Juniper Networks QFX10000 Modular Ethernet Switches Overview

by http://www.fiber-mart.com

The ultra-high density Juniper QFX10000 Modular switches provide the ultimate support and solid ground for today’s most demanding network operations and applications. Their scalability options and their stable performance make them optimal for deployment in medium to large sized Datacenters as well as in private and public clouds. With the custom built ASICs the Juniper QFX10000 switch can deliver from 3 to 96 Tbps of throughput in your network thus becoming a safe and long-term investment. The leading network architects in Juniper are also looking to increase this capability to even 200 Tbps in the near future. With the option to use up to 480 100GB/s ports in a single chassis the Juniper QFX 10000 is the industry leading switch in its class. This eventually enables you to evolve your network infrastructure and boost the performance by upgrading to 100 GB/s and leaving the 40 GB/s and 10 GB/s in the past. This would make the clients extremely happy and will motivate them for even greater creativity. Some of the key features of this switch series are listed below:
The high port density capability of QFX 10000 Modular switches redefine the per-slot economics, enabling customers to do more with less and at the same time simplifying network design and reducing OpEx (operational expenditure)
The custom, built by Juniper, ASICs in each QFX10000 switch delivers unmatched and unparalleled intelligence and analytics
Deep buffers support ensures even greater quality of service options
With the large number of 100GB/s ports this switch series is the optimal solution for future upgrades
The Juniper QFX10000 Modular Switch series gives customers the ultimate architecture flexibility. With both Layer 2 and Layer 3 support the customers can deploy this switch in every part of their network infrastructure. For networks evolving to become software-defined (SDN), the QFX10000 can integrate with VMware NSX SDN controllers and can act as a Virtual Extensible LAN, both Layer 2 and Layer 3,  gateway. You can eventually select out of two available modular chassis:
The QFX10008 Ethernet Switch with an 8-slot, 13 U chassis that supports up to eight line cards
The QFX10016 Ethernet Switch with a 16-slot, 21 U chassis that supports up to 16 line cards
Optionally you can choose some of the following line cards to boost the chassis’ performance:
QFX10000-36Q which can be a 36-port 40GbE quad small form-factor pluggable plus transceiver (QSFP+) or a 12-port 100GbE QSFP28 line card
QFX10000-30C which is a 30-port 100GbE QSFP28/40GbE QSFP+ line card
QFX10000-60S-6Q which is a 60-port 1GbE/10GbE SFP/SFP+ line card with six-port 40GbE QSFP+ or two-port 100GbE QSFP28
Whatever road you take you can’t go wrong with the QFX10000. When it’s fully equipped and configured a single QFX10016 chassis can support up to 480 100GbE ports making it an industry leader.
Key components of the QFX10000
Both versions of the QFX10000 modular switches share some common architectural elements. Both versions run the JunOS operating system which is tasked to handle the Layer 2 and Layer 3 protocols while the Switch Fabric modules have the task to manage the chassis and provide switching functionality for data traffic coming from line cards. The line cards mentioned above include Packet Forwarding Engines (PFEs) and independent line card processors. Due to the Virtual Output Queue (VOQ)-based architecture, the QFX10000 can scale up to a very large deployments with no head of-line blocking, a single-tier low-latency switch fabric and deep buffering to ensure high performance throughout the whole network. One neat feature is the option for easy future upgrade thanks to the direct connection between the horizontal line cards in the front of the chassis and the vertical switch fabric cards in the rear of the chassis. They are connected with each other in a so called orthogonal interconnects which eliminates the need for a midplane. This makes it extremely popular with customers looking to invest in 100 GB/s connections and even 400 GB/s in the near future. The line cards provide an uninterrupted support of cold air due to the redundant and variable-speed fan trays. The power supplies have the capability to convert the external power to the internal voltage needed for safe and stable operation. Each and every component of the QFX10000 switches is hot swappable and interchangeable.
QFX10000 Line Cards key features
Every set of line cards supported by the QFX10000 provides an extensive set of features to compliment your network. They can be deployed in any combination of Layer 2 and Layer 3 network. They have a unique ability for seamless transition between the supported speeds due to the support for tri-speed 10GbE, 40GbE and 100GbE connections. Each line card is built by Juniper Networks making it a recommended component for your QFX Modular switch. The line cards support many technologies including 802.1Q VLAN and VXLAN, link aggregation, VRRP, L2 to L3 mapping. In addition the line cards support filtering, sampling, load balancing, rate limiting, CoS, MPLS, Fibre Channel over Ethernet (FCoE) transit functionality and other key features needed to deploy a high-performance and stable Ethernet infrastructure.
For redundancy in deployments where the power is not stable, the QFX10008 contains six power supply bays while the QFX10016 has ten power supply bays. This offers the much needed flexibility and redundancy. Each power supply has its own internal fan for cooling. In addition all QFX10000 chassis provide both AC and DC power supplies however, AC and DC supplies cannot be mixed in the same chassis. The QFX10000 chassis has front-to-back cooling with hot air being exhausted through the fan trays placed in front of the line cards.
Deploying the QFX10000 switch will benefit and boost your network’s performance. Among the many features that this switch provides, the below few are extremely important to be mentioned:
Each QFX10000 chassis comes with a very important redundant feature: an extra slot to accommodate a redundant RE module that serves as a backup in hot-standby mode. This module is ready to take over in the event of a master RE failure. The taking over of the backup module will be seamless thanks to the integrated Layer 2 and Layer 3 graceful Routing Engine switchover feature implemented in JunOS, while working in conjunction with the nonstop active routing and nonstop bridging features
The support for a virtual output queue for deployment in large size Datacenters. This feature allows for packets to be queued and dropped on ingress traffic during congestion with no head-of-line blocking
The QFX10000 provides many MPLS features to suit your needs. Among many the most important are L3 VPN, IPv6 provider edge router (6PE, 6VPE), RSVP traffic engineering and LDP for segmentation and virtualization
The QFX offers you the support for Fibre Channel over Ethernet (FCoE) together with priority-based flow control (PFC) and Data Center Bridging Capability Exchange (DCBX). All of these features are included as part of the default software which comes with the switch
The QFX10000 Switch Series offers industry leading scale options and high performance with a design capable to seamlessly upgrade your Datacenter to 100 GB/s operating Datacenter. The QFX10000 Series Switches have been designed and destined for the future deployment of 400 GB/s Ethernet solutions. With the deployment of this switch series you will help your cloud and Datacenter operators extract maximum value and intelligence from their network infrastructure.

Advantages and Disadvantages of OM5 Fiber in Data Center

by http://www.fiber-mart.com

Data centers are the main storage houses to store and distribute data on the internet.  With the growing demands like high bandwidth videos, it is hard to say how much data is being uploaded and downloaded every second on the internet. To cope up with the demand of great bandwidth and high- speed applications, data centers of modern times needs to be upgraded. One such up gradation is OM5 fiber cable also known as wideband multimode fiber(WBMMF). In this post we will see both sides of the same coin, means adoption of OM5 fiber in data centers is beneficial or not.
Revolution in Data Center Deployment
With the data centers shifting to support more digital strategies, we have to gear up to tackle new challenges, more users and an ever-evolving digital world. Data rates have grown from 10G, 40G and now beyond 100G in many data center networks. According to Cisco Forecast overview 94 percent of workloads and compute instances will be processed by data centers by 2021. This clearly shows higher bandwidth and greater access are major elements for a data center’s architecture.
What is OM5 Fiber and will it benefit a Data Center?
OM5 is a big innovation in fiber optics and is a new generation of multimode fiber. It is designed to complement Short Wave Division Multiplexing (SWDM) to reduce the parallel fiber count to allow continued use of just two fibers that transmit 40G and 100G. OM5 fiber jumper is defined as wideband multimode fiber (WBMMF) as per new standards in IEC and TIA. It can work over a wide range of wavelength which varies between 850 nm and 950 nm. Let’s have a look
Pros and Cons of OM5 in a Data Center
GBIC-SHOP summarizes the pros and cons of OM5 cable and it cannot be denied that OM5 is to meet the high bandwidth challenges.
Advantages:
Compatibility – OM5 cable has the same fiber size and shape as of its predecessors OM4 and OM5. Therefore one doesn’t have to root up the existing cabling setup and already existing ports can be used shifting it to a high-speed data center network. We can say OM5 is fully compatible with the existing cabling infrastructure of data centers.
Extendibility – OM5 fiber jumpers are considered as an advanced version because of their ability to combine 100G-SWDM4 and parallel transmission technology.  OM4 patch cord can support link length only up to 100m. OM5 can achieve 200/400G Ethernet applications by using this 8-core WBMMF standard patch cable.
Cost – SWDM technology has reduced the number of fiber, the overall cost of achieving high bandwidth with OM5 is considerably reduced. OM5 fiber is beneficial for data center deployments when it comes to cost parameter. In most of the data centers, short reach connection is common, so multimode fiber cable(MMF) is the most cost-effective.
Disadvantages:
As we know that OM5 fiber cable has been standardized last year, its price is higher than OM4. Moreover, its corresponding transceiver 100G-SWDM4 has limited production. However, OM5 will surely be popular in coming times and increase in demand will bring a reduction in price.

Passive and Active Direct Attach Cables – What is the Difference?

by http://www.fiber-mart.com

Introduction of Direct Attach Cables
Direct attach cables (DAC) are an alternative to the fiber optic transceivers, they eliminate the need of using the transceivers by permanently attaching both ends of the cable with transceivers that can be terminated in the SFP+ slot in the communication equipment such as switches, routers, storage and servers. Figure 1 shows a typical direct attach cable.
Direct attach cables are used for smaller distance links, normally direct attach cables are available in lengths of 1 meter, 3 meters, 5 meters, 7 meters, 10 meters and 15 meters. General uses of direct attach cables are connecting the equipment in the same rack, connecting the equipment installed in adjacent racks or connecting the equipment within a mid-sized datacenter. Direct attach cables use both copper and fiber cable assemblies. The decision to use copper or fiber cable is dependent on various factors such as electro-magnetic interference and space availability.
Types of Direct Attach Cables
There are two main types of direct attach cables:
Copper/Twinax Direct Attach Cable
Fiber Optic Direct Attach Cable
These are further classified as Passive Copper/Twinax Direct Attach Cable and Active Copper/Twinax Direct Attach Cable. Fiber optic direct attach cable is available as Active Fiber Optic Direct Attach Cable only. In the next section, we will compare the active and passive types of direct attach cables and look at their specific uses.
Passive Direct Attach Cables
Passive direct attach cables are copper cables with fixed transceivers at both ends, these cables terminate on the communication equipment and provide connectivity between devices. Passive direct attach cables are usually available in lengths of up to 5-7 meters. Passive direct attach cables are thicker and consume more space. They are difficult to manage if a large number of cables are terminating on a single equipment. Passive direct attach cables have length limitations because copper cable cannot be used for longer distance 10G connectivity.
Active Direct Attach Cables
Active direct attach cables can be constructed of either copper or fiber with transceivers fixed at both ends. Active direct attach cables are available in lengths of up to 15 meters. Active direct attach cables have lesser thickness than passive direct attach cables primarily due to lesser thickness of the fiber optic as compared to the copper cable used in passive direct attach cables.
The primary difference between active and passive direct attach cables is the additional components of active transmitter and active receiver present in the active direct attach cable. On the contrary, passive direct attach cables do not have any active component in them rather they rely on the signals provided to them by the communication equipment.
Conclusion:
The choice of using passive direct attach cable or active direct attach cable is purely circumstantial. Below are some pros and cons of using either type of cable.
Passive Direct Attach Cables
Pros:
Cost effective
Flexible to bend
Cons:
Thick – difficult to manage/harness
Shorter length
Electro-magnetic interference can cause packet loss and other issues
Active Direct Attach Cables
Pros:
Cost effective
Longer lengths than passive direct attach cables
No electro-magnetic interference in active optical direct attach cables
Thinner – easy to manage and consume less rack space
Cons:
Fiber cable cannot bend beyond a certain limit
Higher chance of failure than passive cable due to presence of active component

What is InfiniBand and what is it used for?

by http://www.fiber-mart.com

What is InfiniBand and what is it used for?
INTRODUCTION :
InfiniBAnd (IB) is actually a trademark term used since 1999, it was formerly called System I/O. InfiniBand was coined surprisingly when two dueling designs in the market merged, this happened after realizing that it is the right approach to prevent future limitations in the industry, because the existing designs would no longer meet the needs of future servers.
The two competing designs were:  Future I/O – developed by IBM, Compaq and Hewlett-Packard, and Next Generation I/O – developed by Microsoft, Intel and Sun Microsystems.  With confidence that both the industry and the end-users will benefit from the merging, they formed the InfiniBand Trade Association or IBTA which has over 220 members, currently.
Future I/O and Next Generation I/O are Input/output architectures were expected to replace the traditional PCI or Peripheral Component Interconnect system. Why is there a need to replace the PCI bus?  Mainly because, the PCI bus became the bottleneck limiting the performance of high-speed data servers for the reason that, it is restricted to about 500 Mbps of shared data only. PCI dominated the industry since early 90’s with one major upgrade during the period: from 32 bit/ 33 MHz to 64bit/66Mhz. The PCI-X which made the technology one step advanced to 133 MHz was projected to pro-long PCI architecture usage in the industry. However, Internet became so popular globally, continually increasing in demand to almost no downtime at all. The need to constantly accessible, dependable high-performance with fail-safe system which are services provided by the web, data storage features of internet, applications, database servers and enterprise computing software systems, etc. has changed the game plan of the market players.  Moreover, many opt to move storage out of the server to isolated storage networks and distribute data across fault tolerant storage systems is now a trend in the industry.  Such demands require more bandwidth and that bus system have reached the level that PCI interconnect architecture can no longer cater.
So IBTA, came up with the so-called, InfiniBand. What is Infiniband?
InfiniBand is a switch-based point-to-point serial I/O interconnect architecture developed for today’s system with the ability to scale next generation system requirements.  It is operating based on a four-wire 2.5 Gb/s or 10 GB/s base speed per individual port link in each direction.  It is a low pin count serial architecture connecting devices on the PCB as a component-to-component interconnect and enabling “Bandwidth Out of the Box”, chassis-to-chassis interconnect, traversing distances up to 17m over common twisted pair copper wires.  Compared to ordinary fiber cable, it can go over distances of a number of kilometers or more. Its architecture described a layered hardware protocol; Physical, Link, Network, Transport Layers and a software layer to manage initialization and communication between devices.
Different USES of  InfiniBAnd
RAS(Reliability, Availability, Serviceability) provider
InfiniBand provides RAS (Reliability, Availability, Serviceability) capabilities designed into the InfiniBand. RAS refers to a fabric that works both in-the-box and allows Bandwidth Out of the Box.  Because of this RAS feature, it is projected that InfiniBand architecture will be able to serve as the common I/O infrastructure for the next generation of computer server and storage systems at the heart of the Internet. Hence, this will fundamentally alter the systems and interconnects the Internet infrastructure.
Supports Application Service Providers or ASP
The Internet, from simple online data search engine to supporting numerous applications, creating international market for media streaming, business to business solutions, E- commerce and interactive portal sites.  The demand for reliability of each application created tremendous pressure to service providers. Application Service Providers or ASP entered in, a group offering quality services with the capacity to intensely gauge in a short period of time to accommodate drastic growth of internet despite possible congestion using the cluster to support above requirements.  A cluster is group of servers connected by load balancing switches working in a parallel to serve a particular application.  InfiniBand makes application cluster connections simplified by interconnecting or fusing network with a feature-rich managed architecture. It delivers native cluster connectivity, devices can be attached and multiple paths can be used with the addition of switches to the fabric.
QoS or Quality of Service
 InfiniBand at the same can deliver and process transactions of high importance between devices prioritizing over the less significant items through built-in QoS or Quality of Service mechanisms.
Scalability for IPC or Inter-Processor Communications
The switched nature of InfiniBAnd offers connection reliability for IPC or Inter-Processor Communications systems by allowing multiple paths between systems. Scalability is sustained with fully convertible connections managed by a single unit which is the subnet manager. With multi-cast support feature, single transactions can be made to multiple destinations. Consequently, InfiniBand served as a backbone in the capabilities for IPC clusters by allowing multiple servers work together on a single application without the need of secondary I/O interconnect because of the higher bandwidth connections (4X/12X) it can provide.
Storage Area Networks (SAN) simplified
These are groups of complex storage systems linked together to managed switches to allow vast volumes of data to be stored from multiple servers. They provide dependable connections to large database of information that the Internet Data Center requires. Basically, SAN are built using Fibre Channel switches, servers and hubs attached through Fibre Channel Host Bus Adapters (HBA). But the emergence of InfiniBAnd results from removal of Fiber Channel Network and lets servers connect directly to a storage area network eliminating the pricy HBA. With features like the Remote DMA(RDMA) support, simultaneous peer-to-peer communication and end-to-end flow control, InfiniBand through its fabric topology overcomes the deficiencies of Fibre Channel; such as, restriction of data that individual servers can access which arises to “partitioning mechanism”, or sometimes termed as zoning or fencing without the aid of a costly and complex HBA.

CAN THE 10G SFP+ RJ COPPER TRANSCEIVER BE A GAME CHANGER IN 10GBASE-T?

by http://www.fiber-mart.com

Right from the unveiling of the, then, new IEEE standard for 10 Gigabit Ethernet (10GbE), mostly known as IEEE Standard 802.3ae, large corporations started preparing their network infrastructures for the much needed performance boost and almost immediately they started deploying the new standard in their backbones, Datacenters, and server farms with a single and very important goal at their mind, to evolve their network and make it capable of supporting the growing number and demand of business and mission-critical applications. Today we can safely say that the 10GbE standard evolved in a main competitor when it comes to achieving a reliable, affordable, and simple network architecture.
Even though the 10GbE standard is significantly cheaper to deploy today than when it first came up, many leading corporations are still trying to find a way to reduce costs while gaining performance. Mainly they focus their attention on the copper part of the 10GbE and rely on the proven characteristics of copper transceivers in the past couple of years.
When it comes to transceiver and cabling options, 10GbE has you covered in every single aspect of your network. It can work with either copper of fiber solutions and it offers a wide range of distances for your convenience. With latest trends in the networking world and the noticeable improving switching technologies, copper 10GbE solutions are gaining speed and popularity. Currently the most important 10GbE copper technologies are shown in the table below:
The 10GBase-CX4 is the first ever 10GbE copper standard introduced in 2004. Even though it offered low latency for a very low cost, the main disadvantage was its unusually large form factor which was causing high density configurations to be almost impossible.
The CX4 standard has been replaced with the latest SFP+ standard. This standard offers the same latency characteristics, over longer distances. Together with the small form factor these characteristics make it one of the favorite transceivers used in today’s demanding networks.
The 10GBase SFP+ copper transceiver has been developed for greatness. It offers a high performance bi-directional communication over the cheaper and widely deployed standard copper cables.  In order to achieve the maximum performance, the use of Cat 6a or Cat 7 copper cable is a must. One of the crucial points in its advantage is the low power consumption. When properly deployed and maintained the SFP+ copper transceiver can save 0.5W per port compared to an embedded 10GBASE-T RJ45 port. This is especially noticeable with distances up to 30 meters. In addition with the base of its technology being the copper, you can worry no more for any performance loss if the cable is not positioned straight.
When planning your network infrastructure it is important to make sure that the physical infrastructure will support future application needs, and future technology developments. This is proving to be the main challenge of 10GbE copper transceivers even though they use the traditional RJ45 connector which is the most widely used and known connector in the world. However new dynamics in Datacenters and Service Providers mandate that the cable infrastructure handles latency sensitive applications anywhere in the network architecture.  This leaves the impression that when comparing 10GBase-T technology with the alternative SFP+ technology, it is evident that SFP+ is the right technology to choose to ensure optimal performance with lowest latency in the Datacenter and will for sure become the leading transceiver to use when deploying a high performance network architecture.

THE 4 ADVANTAGES OF CWDM TECHNOLOGY NETWORKS

by http://www.fiber-mart.com

Coarse Wavelength Division Multiplexing s one of the optical transport technologies that make use of the light wavelengths and fiber high band capacities along with SDH and DWDM technologies. CWDM is not the latest technology developed for optical transmission but it has its own advantages for choice in particular circumstances.
1.Simpler implementation and operation versus the DWDM implementation.
Simpler refers in this case to simpler optical hardware components necessary to implement the transmission system. Wavelengths spacing is much wider than in classical DWDM systems. Spacing is usually 20nm between lamdas instead of 50GHZ and 100GHZfrom DWDM. CWDM systems are using 8 or 16 or 32 lamdas versus 96 channels in DWDM systems. In 2002 the ITU standardized a channel spacing grid for use with CWDM (ITU-T G.694.2), using the wavelengths from 1270 nm through 1610 nm with a channel spacing of 20 nm. (G.694.2 was revised in 2003 to shift the actual channel centers by 1 nm, so that strictly speaking the center wavelengths are 1271 to 1611 nm..
Lower cost:
Fewer channels to transmit are reflecting in ⅓ fewer costs to implement. Transponders use a wider band to transmit channels being less sophisticated design. There is no need for optical amplifiers since the spacing between the channels are not making them suitable for EDFA amplification. The resulting distances are smaller, like 60Km for 2.5Gbit/s signal. Passive CWDM is an implementation of CWDM that uses no electrical power. It separates the wavelengths using passive optical components such as bandpass filters and prisms. Many manufacturers are promoting passive CWDM to deploy fiber to the home. CWDM is based on uncooled distributed-feedback (DFB) lasers and wide-band optical filters. These technologies provide several advantages to CWDM systems such as lower power dissipation, smaller size, and less cost. The commercial availability of CWDM systems offering these benefits makes the technology a viable alternative to DWDM systems for many metro and access applications.
Easy to expand
Upgrading a 8 channel CWDM system to 16 channel system is easy  and is a matter of combining the mux/demux filters without the pain of adjusting the optical power or dispersion compensation DCM – modules like in case of DWDM. Low incremental cost: “Pay as you grow” Architecture. In DWDM systems, one connector can carry a whole cable’s-worth of traffic. If more than one connector is pulled and several incorrectly reconnected, the crap can truly hit the fan. Replacing, for instance, an amp (with DCM, OSC, and local connections) means everything disconnected has to be put back exactly as it was – or it might simply not work at all. Provisioning can be equally similarly disastrous. And these days, training exposure to these systems is minimal.Another benefit to the passive CWDM technology is that no configuration is necessary.The most complex step in CWDM integration is aligning and connecting the patch cables from the correct wavelength optic to the correct port on the multiplexers on each end of the link.
Specialized application evolve
CWDM – continues to evolve into specialized applications. Combination transport and optical routing or switches are being developed now. Add- on CWDM cards are being included in more transport devices as low cost options. Suppliers are continuing to drive down costs and increase capacity. CWDM and DWDM provides a unique “fit” and will complement not replace the other.
To sumarize the advantages of CWDM technology, it is worth to mention the following :
Lower power consumption – 20%
Smaller space requirements – 30%
Can use SMF or MMF cable
Can use LED or Laser’s for power
Smaller and cheaper wave filters
Cost saving on start up and expansion

Tunable DWDM Lasers – a short overview

by http://www.fiber-mart.com

A tunable laser is a laser whose wavelength of operation can be altered in a controlled manner. While all laser gain media allow small shifts in output wavelength, only a few types of lasers allow continuous tuning over a significant wavelength range.
In order to enable high-capacity optical networks, DWDM systems which use a single optical fiber for optical signals of several different wavelengths are being utilized.Wavelength tunable optical transceivers are becoming important as components that enable ROADM –  Reconfigurable Optical Add/Drop Multiplexer – functionality in next-generation networks. These transceivers have the characteristic that their wavelengths can be switched between different DWDM channels while in use in the network. Tunable transceivers are only available in DWDM form because the CWDM grid is too wide. Typically, these tunable optics are for the C-Band 50GHz. Around 88 different channels can be set with intervals of 0.4nm, which is the 50G Hz band. These optics usually start from channel 16 up to 61 but this depends on the manufacturer of the Router/Switch and which channels it supports.
Working principle
Multiple individual lasers, are built into one piece of silicon.
Tunable Distributed Bragg Reflector (DBR) Laser
One of the earliest types of tunable lasers is the distributed Bragg reflector laser. More modern tunable devices still share the same basic concepts and can be considered an evolution of DBR lasers. Like in DFB laser a DBR introduces a periodic variation of the refractive index effectively generating a Bragg grating or reflector. The cleaved front facet of the device acts as a second mirror. Only the wavelengths in a specific relation with the Bragg period survive in the cavity. Tuning is achieved by injecting current into the Bragg reflector. This results in a modification of the refractive index, which causes the Bragg peak to tune to different wavelengths. The Phase Section is primarily designed for fine-tuning the output Wavelength. The tuning range of these devices is proportional to the maximum change in the refractive index, typically below 20 nm.
Grating-Assisted Co-directional Coupler (GACC) Laser
The grating assisted codirectional coupler (GACC) laser is very similar to a DBR in operation. The purpose of this structure is to extend the tuning range of a DBR. The tuning element is a pair of vertically stacked waveguides with different material properties and a grating. This change leads to a larger tuning range in excess of 60 nm.
Sampled Grating DBR (SG-DBR)
The sampled grating DBR is another variant of DBR laser whose main difference is the presence of a pair of grating mirrors at either end of the cavity. The gratings are periodically sampled or blanked out, which results in a sequence of equally spaced short grating bursts. Just as in DBR, the gratings can be tuned by current injection. It can be proven that by differentially tuning the mirrors it is possible to achieve a wider tuning range than with a simple DBR.
External Cavity Tunable Laser (ECL)
The main characteristic of this architecture resides in moving out of the gain cavity the wavelength selection device, which is typically a MEMS or a thermally tunable filter. There is no integrated grating in the laser cavity like in a DFB or DBR. Tunable Lasers fabricated with this technique are usually very high-power (13 dBm of output power) and have a high spectral purity (SMSR > 50 dB). Among the disadvantages, an ECL is usually very slow to switch from one wavelength to another (in the order of seconds), furthermore in MEMS-controlled ECL mechanical reliability is a concern.
The operating frequency may be defined by a frequency selective feedback element that is thermo-optically tuned by the application of heat from an actuator without substantially tuning the cavity modes. Configuration is controlled by the operating system software in use for the DWDM system.
Thermal compensation of laser resonators is a requirement in components that must operate robustly within the narrow absolute frequency bands of the DWDM specifications.
Application of tunable DWDM lasers:
Sparing
Use tunables to reduce the number of line cards needed to back up all the different wavelengths in a system.
Dynamic provisioning
The wavelength of the tunable transmitter can be changed once the system has been deployed.
Reconfigurable optical add/drop multiplexers (ROADMs)
A simple, more flexible architecture for ROADMs has been proposed, which relies on the use of both tunable lasers and tunable filters.
Optical crossconnects
Tunable lasers can remove wavelength-blocking issues in OXCs.
Dynamic restoration
When a DWDM channel fails, a tunable laser could automatically restore service