How to Build a Lasting Optical Network
By Eugene Park | Posted on September 2, 2020
Introduction
Bandwidth-intensive data services and applications have been driving continued demand for service providers to architect their networks for growth and flexibility. Today, traffic from large bandwidth users such as hyperscale data centers can reach 400G, thanks to recently introduced high-capacity switches and routers with 400G I/O ports that can accommodate standardized 400ZR coherent optical modules. Service providers are able to leverage the 400G technology derived from these standards efforts for their transport applications. However, even before these standardization efforts were completed, industry advocates were already looking ahead towards higher data rates. Network operators are under pressure to build and deploy multi-generational architectures comprised of terminal equipment and optical line-systems to support current and emerging data rates in an efficient, scalable and cost-effective manner. This blog post examines how recent advancements in optical coherent technology within terminal equipment and a trend towards line-systems with wide channel (bandpass) capabilities can help them achieve that goal.
Coherent Optical Transmission
Today’s coherent optical transport terminal equipment provides channel capacities ranging from 100G to 1.2T. Currently available coherent digital signal processors (DSPs) enable network operators to optimize capacity and reach, while software controlled modulation and baud rate enabled by these DSPs support multi-haul applications whereby the same hardware can be used in multiple parts of a network (data center interconnect edge, metro, long-haul, and submarine). This provides a cost-effective way to leverage common hardware for multiple applications. For example, Acacia’s Pico DSP, which is the “brain” inside Acacia’s AC1200 module, can provide up to 1.2T of transmission capacity in a single channel, enabling large bandwidth pipes for applications such as data center interconnect (DCI) edge. The same DSP/module can also be used at a lower modulation order for ultra-long distance submarine links at lower data rates.
Figure 1. Acacia’s 3D Shaping optimizes transmission per-channel capacity, reach, and spectrum utilization.
These DSPs also feature transmission capabilities to enhance performance and flexibility, such as Acacia’s 3D Shaping which includes shaping of constellation points as well as modulation and Adaptive Baud Rate fine-tuning to optimize capacity, reach, and spectral utilization for any line-system channel. Innovations, such as 3D Shaping (Figure 1), have helped contribute to narrowing the gap to the Shannon Limit, the theoretical maximum capacity for a given channel.
Minimizing Stranded Bandwidth by Adjusting the Transmission
The Adaptive Baud Rate feature in 3D Shaping allows for the transmission spectrum to more closely match the channel passband to maximize channel spectrum utilization, which avoids stranded bandwidth.
It is very important to minimize stranded bandwidth because when the optical transmission is not matched to the line system that carries this transmission, unused bandwidth essentially goes to waste. Adaptive Baud Rate provides the fine-tuning capability to continuously adjust the transmission baud rate, resulting in its spectrum filling up previously unused channel bandwidth. In addition, for fixed grid networks, Adaptive Baud Rate is important for maximizing spectral utilization when the optical path encounters cascaded passband filters.
Optical Line Systems and Wide Channels
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, given the various transmission options available. The introduction of reconfigurable optical add-drop multiplexers (ROADMs) provided line-systems with the ability to route individual and groups of wavelengths through optical fiber infrastructure. These ROADMs also provided the ability to adjust channel bandwidth via software control. Even though today’s 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 fragmentation. Earlier generations of dense wavelength-division multiplexing DWDM networks have deployed 50, 75 and 100GHz channels.
Avoiding Stranded Bandwidth by Planning Ahead
While Adaptive Baud Rate provides a method to minimize stranded bandwidth by adjusting the transmission characteristics as previously mentioned, there are additional steps that can be taken to reduce stranded bandwidth by proper planning of the line system channel grid for future networks. As the gap to Shannon’s Limit decreases, a trend seen in current coherent technology advancements indicates a shift in focus from achievable modulation orders to achievable aggregate baud rates. The line system channel plan should evolve to follow this trend.
Today’s high performance DSPs are capable of a modulation order of 64QAM, such as Acacia’s Pico DSP. High levels of industry investments have recently been made in 400G 16QAM ~64Gbaud pluggable coherent solutions to address the growing markets of 400G DCI edge and metro optical interconnects. This has resulted in a trend towards leveraging these investments in 400G pluggable components and technology, while increasing aggregate baud rates for increased channel utilization and network optimization.
To accommodate future transmission schemes while minimizing stranded bandwidth, a logical step is to ensure wider channel plans are integer multiples of recently established plans, such as two times 75GHz as illustrated in Figure 3.
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 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.
Benefits of a 150GHz Channel Plan
As previously mentioned with regards to the Shannon Limit gap decreasing, industry investments are shifting towards increasing the transmission aggregate baud rate to increase capacity while maintaining usable reaches. For example, leveraging the investments in 400G 16QAM solutions, by dialing down the modulation order from 16QAM to QPSK and doubling the aggregate 64Gbaud rate to 128Gbaud and beyond, these solutions can expand the 400G capacity over a much greater distance. A 150GHz channel plan can accommodate these higher aggregate baud rates. Although intermediate steps of modulation order reduction and baud rate increase are possible to increase capacity and reach, Figures 4 and 5 illustrate how intermediate steps can result in stranded bandwidth if a 150GHz channel plan is not used.
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 fewer channels required to maximize fiber capacity. This can reduce operational cost/complexity and result in a reduction in required ROADM ports.
To ensure a meaningful return on investment (ROI), it is desirable to deploy a line system with not only flexible channel widths, but also with wide channel capability in order to drive cost-efficiency and accommodate future higher capacity client rates and wider transmission spectrum.
Cost/Power/Size Benefits of Opto-electronic Integration
Coherent optical modules have enjoyed a healthy year-over-year reduction in per-bit area, power, and cost as shown in Figure 6. This has been achieved in-part through advancements in DSPs, silicon photonics and opto-electronic integration. Silicon photonics technology in particular has enabled high levels of integration, low power consumption, and high-volume manufacturing capabilities.
These reductions in size, power, and cost have enabled coherent applications to migrate towards shorter reach access and edge portions of the network. An example are the recently available 400ZR pluggable solutions, which are housed in QSFP-DD and OSFP form-factor modules. The benefit of these modules is that they can be plugged directly into switches and routers, offering the same density for both coherent DWDM and client optics in the same chassis. In addition, 400G CFP2-DCO form factor modules have also been introduced with carrier-centric, high-performance features utilizing similar technology. Recently, Acacia Communications introduced 400G coherent pluggable modules in all three of the aforementioned form factors.
These modules leverage Acacia’s 3D Siliconization approach, which takes advantage of mature silicon photonics technology and vertical integration, utilizes high-volume electronics manufacturing processes, and is designed to support a wide range of market applications. With the ability to address multiple markets such as access, hyperscale, metro, and long-haul, these 400G integrated, high-volume solutions enable the alignment of component investments across a range of these applications.
Building a Lasting Network
As this blog post highlights, service providers can stay ahead of rising bandwidth demands by building cost-effective, flexible and scalable networks using coherent optical solutions with advanced transmission shaping and wide-channel capable line systems. This maximizes channel utilization and prepares for future increased capacity requirements, while achieving a desirable multi-generational ROI. Continuing the trend of reducing size, power, and cost in coherent transmission technology is paramount to enabling solutions that support scaling-up service provider network capacities. Achieving these objectives will require a balanced long-term approach to technology roadmaps that leverages broader industry investments such as silicon photonics, opto-electronic integration, and volume manufacturing processes.