision FEC, called openFEC (a.k.a, oFEC), that was contributed by Acacia Communications. OpenFEC offered higher coding gain when compared to 400ZR, which was comparable to proprietary bookended implementations typical in long-haul applications. OpenROADM MSA also defined a framing that supported 200G, 300G or 400G line rates all utilizing oFEC, with greater host-board interface flexibility.

In the same timeframe, CableLabs, a standardization body representing the cable/MSO industry, developed a coherent interconnect standard to address the cable access aggregation requirements, targeting data rates at 100G and 200G. Although this application had its own unique set of requirements, CableLabs recognized the benefit of aligning with other standards bodies. CableLabs decided to align with the 100G and 200G standards in OpenROADM because it offered the best combination of benefits for this application, including the high-performance oFEC for 200G. The CableLabs standard utilized a common protocol with OpenROADM, while keeping with optical specifications that addressed cable access aggregation requirements.

OpenZR+: Taking the Ambiguity Out of “ZR+”

While the OIF focused on specifications, leading switch equipment manufacturers diligently worked on increasing the QSFP-DD/OSFP power consumption ratings that their next generation equipment could accommodate. As a result, the per-port power consumption tolerance was increased beyond the 15W target set by OIF 400ZR. Thanks to this achievement, support for additional features and higher performance modes that required higher than 15W became possible. “ZR+” became a term loosely used by the industry to specify or capture potential broader modes of operation beyond 400ZR. Unfortunately, due to lack of a uniform definition, the use of the term ZR+ created confusion within the industry.

While the term ZR+ created confusion, the need behind the desire for higher performance beyond 400ZR remained strong. A key requirement was to accommodate hyperscale DCI links beyond 120km, while maintaining the same QSFP-DD/OSFP form-factors. A survey of 400G standardization efforts pointed to elements of OpenROADM that could provide a standard-based high-performance addition to the 400ZR standard. Thus, the industry began looking at the next logical step, which would be to combine these vetted specifications and achieve the goal of addressing 400G Ethernet-centric solutions beyond 120km. This would enable the extension of hyperscale DCIs beyond the edge to regional distances, and expand the addressable market for module suppliers, providing greater economies of scale that benefits the entire distribution chain. This combination of the 400ZR standard with elements of OpenROADM became known as OpenZR+.

Figure 2. OpenZR+ is the logical combination of two industry standardization efforts that enables high performance DCI pluggable modules supporting multi-vendor interoperability.

As illustrated in Figure 2, OpenZR+ is a combination of two industry standardization efforts created to maintain the simple Ethernet-only host interface of 400ZR while adding support for features such as: (1) higher coding gain using oFEC from the OpenROADM standard, which extends the reach capability; (2) multi-rate Ethernet, which enables the multiplexing of 100GbE and 200GbE clients over the line-side link, providing optimization options for the switch/router equipment to channelize the traffic over the transport link; (3) adjustable 100G, 200G, 300G or 400G line-side transport links (using QPSK, 8QAM, or 16QAM modulation), which enables reach/capacity optimization over various fiber links; and (4) higher dispersion tolerance. All of these enhanced capabilities would exist in a QSFP-DD or OSFP module designed to utilize OpenZR+, supporting reaches well beyond the 120km supported by 400ZR. Table 1 summarizes the benefits of modules designed to use OpenZR+, compared to 400ZR and OpenROADM.

*Reach and power consumption are dependent on modulation format and data rate.
Table 1. Next generation pluggable coherent module types

OpenZR+ Applications and Benefits

The higher performance of OpenZR+ can benefit hyperscale network operators by enabling a wider geographical extension of the physical connections to cover not only data center edge interconnects, but also regional interconnects (see Figure 3). Because modules designed using OpenZR+ are housed in QSFP-DD/OSFP form-factors, hyperscale operators can also reduce operational complexity by having one module type to cover 400ZR and OpenZR+ modes.

Figure 3. OpenZR+ enables the expansion of a DCI network’s geographical footprint over regional distances beyond 120km in the same coherent pluggable form-factor as 400ZR.

Hyperscale networks with 75GHz DWDM grid channel spacing line systems can also benefit from modules utilizing OpenZR+. Because 400ZR was initially targeted to operate over 100GHz channels, a 400ZR transmission might suffer from degradation due to the roll-off frequency response of optical filters when deployed with 75GHz channels spacing. The additional gain from oFEC, specified in OpenZR+, can be used to compensate for these degradation effects over the 75GHz channel.

Efforts by carriers to transform or create a data center topology, such as CORD within the Open Networking Foundation (ONF), may also benefit from OpenZR+ utilizing this Ethernet-centric coherent optical solution in a similar fashion to a hyperscale edge/regional interconnect as an overlay to an existing carrier infrastructure.

OpenZR+ can also provide network operators with operational advantages. Because OpenZR+ specifications are derived from existing standards, network operators can benefit from interoperability among multiple suppliers that adopt OpenZR+. Module suppliers can also benefit from OpenZR+ by supporting applications beyond 400ZR, reaching a wider range of customer applications resulting in a larger addressable market.

Conclusion

Module vendors have quickly recognized the benefits of the OpenZR+ mode. For example, Acacia and NEL announced support for the OpenZR+ mode, and have exchanged DSP test vectors to confirm interoperability of their implementations. This is a milestone in the industry as it paves the way for the development of OpenZR+ modules in a QSFP-DD or OSFP form factor that support performance and functionality beyond 400 ZR.

OpenZR+ utilizes a well-defined common sense implementation of functionality that is already well defined in existing standards. OpenZR+ allows network operators to achieve all of the benefits of 400ZR with enhanced features that can expand the reach and functionality without sacrificing interoperability. Full OpenZR+ implementation details are available to vendors interested in supporting this market-expanding ecosystem.

References

¹To find out more about the origins and applications of 400ZR, please read 400ZR: Accessible 400G for Edge DCIs and Beyond

Bandwidth-intensive applications such as video streaming, virtual and augmented reality, artificial intelligence, and enterprise cloud services have been driving the ongoing rise of optical networking bandwidth requirements. In today’s high-capacity optical networks linking together data centers, high-bandwidth client traffic commonly terminates at 100G switch/router interfaces. This client traffic protocol is typically 100 Gigabit Ethernet (GbE). In addition, with global reaches of data center interconnects (DCI), it has become increasingly important to have the ability to terminate these 100G clients not only for intra- and inter-city distances but also for intercontinental distances.

A new generation of switch ASICs supporting 400G I/O is now leading network operators to question how their optical transport networks can support this emerging rate. With standards-based activities around 400GbE and advances in coherent solutions to support multiple 400G client traffic over an optical channel, network operators are looking for ways to support today’s 100G client traffic while also supporting emerging 400G client traffic across all key segments (DCI edge, metro, long-haul, submarine) of their network in an efficient, scalable, and cost-effective manner.

400GbE Anywhere and Everywhere

Figure 1 illustrates an example of how multi-haul coherent solutions, software configurable transponders that provide various transmission capacities and reaches by varying the modulation order and baud rate settings, can provide flexible options for network operators.

A common method of increasing throughput of a coherent channel is to increase the modulation order. However, this may result in a reduction in reach due to reduced optical signal to noise ratio (OSNR) tolerance for the higher modulation orders. This is shown in Figure 1. Alternatively, the baud rate could be increased while maintaining a lower modulation order which provides additional capacity per channel with minimal sacrifice to reach. The goal here is to fill up the available channel spectrum, but as discussed in an earlier whitepaper, increasing baud rate doesn’t increase fiber capacity once the transmission is well-matched to the channel (see Figure 2).

To ensure a smooth migration from 100GbE to 400GbE, it is important to have a solution that can efficiently transport either type of traffic, or a combination of both, without restrictions on performance or functionality. Not only is it important to be able to transport multiple 400GbE client signals, but an effective multi-haul solution should be able to support a wide range of modulation formats that can be optimized to match channel and reach requirements.

Acacia’s AC1200-SC² (SC ‘squared’) coherent transponder module is the first coherent module that delivers up to three 400GbE client signals over a 1.2T single channel transmission utilizing a single-chip 1.2T DSP. Powered by Acacia’s Pico DSP, the AC1200-SC² transponder module was built to enable network operators to support today’s 100GbE traffic with a 400GbE-ready network that can efficiently and cost-effectively accommodate applications from high-capacity DCI edge to terrestrial long-haul and submarine based on QPSK modulation, designed to ensure future scalability.

The AC1200-SC² utilizes Acacia’s 3D shaping technology to optimize fiber capacity and reach by filling gaps in margin and spectrum, while also simplifying network planning with a scalable multi-generational channel plan, enabling network savings. 3D shaping allows fine-tune adjustment of the modulation order and baud rate to provide network operators with the flexibility to customize the transmission to their network requirements and improve capacity utilization.

As shown in Figure 3, the single-chip, single-channel AC1200-SC² supports a range of services from 3x 400G using 64QAM transmission for DCI edge applications to 1x 400G using QPSK transmission for long-haul and submarine applications. This application flexibility facilitates network savings by enabling common hardware to accommodate multiple applications, as well as reducing the need for costly regeneration nodes for long-haul and ultra-long-haul 400G links.

150GHz Windows

As flexible WSS technology has become widely deployed, network operators are considering how to configure the spectral window/channel spacing requirements for next-generation line systems. Even though a WSS has the flexibility to support a wide range of desired channel spacings, it is preferable to utilize spacing based on regular multiples that can be combined without stranding spectrum due to fragmentation. Earlier generations of DWDM networks have deployed 50, 75 and 100GHz channels. As baud rates are increasing, transitioning to wider channels becomes a necessity. 150GHz channels fit nicely with legacy solutions (e.g., 50GHz, 75GHz, 100GHz), as well as next generation optics technology that could be based on baud rates in the range of 120 to 130Gbaud. Acacia’s AC1200-SC² utilizes a single channel that can be optimized for legacy channel spacing up to 150GHz. In fact, in a wide range of channel spacings for 400G transmission, the AC1200-SC² can be configured to utilize the lower modulation orders compared to other solutions.

Conclusion

Network operators are looking beyond transporting current 100G client traffic toward higher speed 400G traffic and seeking solutions that handle the steadily growing bandwidth demands being driven by new and exciting applications and services.

Acacia’s new AC1200-SC² is built for transport systems to address the needs of network operators by enabling this migration to 400G with a flexible, high-performance solution that simplifies network operations. Higher speed 400G networks provide support for accommodating the growing client traffic across the entire network—from DCI edge, metro, long-haul, and all the way to submarine. Future proof your network today with the AC1200-SC².

Introduction

Ever since the invention of single mode fiber optic cable decades ago, the industry has continued to develop new ways of increasing the amount of data that can be transmitted over an optical fiber link. Single-wavelength, on-off-keying (OOK) modulation, in which laser light is turned on and off to represent a digital “1” and “0”, offered significant improvements over electrical transmission, but achieved only a fraction of the capability of the fiber optic cable. Two significant developments have significantly improved fiber utilization: (1) the simultaneous transmission of multiple lasers of different wavelengths over a single fiber — a technique called wavelength division multiplexing (WDM), and (2) coherent transmission using digital signal processors (DSPs) to more efficiently modulate and detect multi-levels in both phase and amplitude of laser light on two polarizations, resulting in increased spectral efficiency. More recently, coherent optical transmission shaping techniques have further advanced the cause, pushing capabilities closer to the maximum theoretical transmission capacity per channel, referred to as the Shannon limit.

This white paper reviews the technological advancements that have played a role in increasing the capacity of information that can be transmitted over a single mode fiber link. It also discusses how parameters in coherent transmission such as modulation order, baud rate, and transmission shaping determine overall fiber capacity.

Historical Perspective

The Gray Period
The scientific and engineering breakthroughs that ushered in the first generation of long-distance fiber optic transmission included (but were not limited to) the ability to reliably manufacture single-mode lasers with a transmission wavelength that coincided with low-loss wavelength windows of SiO₂ glass fiber, and the ability to reliably manufacture optical fiber that could contain single-mode light transmission within the fiber’s core structure.

Over three decades ago, terrestrial single-wavelength single mode optical transmission operated in the low gigabits-per-second (Gbps) range with the transmission capacity defined by this single-wavelength data rate. Optical transceivers with this characteristic are called “gray optics” because they do not require light to be transmitted on a specific wavelength, or color. Distance extension was done by means of optical-to-electrical-to-optical regeneration. Initial deployments, which utilized this method of regeneration, operated at transmission speeds in the range of a few Gbps for terrestrial applications and a few hundred megabits per second (Mbps) for submarine applications.

EDFAs and DWDM
A major breakthrough in the 1980s that accelerated the deployment of optical networks occurred with the invention of the erbium doped fiber amplifier (EDFA). The EDFA enabled wavelengths within the 1550nm low-loss window to be amplified in the optical domain, eliminating the cumbersome and costly need to electrically regenerate the optical signal. Because the EDFA was a broadband amplifier, multiple wavelengths within the window could be amplified with a single amplifier. Around the same time, designs of lasers and wavelength stabilization techniques were maturing to the point that multiple wavelengths could be densely packed within the same 1550nm window. EDFAs and stable lasers brought dense wavelength division multiplexing (DWDM) into mainstream applications. The invention of the EDFA provided a tremendous boost to the adoption of DWDM for optical networks, especially submarine and long-haul networks and later metro networks. Commercial terrestrial DWDM deployments in 1995 were comprised of eight wavelengths at 2.5Gbps, and by 1999, systems with 40 wavelengths at 10Gbps were being deployed¹.

Multiple techniques were incorporated into the links to overcome various dispersion effects due to the interaction of the laser light with the glass fiber. This interaction limited the overall distance achievable in these amplified links. Techniques included incorporating dispersion-shifted fiber or passive dispersion compensators, to name a few. However, since the topic of dispersion is quite expansive and beyond the scope of this paper, we will not discuss it in detail.

The reach effects of fiber non-linearities as well as the electronic capabilities of directly modulated lasers and direct-detect detectors could support practical commercial OOK transmission deployments to 10Gbps per wavelength. Phase modulation was introduced in the 2000s to help commercial deployment push towards 40Gbps per wavelength using differential phase shift keying (DPSK) and differential quadrature phase shift keying (DQPSK). These schemes were able to leverage existing direct detect receiver technology.

DWDM Standardization
To help advance the commercialization of DWDM systems, standards activity played a critical role. The international standards body called the International Telecommunication Union (ITU) released a frequency grid plan for DWDM transmission in 2002 known as Recommendation ITU-T G.694.1. This standard called out a grid with defined spacing of 100GHz, 50GHz, and 25GHz, with granularity of 12.5GHz.

The grid spacing that initially became widely used was the 100GHz grid. This enabled lasers and optical passive filters to be manufactured in volume with ample margin and still adhere to this grid. A standardized DWDM grid enabled network operators to design an optical network topology that allowed for traffic to be dropped and added at various network nodes. These nodes relied on fixed optical add-drop multiplexers (FOADMs) to accomplish these tasks. Changes in network topology required manual changes (“truck rolls”) in order to modify the FOADM fiber connections or the transmission sources. Later, using tunable lasers or optical filters in front of the receiving detectors, changes could be performed remotely via software commands.

Improvements with laser stability designs, along with optical filtering technology (to increase isolation between adjacent DWDM channels), led to deployments using the tighter 50GHz grid. This not only allowed a way to double the amount of DWDM transmissions compared to using the 100GHz grid, but it also enabled a way to upgrade an optical network from a 100GHz to 50GHz grid as bandwidth demand grew by pre-planning the installation of FOADMs and interleavers.

Figure 1. Illustrations of 100GHz channels and 50GHz channels adhering to the ITU DWDM grid.

The Rise of Coherent and Flexible Grid WSS Technology
Let’s focus on the progress of single-wavelength transmission again, before going back to DWDM. Another confluence of technology advancements in the 2010s brought about the generation of mainstream coherent optical transmission. Aided by advancements in CMOS DSP technology (to implement complex detection and error-correction algorithms) which mitigated dispersion effects using signal processing, the capability to transmit long reaches at >100Gbps with modulation schemes much more complex than OOK, DPSK, and DQPSK were achievable.

Early coherent solutions utilized fixed modulation order and fixed baud rate transmission. With advancements in DSP technology, software programmable modulation orders and baud rates became achievable, providing the ability to address multiple applications with common hardware, referred to as multi-haul solutions. Modulating via QPSK (2 bits mapped into a symbol), 8QAM (3 bits mapped into a symbol), 16QAM (4 bits mapped into a symbol), and higher orders provided a means to increase the amount of data that could be transmitted. In addition, the rate at which these symbols were transmitted (aka baud rate) could also be selected. Selectable modulation and baud rate were steps towards achieving a basic level of capacity and spectral optimization over a coherent optical channel.

Early coherent transmission was able to co-exist over legacy 100GHz line systems (with some limitations due to detrimental interactions between an OOK and coherent transmission over a common amplified link). The spectral width of coherent 100G QPSK modulation fit within a 50GHz channel. However, just as increased bandwidth demands drove a need to increase the number of DWDM channels on a fiber link, a continuing increase in bandwidth demand drove coherent modulation requirements to achieve higher capacity and reaches. This meant that 50GHz was no longer wide enough to support various coherent modulation transmission.

As previously mentioned, early DWDM networks relied on FOADMs to add or drop wavelengths at network nodes. The advent of wavelength selective switch (WSS) technology ushered in the era of reconfigurable optical add/drop multiplexers (ROADMs) allowing wavelengths to be dropped or added via software control. Another benefit of WSS technology was that it enabled line systems to create a tunable/flexible DWDM grid rather than a fixed 50GHz or 100GHz grid. This allowed different coherent modulation transmission with different spectral widths to coexist on the same fiber. This capability provided a tremendous level of network flexibility and future proofing.

Figure 2. (a) Example of flexible grid, (b) 75GHz channels.

In response to a need for additional network planning options beyond a fixed-grid system, the ITU revised ITU-T G.694.1 to include a flexible grid framework (Figure 2a). More recently, in 2017, the Optical Internetworking Forum (OIF) proposed a framework to aid in the development of coherent DWDM transmission with flexible characteristics which included the adoption of 37.5GHz, 50GHz, 62.5GHz, and 75GHz channels. Although current unrelated 400ZR OIF efforts to provide 400Gbps data center interconnect standardization using coherent 16QAM was not an application focus at the time, it turns out that 75GHz channels are a good fit for this application (Figure 2b). As 400G transmission becomes more widely deployed in data center and carrier networks, the contribution of 75GHz channels may begin to increase, creating an impact to network planning activities.

Coherent Shaping
Another advancement in coherent transmission was the introduction of transmission shaping solutions, such as Acacia’s 3D Shaping², which addressed a wide range of multi-haul applications and allowed for granular control of the transmission characteristics to improve performance, reach and capacity.

Higher modulation orders (higher bits-per-symbol) provide increased capacity at the expense of reach, while lower modulation orders (lower bits-per-symbol) provide reduced capacity with farther reach. Integer (quantized) bits-per-symbol steps such as QPSK, 8QAM, and 16QAM may result in sub-optimal capacity utilization due to gaps in link margin. Granular modulation techniques such as 3D Shaping enable non-integer bits-per-symbol modes which helps to close these capacity gaps.

Using 3D Shaping, the probability and location of coherent constellation points can be adjusted to optimize for reach and capacity. 3D Shaping also provides the ability to fine tune the transmission spectrum using Adaptive Baud Rate, reducing the stranded spectrum to help increase capacity utilization over a fiber link. This continuously variable method of adjusting baud rate allows for more efficient optimization of a channel by filling up the channel spectrum, since the spectrum of the transmission varies proportionally with baud rate. By adjusting baud rate to achieve the maximum spectrum supported by the channel, network operators can either increase the capacity of the channel for a given reach or achieve greater reaches at a given data rate by operating at a lower modulation order. Modulations with fixed baud-rates (high or low) leave unused channel spectrum, resulting in a waste of fiber capacity.

An understandable misperception is that overall fiber capacity can always be increased by simply increasing baud rates of each transmission of the DWDM system within the fiber. While this might be true when spectral gaps exist between the transmission spectrum and a fixed channel window (e.g. when using 100GHz channel spacing with 32Gbaud or 64Gbaud transmission), it is not the case when the transmission spectrum is tightly packed. In this latter case, increasing transmission baud rate has no impact on total fiber transmission capacity because proportionally fewer total channels can be supported within the fiber.

Conclusion

Benefiting from multiple technology advances described above, the amount of information that can be transmitted across a single optical fiber has increased by more than 10,000 since the 1980s. The latest transport solutions offer exceptional performance and adjustable transmission features that approach the practical limits of fiber capacity. Network operators are looking to address their specific traffic requirements in the most cost-effective manner possible. Addressing these requirements will require further advances in integration, packaging, automation, scale, and flexible architectures.

References

¹P. Winzer, D. Neilson, A., Chraplyvy, “Fiber-optic transmission and networking: the previous 20 and the next 20 years [Invited],” Optics Express, September, 2019, pp. 24190-24239.

Rapidly growing content demand is pressuring carriers, cloud providers, and traditional data center operators to boost the bandwidth of the data center interconnects (DCIs) that link their facilities. A number of factors contribute to the situation. By the year 2022, roughly 82% of all IP traffic is forecasted to be in the form of streaming video.¹ An estimated 4.8 billion global Internet users are forecasted to be accessing content from an Internet comprised of 28.5 billion networked mobile devices and connections. Meanwhile, emerging applications such as the Internet of Things (IoT), 5G wireless, and ultra-high-density (8K) video are poised to intensify demand. Network operators are planning changes to their architectures to keep pace with this growing demand for content.

To address the need for higher DCI bandwidth requirements to meet this growing demand, the optical networking industry began working on a solution known as the 400ZR implementation agreement, with a goal to combine optical line-side fiber capacity with the benefits of client-side solutions.

Spearheaded by the Optical Internetworking Forum (OIF), 400ZR aims to deliver accessible 400 gigabit-per-second (Gbps) Ethernet links for edge DCI applications (see figure 1). The 400ZR implementation agreement addresses edge-DCI applications with link distances targeting 80 km to 120 km and can be implemented in pluggable 400Gbps optical transceiver module form-factors used for client optics.

Figure 1: The primary 400ZR use case is to apply the technology to DCI edge networks.

The 400ZR standard leverages the advantages of coherent optical technology, in which optical laser light is manipulated using a combination of phase and amplitude modulation to transmit multiple bits per symbol (see Coherent Optics Basics sidebar). Optical coherent technology capable of 400Gbps operation is already deployed in carrier networks. The challenge is to transform this technology to meet the requirements of 400ZR and implement the technology in pluggable module form-factors used for client optics.

This market backgrounder reviews the industry drivers behind the development of 400ZR, the key benefits of the technology, and a product-development roadmap for bringing 400ZR transceivers to market. It also provides a few predictions on how the technology could potentially change the industry.

Practical high-bandwidth DCIs

Link to Practical

Figure 2: In today’s hyperscale data networks, east-west traffic (inter- and intra-data center) dominates total traffic.

Data centers began as off-site installations intended to support the primary business facility, whether for backup, disaster recovery, or because of space or power constraints. Today, data centers have become the primary business facilities for enterprises such as Internet content providers, cloud services providers, communications services providers, and more. Although data transmission to the end-user (north-south traffic) remains of primary importance, these new business models require a far greater amount of internal data transmission (east-west traffic). Enterprises want to be able to add additional facilities but manage the collection as a single asset. By linking inter-data center networking equipment over high-speed optical interconnects, data center operators plan to create scalable, resilient virtual hyperscale data centers.

In addition, the most recent generation of Ethernet switching chips, which offer 12.8Tb operation and 400Gbps I/O (input/output) capabilities, should soon see broad deployment in these hyperscale data centers. These I/O ports depend on 400Gbps optical transceiver modules to connect these switches over the data center’s fiber optic infrastructure. This is where 400ZR plays a serious role in the growth of these data centers.

Practical Solutions

Deployment of 12.8Tb switches with 400ZR optical transceivers in hyperscale data centers should address a number of data center pain points such as cost per bit and density. Let’s discuss how 400ZR addresses these pain points.

Fast, Streamlined Architectures
Cost per bit is a key metric for data center operators. With the increased focus on transmission capacity, the challenge becomes how to develop network architectures suitable for the data center with hardware and software solutions capable of delivering bandwidth at an acceptable cost per bit. Data centers primarily use the Ethernet protocol, which provides an economical approach for switching/routing packets of data, referred to as Internet Protocol (IP) packets.

Given the scale of the modern data center, bandwidth alone is not sufficient. Network architectures also need to minimize the installation and management burden. With respect to the optical portion of the network connecting multiple data center sites, there are generally two ways to transport the IP/Ethernet traffic over an optical infrastructure. A traditional method is to create an optical transport dense-wavelength-division-multiplexed (DWDM) network layer and then attach IP/Ethernet switch/router equipment to this network via short optical cables (see figure 3). For ≤100Gbps links short copper cables may also be used. Using the optical transport terminal equipment, data from each switch/router port is “mapped” onto a distinct DWDM wavelength. The IP/Ethernet layer and the DWDM optical layer are treated as separately controlled networks.

Figure 3: Two common architectures for transmitting IP/Ethernet traffic over an optical infrastructure are maintaining a separate DWDM optical transport network and IP/Ethernet transport network (left) or plugging DWDM transceiver modules directly into the switch/router boxes (right).

Another implementation method is to create what is called an IP-over-DWDM (IPoDWDM) network in which optical transceiver modules with DWDM lasers are plugged directly into each IP/Ethernet switch/routers port. An IPoDWDM network has the potential of reducing both cost per bit and operational overhead since less equipment is required and network management can be consolidated.

Some network operators may find it necessary to maintain separate optical and IP networks, especially to support legacy infrastructure. Other operators may focus on the benefits of reducing the amount of equipment they have to manage, especially given scalability concerns.

400ZR optical transceiver modules enable data centers to deploy an IPoDWDM infrastructure by plugging 400ZR modules directly into 400Gbps Ethernet I/O ports on their switch/router platforms.

Familiar form factors
Space is typically a premium in the data center. Operators want to maximize the density, which refers to the number of data I/O ports that can fit within a rack. This translates to maximizing not only the amount of equipment that can fit within a rack but also the number of optical module ports each rack-mounted terminal equipment can support. Data centers have widely adopted the multi-source agreement (MSA) defined quad small form factor pluggable (QSFP) family which provides high faceplate densities. From the 40Gbps QSFP+ variant to the 100Gbps QSFP28 variant, the data rate supported has increased while the mechanical size has remained relatively unchanged. The latest variant, QSFP-DD, is under development to support 400Gbps.

The 400ZR implementation agreement for data center transceivers is intended to support QSFP-DD, as well as a competing form factor called octal small form-factor pluggable (OSFP, see figure 4). Most 12.8Tb switches and routers designs are expected to incorporate slots to accommodate QSFP-DD or OSFP modules. This is to enable 400ZR transceivers to be plugged directly into the switches to support IPoDWDM architectures. The central challenge for transceiver manufacturers is to find a way to implement coherent technology in these very compact packages.

Figure 4: The two form factors under consideration for the 400ZR implementation agreement are the quad small form factor pluggable, double density (QSFP-DD, left) and octal small form factor pluggable (OSFP, right).

Why coherent optical transmission?

Traditional optical transmission uses a form of intensity modulation called non-return to zero (NRZ) where the laser light is turned on and off at the bit rate, in other words, one bit per symbol. This NRZ signal is received using a technique called direct detection. Short distance 400Gbps pluggable transceivers are being developed using a form of intensity modulation with four-level amplitude transmission called PAM4 (4-level pulse amplitude modulation), in conjunction with a direct detect receiver, resulting in two bits per symbol (see Coherent Optics Basics sidebar). These direct detect solutions rely on multiple lasers to achieve throughput as high as 400Gbps. However, they do not have the range required to achieve the distances needed for 400ZR. In contrast, coherent transmission sends multiple bits per symbol using a combination of phase and amplitude modulation, as well as transmitting in two orthogonal polarizations. For example, the 400ZR implementation agreement states the utilization of dual-polarization 16QAM (DP-16QAM) modulation with a baud rate of approximately 60Gbaud. This enables transmission at 400Gbps and beyond using a single laser source, while offering performance benefits that enable improvements in reach compared to direct detection.

Earlier generations of coherent modules tended to be significantly larger than their intensity-modulated counterparts because of the need for powerful digital signal processors and the use of discrete optical components. These solutions targeted long-haul applications where power was less sensitive. However, advances in CMOS, integrated optics, coherent digital signal processor(DSP) designs, as well as DSP coding and equalization algorithms, have significantly reduced the power and size of coherent interfaces for edge DCI applications while maintaining a high level of performance. Today, the implementation complexity between PAM4/direct-detect and DP-16QAM/coherent solutions are similar. Both solutions require integrated optics and custom DSPs, similar RF electronics and advanced packaging technology.

With rising data demand, data center operators are seeking a low power, small form-factor solution to cover edge-DCI distances. Hence, the industry efforts to deliver a viable coherent solution that viable solution is 400ZR.

The importance of interoperability

As previously stated, the 400ZR implementation agreement calls for a pluggable coherent 400Gbps solution based onDP-16QAM signaling with approximately 60Gbaud transmission rates. These devices allow for supporting 80 to 120 km links. A key advantage of standardization is the expectation of interoperability of 400ZR modules from different suppliers.

Historically, coherent optical networking equipment has been built around proprietary (closed) solutions. With 400ZR, coherent solutions more closely resemble MSA client optics with the expectation of interoperability such that a module from vendor A on one end can optically link up to a module from vendor B on the other end. This interoperability further supports an IPoDWDM infrastructure where switches are optically linked to each other over an open line system using 400ZR modules. An open line system built around 400ZR would give data center operators and carriers greater choice in their DCI network architectures and in the components they use to build them.

The roadmap to 400ZR

The expectation of the wide adoption of 400ZR by hyperscale network operators for edge applications is generating significant industry attention in 400ZR. Practical devices need to be designed and fabricated, and then produced in volume. Similar to the shorter reach PAM4 solution, packaging optics into the QSFP-DD/OSFP form factors is challenging. Complying with these mechanical designs while meeting requirements for performance, power consumption and cost requires an innovative focus on three key areas: the DSP, optical/electrical component consolidation, and high-density packaging.

The investment in 400ZR is also anticipated to benefit other applications in the broader optical networking community. For example, the technology developed for the 400ZR small-form-factor pluggable transceivers can also be applied to next-generation CFP2 modules. Despite the march toward smaller form factors, the CFP2 is expected to remain widely used in carrier applications at 400Gbps. The extra space available in the CFP2 package can enable vendors to offer not just coherent transceiver technology, but also introduce additional features and functionality to better serve the carrier space. We refer to these pluggable modules targeting carrier applications as an Open ROADM 400Gbps variant (see 400Gbps Pluggable Variants sidebar). Let’s take a look under the hood of a module to learn the challenges of manufacturing a 400Gbps transceiver in a QSFP-DD/OSFP form factor.

The DSP and Moore’s Law
Complex optical modulation schemes and advanced error-correction algorithms require a high-functioning DSP. Thanks to Moore’s law, which predicts transistor density doubling approximately every two years, a new generation of high-functionality, size-reduced DSPs can be leveraged for 400ZR and other 400Gbps coherent transceiver applications. These DSPs are key to achieving a compact size module with low power consumption.

Optical/Electrical Component Consolidation
Figure 5 illustrates how advances in optical/electrical component consolidation results in size reductions of coherent modules leading to QSFP-DD/OSFP. In addition to the size reduction of DSPs based on Moore’s law, the use of silicon photonics to take discrete bulky optical components and integrate their functions into a CMOS-based silicon chip is largely responsible for the module footprint reduction. Let’s take a look at how silicon not only provides a high level of integration, but also why silicon is an economical solution.

Figure 5: Heterogeneous integration in silicon enables a path to coherent transceivers in a QSFP-DD/OSFP package.

Materials systems
The high-speed optical components for coherent designs are generally based on either indium phosphide (InP) or silicon. InP is a direct bandgap semiconductor material well-suited for use in optical components such as lasers and modulators. Using InP, these functions can be fabricated on the same chip (see figure 6). In contrast, silicon is an indirect-bandgap semi-conductor material, commonly used to fabricate various electronic components including DSPs and other ASICs. Silicon can also be used to create devices with optical functions (aka silicon photonics) such as modulators and coherent receivers. However, silicon does not allow for the straightforward fabrication of lasers. As a result, in a silicon photonics coherent design, the laser is typically separate from the rest of the optical circuit and made from a different material such as InP. This separation can be an advantage in thermally challenging environments because the temperature-sensitive InP laser can be designed to be further away from heat-producing ASICs, while the temperature insensitive silicon photonics can be placed closer to the heat source.

Figure 6: Indium phosphide (left) enables integration of everything but the DSP; and silicon (right), which enables the integration of many electronic components, but not the laser.

Whereas InP has limited application beyond optics, silicon photonics benefits from leveraging common CMOS processes. Transceiver manufacturers can draw on the resources of a mature electronics industry, benefiting from volume manufacturing, lower costs, broad availability of fabs, and ongoing technology improvements.

Packaging Technology
An effective path to small-form-factor pluggable coherent transceivers is to follow the example of the electronics world and apply heterogeneous integration techniques such as 2D and 3D co-packaging. For example, electronic circuits and a silicon PIC can be packaged side-by-side on a common substrate or stacked (see figure 7). The heterogeneous integration approach makes it possible to integrate the PIC and other crucial components such as transimpedance amplifiers (TIAs), analog-to-digital convertors (ADCs), and digital-to-analog convertors (DACs) into a compact package. An advantage of this approach is the reduction of electrical inter-connects while preserving robust signal integrity. Heterogeneous integration techniques are already used in the electronics industry for compute and memory chips. Heterogeneous integration involving silicon photonics follows progress already made in the semiconductor fabrication process suitable for volume production and high yields.

Figure 7: Heterogeneous co-packaging of silicon enables the integration of high-speed optics and electronics.

Meeting the 400ZR challenge
The 400ZR implementation agreement is a customer-focused effort to enable data center operators to meet increasing demand for content. Speed alone is not enough–data center operators are also looking for cost-effective, compact form factors with low power consumption. 400ZR transceivers are viewed by many as the right solution to satisfy these requirements, but requires expertise in low-power DSPs, integrated photonics, and advanced packaging.

Since 2011, Acacia has leveraged the benefits of integration to evolve from 100Gbps in a 5”x7” form factor to 200Gbps in a CFP2 form factor (see Figure 5). These advances have been enabled by (1) development of low-power DSPs that integrate more functionality while reducing power; (2) the integration of multiple discrete optical components into a single package using Acacia’s mature and highly integrated silicon single-chip photonic integrated circuits (PICs) that minimize both size and power consumption; and (3) innovations in packaging/integration to achieve high densities. Acacia is currently leveraging these capabilities and investing in co-packaging technology to develop 400Gbps pluggable transceivers in the OSFP, QSFP-DD, and CFP2 form factors.

Acacia was first to introduce digital coherent optics (DCO) modules in the CFP and CFP2 form factors and we anticipate that our expertise in low-power DSPs and integrated photonics positions us well to develop 400ZR modules that can scale to meet the required volumes.

Acacia is a trusted partner for many carriers, data center operators, and network equipment manufacturers (NEMs) with products already on the market. This positions us well to also become a trusted partner for 400ZR solutions.

To find out more about how Acacia solutions can support your optical networking needs, visit our website or contact us.

References

1. Cisco (November, 2018), “Cisco Visual Networking Index: Forecast and Trends, 2017-2022,” Retrieved from: https://www.cisco.com/c/en/us/solutions/collateral/service-provider/visual-networking-index-vni/white-paper-c11-741490.pdf.

Download the Whitepaper

Bandwidth demand from consumer and enterprise applications continues to grow. Cloud-based applications, data center interconnections, streaming videos, as well as applications on the horizon such as 8k ultra-high-density video, 5G and Internet of Things (IoT) are contributors to bandwidth growth. Global IP traffic, per Cisco’s Visual Networking Index¹ (VNI) indicates that by 2022 there is projected to be 28.5B networked devices and connections, as well as 82 percent of all IP traffic as video. The 2018 Global Interconnection Index² analysis by Equinix predicts private interconnection bandwidth between businesses is forecasted to exceed 8,200 Tbps by the year 2021. Clearly, transmission speeds should stay one step ahead to ensure there is sufficient capacity to meet the demand. Data rates of 100G per wavelength once confined to the long-haul core portions of the network are now extending towards the edges of the network. Today, 600G per wavelength speeds are becoming available for use in the core as well as for data center interconnect (DCI) traffic.

Maximizing return-on-investment (ROI) for a network infrastructure deployment involves ensuring the solution has the ability to optimize capacity and reach, as well as provide low cost-per-bit. Different portions of the network may have different optimization requirements (Figure 1). For example, an owned or leased optical interconnect linking two data centers may require the highest raw capacity achievable for low costper-bit between sites. For a multi-haul environment with disparate channel impairments within metro and/or long-haul links, filling up the available spectral channel with a flexible control mechanism to optimize fiber utilization may be required. For a long-haul or submarine link, achieving high capacity without sacrificing reach may be the main network requirement. Fortunately, the technology to achieve high performance network optimization is available today to meet these various application requirements in the same device.

Figure 1. Achieving network optimization with 600G era coherent technology.

Planning a network to maximize capacity and reach requires a multi-faceted approach. Network operators are embracing coherent optical transmission for long-haul, metro, and DCI links due to various advantages. Coherent technology provides a means to overcome many of the degradation effects an optical signal encounters over a fiber optic link, as well as providing transmission speeds at 100 Gbps and beyond per wavelength.

Recent innovations in digital signal processing (DSP), optical, and mixed-signal component technologies have enabled the achievement of 600G-per-wavelength transmission speeds. High modulation orders (e.g., up to 64QAM) and high baud rates (e.g., up to ~70Gbaud) are now possible. In addition to this high speed achievement, these technologies also introduce advanced capabilities that allow the flexible fine-tuning of the optical transmission resulting in capacity optimization. Today’s coherent technology enables common optical hardware to achieve the high-performance finesse of a long-distance link, the sheer raw capacity for shorter DCI/edge links, and everything in between. This is the reason why some refer to 600G era coherent technology as multi-haul technology since the same set of hardware can address long-haul, metro, and DCI/edge networks.

Figure 2. Acacia’s AC1200 1.2T coherent module.

Acacia Communications recently introduced the AC1200 coherent module, powered by Acacia’s Pico DSP, a low-power solution based on 16 nm CMOS technology incorporating algorithms and processing power to address a wide range of applications. The AC1200 also includes a silicon photonics integrated circuit (PIC), and high-speed RF electronics to achieve 1.2Tbps capacity by using two wavelengths operating at 600Gbps each. The AC1200, is a leading product in the 600G era offering key capabilities that feature high-performance and high-flexibility, with the goal of enabling network operators to improve efficiency and maximize capacity utilization while reducing network costs.

The AC1200 is rich with features that address multiple network applications including DCI/edge, metro, long-haul, and ultra-long-haul/submarine networks. Let’s take a look at each type of network application and elaborate on how the AC1200 provides benefits for each of these applications.

Data Center Interconnects

Raw Capacity with Client Flexibility
In a DCI/edge network application, raw capacity at the highest achievable transmission speeds is a primary requirement for connecting data centers within the same campus or between sites across a metro region. The recent introduction of 600Gbps transmission per DWDM coherent wavelength, capitalizing on recent 64QAM technology advancements, is beneficial for DCI/edge links that require the highest channel capacity achievable at the lowest cost per bit. Integrating transceiver functions of more than one wavelength into the same device/module can improve cost-per-bit compared to a single-wavelength device/module.

The AC1200 uses a dual-core modem design to drive two tunable 600G C-band or L-band wavelengths from one DSP device for a total transmission capacity of 1.2Tb per module up to 64QAM. An interconnection fabric within the module enables flexible support to map client interface traffic onto these two line-side wavelengths. A dual-core modem plus interconnection fabric design is an efficient way to map 400GbE (Gigabit Ethernet) client traffic onto 600G wavelengths (i.e., 3x400GbE onto two 600G wavelengths). In data center environments where Ethernet is the predominant protocol, the AC1200 has the capability to not only support 3x400GbE client interfaces, but can also support 12x100GbE clients, as well as FlexE protocols.

In addition to short point-to-point links, more expansive data center networks may have metro and long-haul interconnects over private or leased connections. For these interconnects, the requirements for the optical transmission links resemble what you may find in a more traditional metro or long haul network with ring or mesh architectures traversing amplified links with reconfigurable optical add/drop multiplexers (ROADMs). Let’s explore how these types of networks benefit from 600G era coherent optical technology.

Multi-Haul Networks

3D Shaping for Transmission Adaptability and Network Optimization
In comparison to DCI/edge links, metro and long-haul networks have a more diverse architecture with a range of fiber conditions and add/drop filters that may require the optical transmission to have multiple “knobs” to optimize capacity, reach, and spectral utilization. These multi-haul networks may require transmission flexibility to optimize multiple disparate links to achieve maximum capacity on a per-link and overall network basis. Multi-haul networks may benefit from shaping of the optical coherent transmission signal to optimize network performance. To provide the means to optimize the optical transmission, Acacia has introduced 3D Shaping technology, powered by the Pico DSP, which enables flexible fine-tuning of the line-side optical transmission allowing network operators to optimize capacity, reach, and spectral utilization. 3D shaping is a power-efficient technology solution that pushes optical transmission capacity closer to the Shannon limit, the theoretical maximum capacity that a communication channel can achieve, enabling the optical transmission to adapt to the line-system network.

Figure 3. Three elements of AC1200
3D Shaping Feature.

Capacity Optimization with 3D Shaping
3D Shaping consists of three elements. The first element is the ability to shape the probability of the constellation points associated with the coherent transmission. A constellation diagram of the coherent transmission is a visual indicator of the efficiency and quality of the optical transmission. Each constellation point in the diagram represents a transmitted symbol. Previous generations of coherent DSPs transmitted each symbol with an equal probability, resulting in a uniform probability distribution of the constellation. Modifying the probability distribution of the constellation can increase the capacity of a channel.

The AC1200 shapes probability of the transmission constellation by using Acacia’s patented Fractional QAM (F-QAM). F-QAM enables one to dial-in a more precise non-integer bit-per-symbol modulation-setting to improve capacity utilization on a channel. This technique that enables non-integer modulation orders—instead of being constrained to integer modulation, such as QPSK (2 bits-per-symbol), 8QAM (3 bits-per-symbol), and 16QAM (4 bits-per-symbol)—allows the link margin to be optimized with greater resolution than prior generation interconnect technology. Figure 4 illustrates how quantized integer modulation order settings may result in sub-optimal capacity utilization due to link margin gaps. F-QAM can be used to fill in capacity caps by “dialing in” a non-integer step modulation setting.

Figure 4. Example of capacity utilization improvement on each wavelength link using F -QAM.

Reach Optimization with 3D Shaping
The second element of 3D Shaping is the ability to shape the location (or position) of the constellation points, which optimizes transmission distance. Tolerance to noise can be represented by how the coherent transmission constellation diagram points are positioned next to each other. The closer the constellation points are to each other, the less tolerant the transmission is to noise. The farther apart the constellation points are to each other, the greater amount of tolerance the transmission has to noise. By shaping (or adjusting) the location of individual points within the constellation, the tolerance to noise can be optimized, which translates to a farther reach without costly regeneration.
Channel Spectral Optimization with 3D Shaping
The introduction of coherent transmission and flexible grid wavelength-selective-switch (WSS) technology into metro and long-haul networks enabled flexible wavelength routing using ROADMs and mixed spectral bandwidth DWDM transmission along the same link. Preserving transmission in the optical domain avoided the required CapEx investment into optical-electrical conversion for transport layer switching and routing. WSS technology enabled the line system network to adapt to the optical transmission with flexible grid channel spacing. The converse was limited in the sense that the optical transmission had limited flexibility, allowing only quantized levels of capacity and spectral usage as previously mentioned.

The third element of 3D Shaping, Adaptive Baud Rate, addresses these limitations by providing an improved level of flexibility with greater granularity compared to legacy fixed/quantized baud rate choices which can create spectral gaps in a channel. Using Adaptive Baud Rate, the AC1200 can adjust the width and shape of the transmission spectrum with flexible control of the baud rate over a wide range, enabling unused spectral gaps to be converted into usable bandwidth. This allows network operators to fully utilize the available channel capacity within a metro or long haul network, as well as provide the ability to reduce regeneration stages and increase network margin.

Figure 5. Adaptive Baud Rate enables the minimization of spectral gaps within a channel passband.

Using Adaptive Baud Rate, a network operator can adjust the transmission spectrum to better fit into the aggregate available passband of the channel by having the ability to continuously tune the baud rate, filling in any margin gaps (Figure 5).

This fine-tune Adaptive Baud Rate capability is very useful in a multi-haul network with multiple ROADM nodes as shown in Figure 6. As previously stated, channel passband widths can vary among links in the same network as well as between networks. Spectral margin gaps to account for worst-case cascaded passband conditions can result in stranded bandwidth. By using Adaptive Baud Rate, these margin gaps can be reduced to improve the spectral utilization of the channel.

Figure 6. Adaptive Baud Rate implementation in a Multi -Haul network for spectral optimization.

With Adaptive Baud Rate, rather than adjusting the line system to match the optical transmission, the ability now exists for the optical transmission to more closely match the line system, giving rise to a new level of network utilization. Flexible grid channel spacing in a network can create bandwidth fragmentation as a result of spectral margin gaps across a network. Bandwidth fragmentation can be minimized using Adaptive Baud Rate, thus minimizing stranded bandwidth.

With Adaptive Baud Rate, rather than adjusting the line system to match the optical transmission, the ability now exists for the optical transmission to more closely match the line system, giving rise to a new level of network utilization. Flexible grid channel spacing in a network can create bandwidth fragmentation as a result of spectral margin gaps across a network. Bandwidth fragmentation can be minimized using Adaptive Baud Rate, thus minimizing stranded bandwidth.

High-Capacity Long Haul and Ultra-Long Haul Links

Figure 7. AC1200 simple block diagram. Shared local oscillator (LO) and Tx laser.

Increasing capacity without sacrificing reach
3D Shaping is enabled by the Pico DSP inside the AC1200 module. Figure 7 shows a generalized AC1200 block diagram. In addition to 3D Shaping, the DSP engine includes other capabilities such as non-linear equalization and advanced Forward Error Correction (FEC). Beyond the DSP engine, there are elements (e.g., high-speed RF/mixed-signal) which contribute to high-baud rate performance. In this section we will discuss how highbaud-rate capabilities become important for optimizing capacity, especially for long and ultra-long haul/submarine links. For example, high-baud-rate capabilities enable a link to achieve high capacity without sacrificing reach by operating at a lower modulation order.

One school of thought is that to achieve higher capacity the modulation order should be increased (e.g. moving from 8QAM to 16QAM). While this is true, it comes at a price of decreased reach due to greater optical signal-to-noise ratio (OSNR) requirements of the higher order. A rule-of-thumb is that reach is reduced by at least a factor of two for every step increase in integer bit/symbol (e.g., going from 8QAM to 16QAM). A key to having the flexibility to optimize capacity versus reach is having the capability of operating at a high baud rate. High baud rate capabilities provide a knob that allows the transmission to operate at a lower modulation order, thus preserving reach, while increasing the channel capacity. Figure 8 illustrates an example of how the reach of a 400G link can be extended by moving to a higher baud rate without moving to a higher QAM modulation order.

Figure 8. 400G example with different QAM/baud rate deployment options.

Suppose a network operator would like to increase the capacity of a link with distance of X to 400G. There are two potential options. Install a system that delivers a high QAM modulation order transmission, such as 32QAM. With this higher order modulation comes a reduced reach capability. An optical-to-electrical-tooptical regeneration site may be required to compensate for the reduced reach capability, resulting in incurred CapEx and OpEx expenditures. If the terminal equipment had higher baud rate capabilities, the reach can be maintained at distance X and still achieve 400G capacity by using 16QAM modulation and increasing the baud rate. This assumes that there is a sufficient spectral window to allow for the baud rate increase. Simply stated, the ability to modulate at these high baud rates brings flexibility of increasing capacity and reach.

The AC1200 coherent module achieves high baud rates of ~70Gbaud based on its design of advanced high-speed mixed-signal analog-to-digital converters (ADCs), digital-to-analog converters (DACs), and Acacia’s high bandwidth silicon photonics. The combination of high-baud rate capabilities and the fine tuning of 3D Shaping means that a network operator can achieve a high-capacity long/ultra-long haul link with high spectral efficiency. This is especially important in a submarine network where multiple providers share the same fiber link, and are trying to optimize its capacity utilization of their assigned spectrum.
Reach Extension Based on Enhanced SD-FEC
FEC performance is one of the main elements in advancing the limits of reach and capacity. The introduction of soft-decision FEC (SD-FEC) compared to hard-decision FEC (HD-FEC) provided a significant improvement in net coding gain. While in the 600G era, further enhancements to SD-FEC algorithms have also brought about additional coding gain, resulting in improved performance and longer reaches for any baud and modulation format. The AC1200 features Enhanced Turbo Product Code SDFEC with ultra-high net coding gain (NCG) to extend reach, while maintaining low power dissipation.
Performance Improvement Example
To illustrate the potential performance improvements of 3D Shaping, high-baud-rate technology, and advanced FEC, consider the transmission of 200Gbps across a coherent optical DWDM channel, based on 50GHz spacing, which includes ROADM nodes. Using Adaptive Baud Rate to increase the baud rate to fill the available spectrum, and by using F-QAM to operate at a lower order non-integer bit-per-symbol modulation, nearly 1.5dB improvement in performance can be achieved. High-resolution ADCs/DACs, high bandwidth silicon photonics, additional techniques to reduce implementation penalties, as well as nonlinear equalization, and FEC enhancements all can provide additional system margin improvement of the same order. The resulting system performance improvement potentially doubles the reach in comparison to previous generation coherent technology.

Flexible Client Support and Link Security

In the above DCI, metro, long haul, ultra-long haul network scenarios, there is a common requirement to support multiple client protocols and ensure link security. To support DCI links in which Ethernet is the prevalent protocol, the AC1200 client interface options include 100GbE, 400GbE, and FlexE. The AC1200 also supports OTU4 for network operators with an OTN transport network.

Link security has become an increasingly important requirement for network operators. Link security is achieved by providing encryption of the data traffic between sites. To augment Layer 2/3 encryption, the AC1200 offers Physical Layer 1 AES-256 wire-speed encryption for each optical channel link.

Intelligent Networking with Adaptive Transmission

Figure 9. Acacia’s coherent optics provide the flexibility to optimize networks for maximum capacity.

For network operators that are looking towards network intelligence for operational benefits, the challenge is whether the optical transmission technology has the capabilities beyond sheer capacity and performance to provide any benefits in an intelligent network. The AC1200 is well suited to be an integral element of an intelligent network. The intelligence of the AC1200 module enables operator ease-of-use by incorporating built-in optimization configurations to avoid manual adjustment of every knob. In conjunction with platform software, network equipment suppliers are empowered to develop configured optimization settings that may operationally benefit the network operator.

Having the capability of fine-tuning the coherent optical transmission characteristics is a good match for intelligent networks using some form of software-defined networking (SDN) and artificial intelligence (AI). With modules such as the AC1200 that provide performance monitoring of numerous advanced parameters, eventually, we may see a network that is able to adjust the optical transmission characteristics on-the-fly using AI to match or predict changing conditions of an open line system.

Conclusion

Table 1 summarizes the multiple network applications addressed by key features of the AC1200. Network operators require network flexibility and optimization to maximize ROI. Achieving capacity optimization and low cost-per-bit rely on advanced optical coherent technology to deliver high performance, high capacity, flexibility, in a compact, cost-effective, power-efficient solution. The AC1200 is a feature-rich coherent module that delivers these attributes, addressing multiple network applications.

Table 1. Multiple network applications addressed by key features of the AC1200 .

References

1. Equinix (2018), “Global Interconnection Index: Measuring the Growth of the Global Digital Economy, Volume 2,”Retrieved from: https://www.equinix.com/global-interconnection-index-gxi-report.

Download the Whitepaper

Acacia Communication, Inc., 3 Mill and Main Place, Suite 400, Maynard, MA 01754, USA 2
OFS, 19 Schoolhouse Road, Somerset, NJ 08873, USA *

Abstract: We demonstrate real-time transmission of 16 Tb/s (80x200Gb/s) over 1020km TeraWave ULL fiber with 170km span length using the world’s first 200Gb/s CFP2 module with a record low power consumption less than 0.1W/Gbps.
© 2018 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
OCIS codes: (060.1660) Coherent communications; (060.2330) Fiber optics communications; (060.4510) Optical communications.

1. Introduction

The rise of social media and over-the-top video along with the adoption of mobile broadband have driven fast traffic growth and shifting traffic patterns in the metro and regional networks. For metro applications, the transmission reaches have to be above 50 km for metro/access and below 1000 km for metro/regional applications. Different from long-haul (>2000 km) transmission systems which heavily focus on the reach and the performance, the metro application calls for compact, scalable, cost-efficient, and easy-to-use digital coherent transport solutions [1,2] as shown in Fig. 1. These requirements are met with pluggable CFP2 transceiver modules (“digital coherent optics”) that include all optics and digital signal processing to support plug-and-play coherent transmission with flexible capacity and spectral efficiency (SE).

In order to achieve the capacity and cost-per-bit targets of metro/regional systems, it is essential to be able to increase the data rates from 100Gb/s to 200Gb/s while maintaining sufficient robustness to narrow optical filtering resulting from cascaded reconfigurable optical add-drop multiplexers (ROADM) on 37.5 and 50 GHz grids. To achieve this goal, 200Gb/s CFP2 should support high modulation formats like 8QAM and 16QAM. However, the lower OSNR sensitivity and increased susceptibility to distortions of higher modulation order pose additional challenges to a CFP2, which is constrained by low power consumption and small footprint requirements. To address this challenge, careful power optimization of DSP algorithm and a good trade-off between DAC/ADC ENOB and performance are necessary. Also, a novel design of RF driver architecture further reduces the power dissipation.

In this paper we report real-time transmission of 200Gb/s CFP2 signals over 1020 km TeraWave ULL fiber using a span length of 170km, much longer than typical metro/regional spans. This length was chosen to stress the OSNR budget while showing achievable amplifier spacing with ultra-low loss fiber. We demonstrate transmission of 80 x 200Gb/s polarization multiplexed (PM) 8QAM and 16QAM channels on a 50 GHz grid with 2.8dB and 1.2dB Q2-factor margin, respectively. We further confirm that there is only 0.1dB Q2-factor penalty for 200Gb/s PM 16QAM signals at 37.5GHz grids compared with 50 GHz channel spacing. We also make a 19 hours’ long-term error free measurement for 200Gb/s PM-16QAM signals at 37.5GHz spacing over 1020km. We report <0.1 W per Gb/s power consumption, which is only 1/8 of an OIF compliant 100G coherent 5”x7” module [3]. To the best of our knowledge this is the first demonstration of real-time 200Gb/s transmission enabled by CFP2 modules and the lowest reported power consumption per bit for coherent transmission.

Fig. 1. Transceiver module landscape (Ref [2].).

2. CFP2 transceiver module

The CFP2 module used in this experiment is implemented in multi-source agreement (MSA) 107.5 mm x 41.5 mm form factor. It has twice the faceplate density of CFP-DCO. It integrates a tuneable narrow-linewidth laser, a single-chip silicon photonic integrated circuit for quad-parallel Mach-Zehnder modulators and polarization diversity 90° optical hybrid, trans-impedance amplifiers (TIA), RF drivers, and a DSP ASIC based on 16-nm CMOS technology. It supports 100Gb/s client data rate using 31.4Gbaud PM-QPSK and 200Gb/s client data rate using either 41.8Gbaud PM-8QAM or 31.4Gbaud PM-16QAM.

The ASIC includes four-channel DAC/ADC, a DSP engine which performs pulse shaping and pre-equalization on the transmitter side, and compensation of chromatic dispersion and polarization mode dispersion as well as clock and carrier recovery on the receiver side. One of the key features of the DSP is a soft-decision forward-error correction based on TurboProduct code designed specifically for ultra-low power consumption applications. With 15% overhead, it has up to 10.8dB net coding gain (NCG) for QPSK and 11.3dB NCG for 16QAM. Low-pow silicon photonic circuits [4] and RF drivers together with the carefully power optimized DSP [5,6] enable a total power consumption of the CFP2 module of less than 20W for 200Gb/s capacity.

Figure 2 shows the comparison of AC200 CFP2 product with other Acacia’s coherent transceivers including AC040, AC100 and AC400 in terms of watts per 100Gb/s. The improvement of CMOS technology from 40nm to 16nm together with the advanced optical/DSP design, which increase the modulation format from BPSK to 16QAM, enable to reduce the power consumption per 100Gb/s from 125W to 10W over 5 years. It is important to note that this roadmap will continue in both directions in the future. First, the date rate will be further increased to 400Gb/s and beyond. Second, the module has even smaller form factor like OSFP or DD-QSFP [7,8] and less power consumption thanks to the next generation of 7nm CMOS technology. The industry expects the 80km-plus 400ZR digital coherent optics module will consume around 15W [9], which results in only 3.75W per 100Gb/s.

Fig. 2. Evolution in power and density per 100Gbps for coherent transceiver modules.

3. Experimental setup

The experimental set-up is shown in Fig. 3. The transmitters consist of one loading and one measurement path. For the measurement path, four 200Gb/s CFP2s (either 16QAM or 8QAM) with a root raised cosine (RRC) spectrum (ROF = 0.2) are set to adjacent channels and then combined together. The loading path is composed of 80 CW DFB lasers on 50GHz spacing from 1530.334 to 1561.826nm, and it is modulated by an I/Q modulator which is driven with 32Gbd 16QAM data waveform by a two-channel 64GS/s DAC using a test sequence generated from an off-line DSP. The polarization multiplexing is performed with passive components. The four 200Gb/s CFP2 measurement channels are combined with the loading path which is notch-filtered by a wavelength selective switch (WSS) prior to the combination. The combined 80 DWDM channels are sent to transmission fiber links with equal launch power. The transmission link consists of six 170km spans using OFS TeraWave ULL fiber which has Aeff of 125 um2 and average attenuation of 0.155dB/km at 1550nm.

The long spans are amplified by hybrid Raman/EDFAs. The average span loss including splices, connectors, WDM couplers and isolators for Raman pumps is 29dB. The average EDFA gain is about 16.3dB. The averaged on-off Raman gain is about 12.7dB which is enabled by three pump lasers at 1429nm, 1447nm, and 1465nm with a total pump power of 850mW.

Fig. 3. Circulating loop setup for transmission experiments.

4. Experimental results and discussion

To test the 200Gb/s transmission performance as a function of OSNR for PM-8QAM on a 50GHz grid (4 bits/s/Hz SE) and PM-16QAM on a 37.5GHz grid (5.33 bits/s/Hz SE), we place the 4 CFP2 channels next to each other in the middle of the C-band. One of the middle CFP2 channels is selected by a 50GHz optical demux in front of the coherent receiver and the real-time BER is measured as a function of received OSNR by varying the EDFA output power and subsequently converted into Q2-factor.

Fig. 4. PM-8QAM and PM-16QAM pre-emphasis measurement at 4 bits/s/Hz and 5.33 bits/s/Hz SE in back-to-back and transmission.

Figure 4 show the PM-8QAM and PM-16QAM performance vs. OSNR at 1545.7nm after 1020 km along with the noise loaded WDM back-to-back performance at 4 bits/s/Hz and 5.33 bits/s/Hz SE. At the error correction threshold (~6.6dB Q2-factor), we achieve a minimum required OSNR of 17.5dB/0.1nm and 20dB/0.1nm respectively. After transmission, the optimal OSNR is achieved at 25.8dB/0.1nm for PM-8QAM and 24.6 dB/0.1nm for PM16QAM. Also, the optimal power per channel is achieved at 1.5dBm for PM-8QAM and 0.3dBm for PM-16QAM. The difference of optimal power per channel is approximately equal to the ratio of symbol rate between PM-8QAM and PM-16QAM. This result indicates the optimal power density is the same for different modulation formats, which is consistent with the Gaussian Noise (GN) model for dispersion uncompensated systems [10]. For the same 200Gb/s net data rate, PM-8QAM at 50GHz channel spacing outperforms PM-16QAM at 37.5 GHz channel spacing by ~2dB Q2-factor at optimal power per channel.

To test the transmission performance as a function of the wavelength, we determine the Q2-factor from the measured real-time BER of the two middle channels among 4 consecutive CFP2 50GHz-spaced channels as they are moved across the C-band. The EDFA operates at 19.2dBm output power, which corresponds to an average power per channel of 0.2dBm launched into the transmission fiber. The average OSNR across 80 channels is 23.7dB/0.1nm. The result of this measurement is shown in Fig. 5. It shows that PM-8QAM and PM-16QAM have an average of 2.8 dB and 1.2dB Q2-factor margin above the FEC threshold (~6.6dB) respectively.

Fig. 5. Measured OSNR and Q2-factor performance across 80 channels at 50GHz channel spacing after 1020 km.

Due to the larger symbol rate, PM-8QAM will be more sensitive than PM-16QAM to filtering and misalignment of the transmit laser and the center frequency of the filters. However, Fig. 6 shows that there is only 0.5dB Q2-factor degradation after 6GHz detuning although the received spectrum after 1020km transmission is asymmetric after passing through a 50GHz optical demuxer in front of the coherent receiver.

Fig. 6. Left: Optical spectrum of PM-8QAM after 50GHz demuxer w/o frequency detuning. Right: Performance penalty with frequency detuning after 1020km.

Figure 7 shows the result of the 19 hours long-term performance measurement of PM16QAM at 37.5GHz channel spacing after 1020km transmission. The measurement samples are collected at 2 second interval. All transmitted data are decoded without error. We see only 0.1dB Q2-factor degradation compared with 50GHz channel spacing. It implies that WDM crosstalk penalty is small. We also confirm that the CFP2 module power consumption is less than 20W.

Fig. 7. Long term measurement of PM-16QAM at 37.5 GHz channel spacing.

5. Conclusions

Relying on pluggable CFP2 transceiver modules, we demonstrate successful 200Gb/s real-time transmission of 80 PM-8QAM and PM-16QAM channels at 4 bits/s/Hz spectral efficiency over 1020 km distance using very long 170km spans of TeraWave ULL fiber showing 2.8dB and 1.2dB Q2-margin, respectively. We make a long-term measurement of PM-16QAM at 5.33 bits/s/Hz spectral efficiency showing the potential for further increase of the system capacity. At less than 0.1W/Gbps, we achieve by far the lowest reported power consumption per bit for coherent optical transceivers.

Acknowledgement
We thank many colleagues at Acacia who were essential in the development of the CFP2.

Download the Article PDF

References and links

1. http://www.tmcnet.com/tmc/whitepapers/documents/whitepapers/2013/9378-bell-labs-metro-network-trafficgrowth-an-architecture.pdf
2. http://www.ecocexhibition.com/sites/default/files/file/Market%20Focus%20Presentations%202012/Acacia%20ECOC%202012.pdf
3. http://www.oiforum.com/public/documents/OIF-MSA-100GLH-EM-01.0.pdf
4. C. R. Doerr, “Photonic Integrated Circuits for High-Speed Communications,” presented at the 2010 Conference on Lasers and Electro-Optics (CLEO) and Quantum Electronics and Laser Conference (QELS), San Jose, California, 16–21 May 2010, CFE3.
5. J. C. Geyer, C. Rasmussen, B. Sha, T. Nielsen, and M. Givehchi, “Power Efficient Coherent Transceivers,” presented at the 42nd European Conference on Optical Communication (ECOC), Dusseldorf, Germany, 18–22 Sept. 2016, P109.
6. C. Rasmussen, “Power and Reach Trade-offs Increasing the Optical Channel Rate Through Higher Baud Rate and Modulation Order,” in Optical Fiber Communication Conference, OSA Technical Digest (online) (Optical Society of America, 2017), paper Th3G.1.
7. http://www.lightwaveonline.com/articles/2016/11/osfp-msa-targets-400-gbps-optical-transceiver-module.html
8. http://www.lightwaveonline.com/articles/2016/03/qsfp-dd-msa-targets-200g-400g-optical-transceivers.html
9. http://www.gazettabyte.com/home/2017/6/22/the-oifs-400zr-coherent-interface-starts-to-take-shape.html
10. P. Poggiolini, “The GN Model of Non-Linear Propagation in Uncompensated Coherent Optical Systems,” J. Lightwave Tech. 30(24), 3857–3879 (2012).