PDF of this tutorialLess Complex, More Efficient
The DWDM point-to-point architecture is inherently simple to build and troubleshoot. Unlike some systems, planners do not have to determine the ultimate capacity of the trunk prior to construction. Furthermore, they need not live with their capacity decision for years to come. With its fundamentally different multiplexing architecture, point-to-point DWDM enables protocol transparency, incremental growth, and capacity expansion over time, while dramatically reducing start-up costs.
Because DWDM is not TDM based, all channels are discrete and essentially stand independent of each other. Each channel card represents an individual customer or protocol. These can be added one at a time as colors are added; channel assignments need not be on adjacent channels, and different kinds of data or different speeds can be added over each channel. In the TDM world, the entire backbone must be in place at the outset for the total number of customers that are going to be on the circuit.
Point-to-point solutions are also extremely efficient. Every possible wavelength can be utilized without regard to the rest of the network. Furthermore, as long as the light can originate and terminate end to end, amplifiers are rarely needed. A typical system will transport effectively for up to 100 kilometers without an amplifier, which is a significant distance inside a metro area and more than enough to link central offices and facilities. With no amplifiers or additional equipment required, point-to-point represents a simple, cost-effective, and extremely efficient solution.
Incorporates Both an Add/Drop Multiplexer (ADM) and Digital Cross-Connect System (DCS) The role of the ADM is to determine which channels remain at the site and which pass through. A point-to-point system with an OXC offers the same functionality as a discrete ADM. Even though specific add/drop nodes are used, the OXC can compliment the add/drop function. When the network sophistication requires such complexity, the OXC may be used in place of an OAD. Because additional equipment is not required, costs are reduced. There is also no electronic latency or point-of-component failure when one eliminates the need for a TDM–based DCS.
Amplifiers are not standard equipment in every DWDM network and are only necessary if there is significant decibel loss in the span. If the light that is being generated at the origination site with substantial power sustains enough loss through the fiber route so that the light arrives below the specified receive sensitivity specification at the receiving end, only then are amplifiers required.
OADM nodes built from DWDM filters cause different optical channel-power levels for pass-through and inserted light. This all-optical, pass-through approach requires attenuators and amplifiers at each node to balance individual channel levels. Telcordia’s OADM standards stipulate amplifiers at each node for this reason. Point-to-point systems, however, contain O-E-O conversion at each end and therefore always regenerate the power into balanced channel levels on each hop, thus eliminating the need for custom attenuation and amplifiers. Ultimately, this method eliminates a costly single-point failure from the system while also eliminating the complex step of level balance planning.
Allows for Multiple Customers to Share a Wavelength
Placing an electronic photonic concentrator (EPC) at all nodes improves the economics at lower speeds. An EPC is a combination of optical and electrical technology that allows the division of DWDM channels into point-to-point subrate channels of varying bandwidth while maintaining total protocol and format transparency. As a result, low-speed channels that cannot economically justify their own wavelengths can be combined and placed on one wavelength. In addition to subrate multiplexing, EPCs provide direct optical connections to customer premises equipment (CPE). This technique elegantly provides subrate traffic to the mesh network (see Figure 5).
Figure 5. Allows Multiple Customers to Share
Programmable Regenerators at Point-to-Point Transponders Support Usage-Based Circuit Pricing Through regeneration, every receiver would reclock, retime, and reamplify the signal, cleaning, the pulse on every point-to-point hop. In addition to improving the performance of the signal, the regeneration feature enables the provisioning of speed—if it is selectable at a specific site. As a result, network operators can offer different tiers of service and pricing to their customers. This also provides an excellent control mechanism, ensuring that customers receive exactly the quality of service (QoS) for which they are paying.
Regeneration is an electrical feature built inside each DWDM channel module. As a result, it is extremely cost-effective and requires no additional space. In the TDM world, regeneration typically requires the installation of a brand-new node. Essentially, the service provider must purchase a whole core multiplexer to regenerate the signal. However, no revenue-generating customers are added.
Open Systems and Interoperability, OXC Used at the Core
Differences exist between the light characteristics of DWDM channels as a result of the quality needed for different channel densities. A four-channel system's light used poorer tolerances than a 32-channel system. Frequency accuracy and channel width are just some of the important parameters that vary between systems. Thus, true end-to-end interoperability from different systems’ DWDM light signals will not be achieved unless all systems use the most dense and expensive specifications. This is not economical, and it does not allow for future systems of greater density to interoperate with current systems within the DWDM layer. If all systems accept and emit a standard light such as 1310 nm at their terminal interfaces, however, interoperability between nodes of different types and vendors will become simplified. Point-to-point systems of virtually all vendors can accept and deliver 1310 nm and thus can use a common standard for interconnection. Point-to-point DWDM with optical cross-connection allows for lower-density systems to connect to higher-density systems and supports a mixed-vendor environment for end-to-end circuits.
Fewer Nodes Are Needed
The pay-as-you-grow nature of point-to-point dictates that nodes must only be installed when fiber exhaust is reached or when a new customer is to be added. This contrasts sharply to planning the entire network and building rings for all sites, even though many sites may have no customers.
Metropolitan DWDM and Later OXCs Can Be Provisioned at the Time of Installation for the Required Method of Protection
This enables the transition from bidirectional line switched ring (BLSR) to mesh and 1 for n restoration schemes. BLSR is a method of SONET transport in which half of the working network is sent counter-clockwise over one fiber, and the other half is sent clockwise over another fiber. Protection fibers are shared across the entire ring. If any connection between two sites breaks, the protection fibers are used to reheal the link by traveling around the ring in the other direction. This requires complex software, and its function must be planned in advance of building the network. It also requires knowledge of the entire fiber network that may not be available in the new environment of shared, co-opted, and sublet fiber networks.
As a feature of the AON, protection against a fiber break can be provided in a mesh network on a per-channel basis. Providers and customers can choose which circuits to protect and how much protection is needed at the time of installation. Thus, network operators can offer tiered pricing for different levels of protection. With this higher level of flexibility and greater granularity, service providers can arrange multiple protection routes if the customer so desires and is willing to pay. In the TDM world, one-size-fits-all is still the rule, and all decisions regarding circuit protection must be made prior to the initial build out.
AONs Provide Provisioning, Including Wavelength (Lambda) Assign and Conversion for Many Routes at a Node Site
Traffic arriving on one point-to-point system at a site can be easily transferred to another system via the cross-connect function, which could be simple fiber jumpers or a modular cross-connect system (see Figure 6). Circuits are built to pass between the DWDM nodes, regardless of channel numbers of each span.
Figure 6. Provisioning Flexibility
Small Sites Can Start with Basic Metropolitan DWDM Systems until Traffic Justifies an OXC or Combination of the Two
A start-up CLEC, for example, would find it difficult to justify purchasing an OXC with 512 ports if only two circuits were in use. With point-to-point, it is possible to use small DWDM systems at the outset until there are enough customers to warrant an investment in a software-control system. This ability enables service providers to enter small markets more cost-effectively and be extremely price-competitive with existing providers that are recouping revenues to pay for much larger networks.
Shorter Distances Can Nullify Most Common Barriers
Lower launch power can be used to engineer and terminate a point-to-point system. This leads to lower costs and a safer network. In addition, as a result of lower power, the incidents of cross-talk and mixing are reduced considerably. Furthermore, in a point-to-point scenario, there is no need to use a high-power, long-haul system. In a ring topology, this may be required, as optical nodes are subject to high loss. Combating this would require multiple amplifiers or high launch power.
With point-to-point, distances are also shorter between nodes. If the light is being terminated at every mesh site, then dispersion is not a problem. Dispersion is a particular form of distortion that happens to all light signals as they traverse distance, and, as the signal degrades over every kilometer, errors occur. To combat this, the complicated procedure of dispersion calculations must be completed for circuits without regeneration.
Point-to-point also simplifies circuit management and fault isolation. If every point is monitored, it is an easy procedure to conduct segment-by-segment troubleshooting to identify the breaking point. With a point-to-point metropolitan DWDM system and future OXC combination, it is relatively simple to switch service from the problem channel to a working channel from a central console.
Pay-as-You-Grow Defers Costs
As mentioned previously, point-to-point enables providers to pay to upgrade one channel at a time instead of having to install a costly core system with multiple OADMs and amplifiers.
Investment Protection/Return on Investment
By following a strategy of point-to-point DWDM systems, providers can overcome the immediate challenge of fiber exhaust. Moreover, by adding OXCs to metropolitan DWDM systems, it is possible to evolve seamlessly into a fully optical network, which guarantees interoperability within and also interfaces to other optical networks.
As for reliability, all circuit protection issues can be handled through features of the metropolitan DWDM and the OXC, including wavelength translation between hops. If regeneration is used, a point-to-point circuit that leverages a cross-connect can cover vast distances. For example, a firm in Southern California could build an end-user circuit using cross-connected city metro systems through many communities throughout the state without having to install a specific long-haul route, as long as there is connectivity between the various metro areas along the way. The additional advantage is that the system's many points of presence (PoPs) make it a blended metropolitan and long-distance system so that delivering service to customers along the routes is very convenient. Even alternate routing choices of more than two paths can be built for each channel independently, without resorting to costly ring-layer, protect-all-channels designs.
A mesh topology is also an ideal compliment to the high demand for data networking and packet-based transport. While telecommunications providers have typically built networks based on rings as a result of their self-healing and rerouting properties, data backbones have always been mesh designs. This is also partly a result of the complex interconnection of IP routers. Certainly, a mesh topology shared by both IP and transport networks provides an elegant solution. Moreover, a mesh with optical routing can also provide self-healing capabilities without the complexity associated with ring software. Alternate routing through the mesh either by the router, switch, or optical network becomes an apparent solution.
