As the following figure shows, if only the costsof transmission and regeneration equipment are taken into account (i.e. SDH regenerators in the caseof SDM and WDM TM with transponders with optical in-line amplifiers in the WDM case), the initial link cost of using WDM technology is more than double that of SDH. However, the WDM solution is more cost effective for the deployment of three and more channels in the network, due to the shared use of the optical amplifier in line.
As the following figure shows, if in addition to the above consideration, the cost of fiber is also taken into account, the cost advantage of WDM box becomes even more evident and is amplified as the number of channels increases. WDM solution is more cost effective for the deployment of three and more channels in the network.
WDM in the short haul
Regenerators are not necessary and optical degradations have less impact due to the limited distances in short networksdistance, hence the benefitsof WDM are less obvious than those of SDM or enhanced TDM solutions. However, fiber depletion and low cost optical componentsare now the driving force behind WDM in the metro area.
The short-haul application is linked to the interconnection of several pointsof presence (POP) in the same city. Let's analyze an example. The following figure shows that the transport network has at least two POPs per city, where customers can interconnect. With two-node interconnection techniques, such as drop and continue, customer networks can be interconnected with the transport network through two different POPs.
The result is a very secure architecture that can even survive POP outages without any impact on traffic. Thus, the traffic flow between two POPs in a city consistsnot only of traffic that passes through the city, but also of traffic that ends in the city andis protected by Drop and Continue. These increased needs for intra-city capacity have led to the deployment of WDM in the short-haul section of a transport network.
The main reason why WDM is preferred over SDM is because the fibers in a city must be leased to a third party or a fiber optic network must Leasing or building urban fiber is not only an expensive process, it is also a less flexible approach to improving capacity In a dynamic environment, where distributions and traffic volumes are changing rapidly , the amount of fiber to rent or build is difficult to predict in advance, therefore there are clear advantages of using WDM technology in flexibility as wavelength channels can be activated in a very short time.
Although specific short-haul WDM systems areare available worldwide, it is advantageous to use the same type of WDM system for itslong-haul network. While short haul WDM systems are less expensive than their long haul counterpartsand due to their inexpensive optical componentsthey can be used, they lead to a heterogeneous network which is not preferred by many. reasons. First, using two different systems resultsin increased operating and management costs. For example, a heterogeneous network requires more spare partsof equipment than a homogeneous network. Second, interworking between two different systems could cause problems. For example, a bottleneck can occur because short-haul WDM systems typically support fewer wavelengths than long-haul WDM systems.
Optical transport network architectures
Optical transport networke (OTN), as illustrated in the following figure, representsa natural next step in the evolution of transport networks. From a high-level architectural perspective, one would not expect OTN architectures to differ significantly from those of SDH. However, the fact that SDH involves digital network engineering and OTN involves analog network engineering leads to important, albeit subtle, distinctions. Exploring these distinctions leads us to understand aspectsof OTN that may differ from their SDH counterparts.
Evolving OTN WDM architectures (including network topologies and survival patterns) will closely resemble, if not mirror, those of SDH TDM Networks. This should be surprising, however, since SDH and OTN are both connection-oriented multiplexed networks. The major differences come from the shape of the techMultiplexing nology: digital TDM for SDH vs analog WDM for an OTN.
The digital / analog distinction has a profound effect on the fundamental cost / performance trade-offs in many aspectsof the OTN network and system design. In particular, the complexities associated with analog network engineering and maintenance implications represent the majority of challenges associated with OTN.
To meet the short-term need for increased capacity, WDM point-to-point line systems continue to be deployed on a large scale. As the number of wavelengths and the distance between terminals increase, there is a growing need to add and / or remove wavelengths at intermediate sites. Therefore, Flexible Reconfigurable Optical ADMs (OADMs) will become integral partsof WDM networks.
As more and more wavelengths are deployed in the networks ofs operators, it will be increasingly necessary to manage capacity and transfer signals between networks at the optical channel level. Likewise, DXCs have emerged to manage capacity at the electrical layer level, optical interconnects(OXC) will emerge to manage capacity at the optical layer.
Initially, the need for optical layer bandwidth management will be most acute in the core transport network environment. Here, logical mesh-based connectivity will be supported through physical topologies, including OADM-based shared rings of protection and OXC-based mesh restoration architectures. The choice will depend on the degree of bandwidth desired by the service provider "from the build" and the survival timescale requirements.
Similar requirementsfor bandwidth management are emerging for the environment.Inter-office and metropolitan access, OADM ring solutions will also be optimized for these applications: shared optical protection rings for mesh requestsand dedicated optical protection rings for hubbed requests. Therefore, just as OA has been the technological catalyst for the emergence of WDM point-to-point line systems, OADMs and OXCs will be the catalystsfor the emergence of OTN.
While the elementsof the optical network assume the transport layer functionality traditionally provided by SDH equipment, the optical transport layer will come to serve as a unifying transport layer capable of supporting both legacy and converged packet core network signal formats. Of course, the transition from the service provider to the OTN will be predicted on the transfer of the functionality of the "SDH type" transport layer to the optical layer, at the same time.ps as the development of a maintenance philosophy and associated network maintenance functionalities for the emerging optical transport layer.
Survival is central to the role of the optical network as a unifying transport infrastructure. As with many other architectural aspects, optical network survivability will bear a high level resemblance to SDH survivability, as network topologies and types of network elementsare so similar. Within the optical layer, survivors will continue to provide the fastest possible recovery from fiber cutsand other physical media faults, as well as provide efficient and flexible management of protection capacity. .
OTN is conceptually analogous to SDH, in that sublayers are defined that reflect client-server relationships. Since OTN and SDH are both oriented multiplex networksWhen connected, it should come as no surprise that the restoration and protection schemes of the two are remarkably similar. The subtle but important difference is worth repeating: while the TDM network is based on digital time slot manipulation, the OTN / WDM network is based on an analog or optical frequency slot. Channel manipulation (wavelength). So while we can expect similar protection and recovery architectures to be possible with the two technologies, the types of network outages that may need to be accommodated in a particular survival pattern can be very different. .
Optical Layer Survivability
Telecommunications networks are necessary to provide reliable and uninterrupted service to their customers. The overall availability requirementsare in the order of 99.999% or more, which would imply that the network cannot be in panne more than 6 min / year on average. As a result, network survivability is a major factor affecting the way these networks are designed and operated. Networks must be designed to handle link or fiber outages as well as equipment failures.
The network can be considered to be made up of several layers interacting with each other, as shown in the figure above. Different operators choose different ways of realizing their networks using different combinations of stratification strategies. Incumbentsuse their large installed base of 'SDH equipment and extensive capabilities for preparing and monitoring digital interconnections.
In contrast, an operator offering Internet Protocol (IP) -based services seeks to simplify network infrastructure using IP as a transport layer basic without using SDH. Carriers distinguished by quality (and The optical layer provides light paths to the upper layers, which can be thought of as client layers that use the service provided by the optical layer. Lightpaths are circuit-switched pipes carrying traffic at fairly high bit rates (for example, 2.5 Gb / s or 10 Gb / s). These light paths are typically configured to interconnect client layer equipment, such as SDH ADMs, IP routers, or ATM switches. Once configured, they remain fairly static over time.
The optical layer includes optical line terminals (OLTs), ADMs optical (OADM) and optical interconnections (OXC),as shown in the following figure. OLTs multiplex multiple channels into a single fiber or pair of fibers. OADMs remove and add a small number of channels from / to an aggregated WDM stream. An OXC switches and manages a large number of channels in a high traffic node location.
We look at optical layer protection from a service perspective, in terms of the types of services to be provided by the optical layer to the upper layer, and then compare the different optical layer protection schemes that have been proposed in terms of cost and efficiency of bandwidth depending on the mix of services to be supported. It is somewhat different, who tends to regard the protection of the optical layer as analogous to the protection of the SDH layer.
Why optical layer protection?
The IP, ATM and SDH layers shown above figure, all integratedrent protection and restoration techniques. While these layers were all designed to work with other layers, they can also work directly over fiber and therefore do not depend on other layers to handle protection and restoration functions. As a result, each of these layers has itsown protection and restoration functions. So, the question arises, why do we need the optical layer to provide itsown set of protection and restoration mechanisms. Here are some of the reasons -
-
Some of the layers working on top of the optical layer may not be fully able to provide all the necessary protection functions in the network. For example, the SDH layer was designed to provide full protection and therefore would not depend on the protection of the optical layer. However, professional techniquestection in other layers (IP or ATM) by themselves may not be sufficient to provide adequate network availability in the presence of faults.
There are currently many proposals for making the IP layer work directly over the optical network. layer without using the SDH layer. Although IP incorporates fault tolerance at the routing level, this mechanism is cumbersome and not fast enough to provide adequate QoS. In this case, it becomes important for the optical layer to provide rapid protection to meet the overall availability requirementsof the transport layer.
-
Most carriers have huge investmentsin existing equipment that do. do not provide any protection mechanism, but cannot be ignored. A transparent introduction of the optical layer between this equipment and the raw fiber provides a low cost upgrade of the infrastructure.ture over long fiber links with increased survivability.
-
Protection and restoration of the optical layer can be used to provide an additional level of resiliency in the network. For example, many transport networks are designed to handle a single failure at a time, but not multiple failures. Optical restoration can be used to provide resilience against multiple failures.
-
Protection of the optical layer may be more effective in handling certain types of failures, such as fiber cuts. A single fiber carries multiple wavelengths of traffic (eg 16 to 32 SDH streams). A fiber cut therefore resultsin 16 to 32 of these SDH streams being restored independently by the SDH layer. The network management system is inundated with a large number of alarms generated by each of these independent entities. If the fiber cut is restored sufficiently rquickly through the optical layer, this operational inefficiency can be avoided.
-
Significant savings can be achieved by using the protection and restoration of the optical layer.
Limitations - Optical layer protection
Here are some of the limitations of optical layer protection.
-
It cannot handle all types of failures in the network. For example, it cannot handle the failure of a laser in an IP router or SDH ADM connected to the optical network. This type of failure must be handled by the IP or SDH layer, respectively.
-
It may not be able to detect all types of failures in the network. The light paths provided by the optical layer can be transparent so that they carry data at various bit rates. the optical layer in this case may in fact not know who exactly is transported on these pathsluminous. As a result, it would monitor traffic for degradations, such as increasing bit error rates, would normally invoke a protection switch.
-
The optical layer protectstraffic in unitsof light paths. It cannot provide different levels of protection to different partsof the traffic carried on the light path (part of the traffic may be high priority, the other less priority). This function must be performed by a higher layer that handles traffic with this finer granularity.
-
There may be link budget constraintswhich limit the protective capability of the optical layer. For example, the length of the protection road or the number of nodes crossed by the protection traffic can be limited.
-
If the overall network is not carefully designed, there may be race conditions when the optical layerand the client layer both try to protect the traffic from failure simultaneously.
-
The technology and protection techniques have not yet been tested in the field, and the large-scale deployment of these new protection mechanisms will therefore take a few years.
Finishes of protected entities
Before going into the details of protection techniques and the compromises between them, it is interesting to define the entities that are protected by the optical layer and the client layer. These entities are represented in the following figure.
Client device port
The portson the client device may fail. In this case, the optical layer does not cannot protect the client layer by itself.
Intra-site connections between the client and optical equipment
Cables inside a site can be disconnected, mainlynt due to human error. This is considered a relatively probable event. Again, full protection against such occurrences can only be supported by combined client layer and optical layer protection.
Transponder cards
Transponders are interface cards between client equipment and the optical layer. These cards convert the signal from the client equipment to a wavelength suitable for use within the optical network, using optical-to-electrical-to-optical conversion. Therefore, the failure rate of this card cannot be considered negligible. Given the large number of these cards in a system (one per wavelength), a special protective backing for them is in order.
External installations
This fiber installation between sites is considered to be the least reliable component of the system.e. Fiber cutsare quite common. This category also includes optical amplifiers which are deployed along the fiber.
Entire nodes
An entire node can fail due to errors of maintenance personnel (eg tripping of circuit breakers) or entire site failures. Site failures are relatively rare and usually occur due to natural disasters such as fires, floods, or earthquakes. Node failures have a significant impact on the network and therefore still need to be protected against, despite their relatively low probability of occurrence.
Protection against recovery
Protection is defined as the main mechanism used to manage a failure. It must be very fast (generally the traffic must not be interrupted for more than 60 ms in the event of failure of SDH networks). As a result, the rouProtective roads usually need to be pre-planned so that traffic can be quickly switched from normal routes to protective roads.
Due to speed requirements, this function is generally performed in a distributed manner by network elementswithout relying on a centralized management entity to coordinate protection actions. With the exception of recent (and not yet proven) fast mesh protection schemes, protection techniques tend to be fairly straightforward and are implemented in line or ring topologies. They all end up using 100% of the network access bandwidth.
In contrast, restore is not a primary mechanism used to handle failures. After the protection function is complete, recovery is used to provide efficient routes or additional resiliency against further failures before the first failure occurs.e is corrected. As a result, it can afford to be quite slow (from a few seconds to a few minutes sometimes).
Catering routes do not need to be planned in advance and can be calculated on the fly by a centralized management system, without requiring a monitoring function. More sophisticated algorithms can be used to reduce the excess bandwidth required, and more complex mesh topologies can be supported.
Sub-layers in the optical layer
The optical layer consistsof several sublayers. Protection and restoration can be performed at these different layers. We can have systems that protect individual light paths or optical channels. These diagrams manage fiber cutsas well as breakdowns of terminal equipment, such as lasers or receivers.
We can have some schemes that workat the aggregate signal level, whi ch corresponds to the OMS (Optical Multiplex Section) layer. These diagrams do not distinguish between the different light paths which are multiplexed together, and restore them all simultaneously by switching them as a group.
The term path layer protection is used to refer to schemes that operate on individual channels or light paths and line layer protection to refer to schemes that operate at the path layer level. optical multiplex section. See Table 1 for a comparison between the properties of the path and line layer diagrams, and Table 2 and Table 3 for the different path and line diagrams.
Table 1: A comparison between line protection and path protection
Criterion | Line protection | Path protection |
Protectsagainst | Inter-office services Site / node failures | Inter-office services Site / node failures Equipment failures |
Number of fibers | Four, if single-level multiplexing is used | Two |
Can handle single path failures / degradation | No | Yes |
Support traffic that should not be protected | No | Yes |
Equipment cost | Low | High |
Bandwidth efficiency | Good for protected traffic | Weak for unprotected channels |
Table 2: A comparison between linear layer schemes
Schema | Protectsagainst | Topology | Constraints/ gaps | Benefitsfor the customer |
1 + 1 line | Line breaks | Point to point | Various routes needed to protect fibers | The easiest to set up and use |
1 + 1 ligne | Line breaks | Point to point | Various routes needed to protect fibers | Support for low priority traffic Reduced losses (about 3 dB) |
OULSR | Line breaks Node faults | Metropolitan ring | Optical layer degradation There is additional power loss due to bridging at the line of signals | Simple to set up and use Can be done using passive elements(instead of optical switches) |
OBLSR | Line breaks Node faults | Metropolitan ring | Degradations of optical layer | Reuse protective bandwidth Support low priority traffic |
Mesh line protection | Cutsline Node errors | All | Limited by alterations in the optical layer Based on an all-optical cross-connect Difficult for the elderly | Efficient Low cost |
Table 3: A comparison between path-layer schemes
Schema | ProtAgainst | Topology | Constraints/ Deficiencies | Customer benefits |
Client layer protection | Customer equipment failures Intra-office installations Transponder failures Inter-office installations Node failures | All | Requires Most expensive | Most extensive protection |
1: N equipment protection | td = "vertical-align: middle; "> Transponder faults Linear or ring | | Very low cost Effective bandwidth |
1 + 1 path or OUPSR | Inter-office services Node faults | All | td = " vertical -align: middle; "> Requires Bandwidth consumption
Similar to client protection Easy to develop and use |
OBPSR | Inter-office installations Node faults | Virtual ring | | Bandwidth reuse protection Support low priority traffic |
Protection of mesh paths | Inter-office installations Node faults | Any | Requires an OXC Very complex to implement and use | td = "vertical-align: middle; "> High performance
The physical topology of the network can be any mesh, passing light paths between the nodes of client equipment. The virtual topology from the point of view of the client equipment is restricted according to the client layer (eg rings for SDH). 2The physical topology is any mesh, while the virtual topology of the light paths is a ring.
Consider, for example, the two protection schemes shown in the following figures. These two schemes can be thought of as 1 + 1 protection schemes, i.e. they
Layer protection line and path layer
There are important differences in cost and complexity between the two approaches. Line protection requires an additional splitter and a switch to an unprotected system. However, path protection requires one splitter and one switch per channel. More importantly, trail protection typically requires twice as many transponders and twice the multiplex / demultiplexer resources of the line protection. Therefore, path protection is almost twice as expensive as line protection, if all channels are to be protected. The story changes, however, if not all channels need to be protected.
The basic protection schemes
A comparison of the protection schemes can be found in Tables -1, 2 and 3. Optical layer protection schemes can be classified in the same way as SDH protection schemes and can be implemented at theof the client layer, the path layer or the line layer.
Client protection
A simple option is to let the client layer take care of itsown protection and not have the optical layer perform protection. This can be the case for SDH client layers. Although this is simple from an optical layer point of view, significant cost advantages and bandwidth savings can be obtained by providing optical layer protection. While the client protection method can support point-to-point, ring, or mesh client networks, it is important to note that from an optical network perspective, all of this translates into optical mesh support. , because even a point-to-point link client can In client layer protection, client work and protection paths are routed in a completely Path layer schemes
1 + 1 path protection
This scheme requires two wavelengths on the network, because as well as two setsof transponders at each end. When applied to a ring, this protection is also referred to as Optical Unidirectional Path Switching Ring (OUPSR) or Dedicated OCh Protection Ring (OCh / DP Ring).
Implementation Notes - Bridging is usually done via an optical coupler, while selection is via a 1 x 2 optical switch. The receiving end can decide to switch to the emergency path without coordination with the source.
Switched ring ofbidirectional path
This scheme is loosely based on the SDH 4-fiber bi-directional line switched ring (BLSR) and relies on a shared protection bandwidth around the ring. When a working light path fails, the nodes coordinate and try to send traffic through the designated protection bandwidth in the same direction around the ring (to overcome transponder faults). This is a range switch. If unsuccessful, nodes loop traffic around the alternate path around the ring to the other end of the failure. This action is a ring switch.
The scheme allows non-overlapping light paths to share the same protection bandwidth as long as they do not fail together. This scheme is also referred to as the OCh Shared Protection Ring (OCh / SPRing).
Implementation Notes - This scheme can be implemented in an OXC or,through much smaller switches in OADM. Switches are required for each protection channel. It is similar to the SDH BLSR standard.
Mesh path protection
This scheme allows global mesh protection with very fast switching (in less than 100 ms) for each faulty light path separately from a emergency path, shared by several light paths potentially taking a different route per light path. If unsuccessful, it is signaled to all relevant nodes that set up backup paths.
Implementation Notes - These schemas are being implemented in OXCs. Due to time constraints, the predefined backup paths are stored in the nodes of the network and are activated according to the types of failure.
Meshpath restoration
Unlike meshpath protection, this scheme does notnot work have strict time constraints. This device calculates alternative routes using itstopology and broadcastsnew configuration information to the nodes, which define these routes. Nodes don't need to keep n / w information.
Implementation Notes - The centralized nature of this scheme ensures more optimized protection routes and reduces the complexity of implementation and maintenance.
1: N Equipment protection
One of the most complex (and therefore prone to failure) modules of a typical WDM terminal is a transponder. The 1: N protection designates a spare transponder to take over in the event of failure of the normal transponder.
Implementation Notes - This scheme is more generally based on a designated protected wavelength. In the event of a failure, both ends should fail over using fast signaling protocols, not likeAPS in SDH.
Line layer schemes
1 + 1 linear protection
This sc heme is based on bridging the entire WDM signal in bulk on a pair of Linear protection 1: 1
This scheme requires a configuration similar to the previous one (i.e. 1+ 1 linear), however the signal is switched to the work or protection path, but not on both. Although this increases the coordination load, it allows low priority traffic to run on the backup path (until needed to protect the working path). It also resultsin less optical power loss due to all signal energy being directed to one path instead of two.
Implementation Notes - Switching is usually done at theusing a 1 optical × 2 switch. Coordination is achieved through a rapid signaling protocol.
Optical Unidirectional Line Switching Ring (OULSR)
The scheme is similar to the OUPSR scheme except that bridging and signal selection is performed for the aggregate WDM signal. This allows for a more optimized design, lower cost, and very different implementations.
Implementation Notes - One implementation of this scheme is based on passive couplers which run the optical ring in a broadcast medium. Instead of using OADMs, this scheme is based on simple OLTs, each coupled in rings both clockwise and counterclockwise, so that each of the wavelengths are transmitted and received on both fibers. Under normal conditions, the link is artificially disconnected, resulting in a linear bus, when the fiber-cut link is reconnected.
Bidirectional Line Switched Ring
This scheme is similar to the OBPSR scheme in both aspectsof the protocol and the protection actions used (range and ring switching) As all line layer schemes, the aggregate WDM signal is bulk switched to a dedicated shield fiber (requiring four fibers), or to a different WDM band in a single fiber (allowing only two fibers, but requiring one o scheme) digital multiplex). This scheme is also referred to as the OMS Shared Protection Ring (OMS / SPRing).
Implementation Notes - As the back-up route loops optically around the entire ring, optical line amplifiers may be required along the way. relief to compensate for losses. The circumference of the ring is also limited by other optical degradations. Therefore, this option is best suited for metro applications.
Prmesh line otection / restoration
This scheme is based on fully optical interconnectswhich divert the WDM signal from a failed facility onto an alternate route and back to the other end of the faulty installation.
Implementation Notes - Like OBLSR, this scheme is limited by optical impairmentsthat can develop along alternate routes and requires optical design precautions.
Considerations for choosing the protection scheme
The criteria that could be used by an operator to select the protection schemes to be used in the network. A simplified decision table for this is shown in the following figure assuming equipment and line protection are required.
The cost of protection
Another criterion from the carrier's point of view is the cost of the system in atminus two aspects-
- Cost of equipment
- Bandwidth efficiency
Both depend on the combination of traffic services, i.e. the fraction of the traffic to be protected by the optical layer.
The following figure shows the cost of equipping path layer schemes and equivalent line layer schemes as a function of traffic composition. If all traffic needs to be protected, trail layer schemes require approximately twice the equipment of line layer schemes because there is less sharing of common equipment.
However, the cost of trail layer protection is proportional to the number of channels that are to be protected, as each channel requires an associated multiplexer / demultiplexer and termination equipment. Thus, the cost of protecting the path layer decreases if fewer channels need to be protected. In the event that no canal does not need to be protected, path layer schemes will cost roughly the same as line layer schemes, assuming no additional common equipment is deployed.
The story is different from the point of view of the efficiency of the tape bandwidth, as shown in the following figure. In an in-line protected system, the protection bandwidth is consumed for light paths that require protection as well as those that do not require protection. In path protection systems, light paths that do not require protection can use the bandwidth, allowing other unprotected light paths to use bandwidth that would otherwise have been wasted on unwanted protection.
'therefore that while a large part of the light paths could be left unprotected, the protection of the path layerrecovers the cost by carrying more working traffic on the same network as the line layer protection.