It is the old adage of supply and demand. The commodity is bit rate capacity, often called bandwidth by IT people. Certainly, the demand is there and probably doubling every one to two years. In the supply chain the bits, now in great quantities, must be transported in some manner. The other requirement is that they be directed to or accepted from a user. What we are talking about is information represented by bits. The only transport that can handle these great quantities is an optical link. At every node in the
optical network the stream of bits must be returned to the electrical domain for switching/routing. The goal is an all-optical network without the laborious return to the electrical domain at switching nodes except at the input–output points. Optical links are presently carrying 10 Gbits/s per bit stream per wavelength.
With dense wavelength division multiplexing (DWDM) a single fiber can carry 8, 16, 32, 40, 80, 160, or 320 wavelengths. Within some years of the publication of this book, we will have 40 Gbits/s per bit stream and a single fiber will support in that same time frame 320 wavelengths or more. If each wavelength carries
40 Gbits/s, then a single fiber will have the capacity to carry 40 × 160 Gbits/s or 6400 Gbits/s.
As seen from the network provider, a major drawback in fiber networking technology is the costly requirement for repeated conversions from the optical domain to the electrical domain and back again, called OEO, at every regeneration point and for periodic signal monitoring along the line. Amplifiers replacing regenerative repeaters where the add–drop function is not necessary have improved the cost situation somewhat. When optical switching without OEO is employed in the network, the requirement for regenerative repeaters (regens) in the network will be vastly reduced.
In the PSTN, fiber networks are primarily based on SONET and SDH infrastructures. In such cases the cost of regenerating optical signals can be very expensive, especially when it requires full SONET or SDH termination equipment at every ADM regeneration point. It has been found that even in these relatively
homogeneous all-SONET environments, optical layer management can be a key factor in maintaining system integrity. Present optical networks, unfortunately, require OEO conversion for effective network management.
Even at locations along the fiber line where full conversion is not necessary, partial conversion at key points can be vitally important for monitoring to assure circuit quality. At amplifier locations as in conversion signal points, active signal monitoring capabilities should be included. This will require optical signal splitting and optical-to-electrical conversion of some portion of the light signal. The move toward direct deployment of gigabit ethernet (GbE) over the MAN WAN is a factor that may help to temporarily mitigate some of the push for all-optical switching because the cost of interface equipment for GbE fiber transport
links is significantly less than for a SONET or SDH link. We see it as highly unlikely that GbE could completely replace SONET/SDH in the foreseeable future for anything other than very limited environments. However, the real-world impact of GbE deployment is likely to be increased traffic heterogeneity
that will further drive the need for effective optical-layer management. The ultimate goal behind DWDM deployment is the provision of more bit rate capacity. As a corollary then, MANs and WANs in the real world require optimization of the use of each wavelength’s bit rate capacity. With the PSTN,
where long-haul links traverse the network core, this goal of optimized utilization has typically been accomplished by pre-grooming all traffic such that uniform groups of signals can be efficiently transported long distances with a minimum of intervening decision points along the way. However, for traffic traveling nearer
the edge of the transport network, new-generation equipment needs to provide a higher level of traffic monitoring and grooming capabilities within the optical domain to achieve a balance of flexibility, performance, and capacity utilization. We would argue that in most cases it would make little economic and practical
sense to invest in DWDM and then map GbE connections across individual wavelength carriers on a one-to-one basis. Therefore, the push for aggregation of multiple connections can very quickly lead to a mix of heterogeneous non concatenated traffic traveling within a shared wavelength with a multitude of
different end-point destinations. The goal of the industry is to make the network all-optical except at each
edge transition point. This point will be on a user’s premises. By transition we mean the conversion of light to equivalent electrical information expressed as 1s and 0s. This chapter’s objective is to describe various steps to be taken toward what may now be called the “all-optical” network, its topology, and routing/switching
in the optical domain.

The following is a list of new technology and radical approaches to make an all-optical network a reality:
a. Optical switching
b. Advanced wavelength demultiplexing and multiplexing
c. Tunable filters
d. Stabilized lasers
e. New approaches to modulation
f. Improved optical amplifiers with flat gain characteristics
g. New and larger optical cross-connects (OXCs)
h. Optical add–drop multiplexers (OADMs)
i. Signaling techniques in the light domain

2.1 Derived Technology Applications
The semiconductor optical amplifier (SOA) is one of the most promising technologies for optical networks. By integrating the amplifier functionality into the semiconductor material, the same basic component can perform many different applications. SOAs can perform switching and routing roles in an integrated
functionality within the semiconductor material. Other important elements of an all-optical network are space switches, wavelength converters, and wavelength selectors, all of which can be fabricated from SOAs.
The new-generation managed optical network is moving toward a distributed switching model in which lambda (╬╗) switches with intelligent Layer 1 cross connect capability are distributed at various points along the network border. This concept is illustrated in Figure 20.1. Such an architecture provides seamless
and efficient Layer 1 management of heterogeneous traffic types throughout the network, without sacrificing performance or flexibility in either the core or edge environments. This global distributed-switching architecture is equally adaptable to using dedicated wavelengths packed with homogeneous traffic for long-haul point-to-point transport or for flexibly managing heterogeneous traffic on dynamically
allocated short-haul wavelengths. With cross-connects along the edge of the network cloud, there is an emerging need for supporting a managed optical layer within a distributed optical switching
environment. This outlines the crux of the matter and presents significant opportunities and challenges for both the semiconductor level and module-level developers and manufacturers. To achieve the required performance requirements, next-generation cross-connects need to be closer to the network by providing
Layer 1 switching as opposed to the present traditional Layer 2 switching.
There will be two cross-point designs: asynchronous and synchronous. The higher-speed asynchronous cross-points enable heterogeneous MAN implementations to efficiently support different types of native-mode traffic within the same ring. In long-haul networks, there will be innovative uses of synchronousswitching
cross-points which will provide the necessary performance requirements. These switches are seen as more of a time–space–time switch rather than the more rudimentary space switch. These new-generation synchronous crosspoints will incorporate Layer 1 grooming capabilities that can selectively switch SONET (or SDH) or any other TDM (time-division multiplex) signals between any combination of inputs and outputs.
It is expected that these optical Layer 1 switching capabilities will use highspeed
synchronous ICs. This next-generation synchronous cross-point switch will offer the capability to selectively groom out and switch and STS-1 (STM-1) from within STS-48 (STM-16) or STS-192 (STM-64) bit streams. These devices will provide complete flexibility for provisioning IC-level managed optical crossconnects from any STS-1 input to any STS-1 output. Non-SONET traffic mapped to STS-N-equivalent containers and protocol-independent wrapped traffic can be switched within the same cross-connects. These high-density, high-speed grooming switches are deployed along the edge
of the switching network cloud. They can optimize capacity utilization whileefficiently making Layer 1 access decisions to partition out traffic to outlying internet protocol, GbE, ATM, fiber channel, or other Layer 2 switches. Localized Layer 2 functions such as routing and policy management are appropriately handled
by the outlying switches while Layer 1 access switches provide high-speed performance IC-level switching/grooming of DWDM wavelengths [1].

Today’s data networks typically have four layers:
• IP for carrying applications
• ATM for traffic engineering
• SONET/SDH for transport
• DWDM for capacity
This architecture has been slow to scale, making it ineffective for photon networks. Multilayering architectures typically suffer from the lowest common denominator effect where any one layer can limit the scalability of the other three layers and the entire network.
4.1 Two-Layer Networks Are Emerging
For the optical network developer, an absolute prerequisite for success is the ability to scale the network and deliver bit rate capacity where a customer needs it. Limitations of the existing network infrastructure are hindering movement to this service-delivery business model. It is the general belief in the industry that
a new network foundation is required. This network foundation is seen as one that will easily adapt to support rapid change, growth, and highly responsive service delivery. What is needed is an intelligent, dynamic, photonic transport layer deployed in support of the service layer.
The photonic-network model divides the network into two domains: service and optical transport. The new architecture is seen as combining the benefits of
photonic switching with advances in DWDM technology. It delivers a multigigabit bit rate capacity and provides wavelength-level traffic-engineered network interfaces to the service platforms. The service platform includes routers, ATM switches, and SONET/SDH add–drop multiplexers, which are redeployed from
the transport layer to the service layer. The service layer is seen as relying completely on the photonic transport layer for the delivery of the necessary bit rate capacity where and when it is needed to peer nodes or to network elements (NEs). The bit rate capacity is provisioned in wavelength granularity rather than in PDH TDM granularities. We expect exponential growth rate of the fiber network; to meet these requirements, rapid provisioning is an integral part of the new architecture. While the first implementations of this model will support error detection, fault isolation, and restoration via SONET, these functions will gradually move
to the optical layer.

No comments:

Post a Comment