WDM Optical Communication Networks

Mohamed Abdirahman (0525947) Department of Electrical and Computer Engineering International Islamic University Malaysia

AbstractThe rapid evolution of telecommunications networks is always driven by ever-increasing user demands for new applications as well as continuous advances in enabling technologies. The emergence of time-critical multimedia applications, such as internet, telephony, video conferencing and interactive gaming, requires also a large amount of network bandwidth. To meet the unprecedented demand for bandwidth, an optical multiplexing technology, Wavelenghtdivision Multiplexing (WDM) is introducted. In this paper, Princibles of WDM is presented, WDM network architecture and recent trends is discussed. Keywords: WDM, lightpath, photonic switch

technology in its own right. The telecommunications industry, network service providers, and research communities worldwide are paying close attention [1]. Thus, the basic premise of the subject of an optical wavelength division multiplexing (WDM) network is that as more and more users start to use data networks, and as their usage patterns evolve to include more and more bandwidth-intensive networking applications such as data browsing on worldwide web, java applications, video conferencing and others [2]. In these areas, there is an acute need for a very high-bandwidth transport networks facilities, whose capabilities are much beyond those that current high-speed networks can provide. The problem is that there is no just enough bandwidth in our networks today to support the exponential growth in user traffic. Voice traffic continues to experience a healthy growth of approximately 7% per year [2]. However, the data-traffic is attracting people attention. Fig. 1 shows the past and projected future growth of data and voice traffic. Duo to huge growth of data traffic, the WDM (Wavelength Division Multiplexing) optical networking sysem has been developed to meet the tremendous traffic [3]. The key is designing optical communication Networks is to exploit the fibers huge bandwidth to introduce concurrency among multiple users transmissions into the network architectures and protocols.

I.

INTRODUCTION

We are moving towards a society which needs that we have to access information at our fingertips when we need it, where we need it, and in whatever format. As a result of its huge transmission capacity and countless advantages, fiber optics has fostered a bandwidth revolution, addressing the constantly growing demand for increased bandwidth. Within this growing quickly area, Wavelength Division Multiplexing (WDM) has emerged as a breakthrough technology for exploiting the capacity of optical fibers. Today, WDM is deployed by many network providers for point-to-point transmission. However, there is strong momentum to develop it as a full-fledged networking

Fig. 1: the past and projected future growth of data and voice traffic [2]

In optical communication network, this concurrency may be provided according to either wavelength division multiplexing, timedivision multiplexing (TDM) or code-division multiplexing (CDM). Although TDM and CDM are somewhat futuristic technologies, but the TDM bit rate and the CDM chip rate may be much higher than electronic processing speed. Thus, TDM and CDM are less attractive than WDM, since it has no such requirement.

Fig 2. The low attenuation regions of optical fiber [2]

II.

WAVELENGHT-DIVISION MULTIPLEXING (WDM)

Wavelength-division multiplexing is an approach that can exploit the huge optoelectronic bandwidth mismatch by requiring that each end users equipment operate at electronic rate, but multiple WDM channels from different end-users may be multiplexed on the same fiber [2]. Under WDM, the optical transmission spectrum is divided into a number of non-overlapping wavelength bands, at each wavelength supporting a single communication channel operating at whatever rate one desires. Thus, by allowing multiple WDM channel to coexist on a single fiber, one can tap into huge fiber bandwidth, with the corresponding challenges being the design of appropriate network architectures.

Conceptually, WDM is similar to frequency division multiplexing (FDM) that has already been used in radio communication systems for over a century. The basic principle is to divide the huge bandwidth of an optical fiber into a number of non-overlapping channels and transmit multiple optical signals simultaneously and independently in different optical channels over a single fiber, each signal being carried by a single wavelength. As shown in figure 2, there are two lowattenuation areas. One is centered at 1300 nanometers (nm) and the other at 1500 nm. Both areas has a range of about 200 nm with attenuation loss less than 0.5 dB/km. theoretically, these two areas can provide a total amount of 50 Terabit per second (Tbps) low attenuation transmission bandwidth [1]. However, because the maximum rate at which an end device can access an optical channel is limited to electronic processing speed, it is technically impossible to take advantage of all bandwidth of an optical fiber by using a single optical channel.

Figure 4 shows a block diagram of a basic WDM transmission system. The network medium can be a simple fiber link, a passive start coupler or any type of optical network. The transmitter consists of a laser and a modulator. The laser is the light source, which generates an optical carrier signal at either a fixed wavelength or a tunable wavelength [4]. In the modulator, the carrier signal is modulated by an electronic signal and is sent to the multiplexer. The multiplexer combines multiple optical signals of different wavelengths at its input ports into a single optical signal, which is transmitted to

common output port or optical fiber. The demultiplexer uses optical fibers to separate the optical signal received on the input port into multiple optical signals n different wavelengths, which are sent in to receivers. The receiver consists of a detector (photodiode) that converts the optical signal to an electronic signal. The optical amplifiers are used to maintain the power strength of an optical signal at appropriate locations is the transmission system.

Fig 3: A four-channel point-to-point WDM transmission system [4]

III.

WDM NETWORK ARCHITECURES

WDM networks can be classified into tow broad categories: broadcast-and-select WDM networks and wavelength-routed WDM networks. A. Broadcast-and-select WDM networks A WDM network that shares a common transmission medium and employs a simple broadcasting mechanism for transmitting and

receiving optical signals between network nodes is referred to as a broadcast-and-select WDM network. The most popular topologies for a broadcast-and-select WDM network are the star topology and the bus topology. In star topology, a number of nodes are connected to a passive star coupler by WDM fiber links as shown in figure 4. The information streams from multiple sources are optically combined by the star and the signal power of each stream is equally split and forwarded to all of the nodes on their receive fibers [2].

The star coupler receives and combines all the signals and broadcast them to all nodes in the network. At the receivers node, it is tuned to only one of the wavelengths using optical filter; hence, it can receive the information stream

connect (OXC) or wavelength cross-connect (WXC) to route and switch different wavelengths from an input to an output port [1]. Each fiber link operates in WDM AND supports certain number of optical channels or wavelengths. An access node can be connected to a routing node, which is used as an interface between the optical networks and electronic client networks. A wavelength-routed WDM network is a circuit-switched network in which a pair of network nodes communicates through and endto-end optical connection that may consist of one or more all-optical connection called lightpaths. A lightpath is a unidirectional alloptical connection between a pair of network nodes, which may span multiple links and use one or multiple wavelengths without undergoing any conversion at each intermediate node [5]. Two lightpaths cannot share the same wavelength on a common fiber link, which is referred to as the wavelength-distinct constraint. However, two lightpaths can use the same wavelength on different fiber links. This is referred to as wavelength- reuse property. In the absence of any wavelength conversion capability, a lightpath must use the same wavelength on all the links it spans, which is known as the wavelength-continuity constraint. This constraint is unique to wavelength-routed WDM networks and makes such networks different from conventional circuit-switched networks [6]. Because of this constraint, network performance in terms of wavelength utilization and blocking probability would be largely degraded. For this reason, it is desirable to eliminate the wavelength- continuity constraint in order to improve network performance. This is achieved by deploying wavelength converters at network nodes to provide wavelength conversion capability in the network.

Fig 4: A passive-star-based local optical WDM network

The advantage of broadcast-and-select WDM network lies in its simplicity and broadcasting capability. However, this type of network needs a large number of wavelengths because the wavelengths cannot be reused in the network. So this type of network cannot span a long distance and is most suitable for deployment in local area networks (LANs) or metropolitan area network (MANs) B. Wavelength-Routed WDM Networks A wavelength-Routed (wide-area) optical WDM network is shown in fig. 5. The network consists of a photonic switching fabric, comprising active switches connected by fiber links to form an arbitrary physical topology. Each end user is connected to an active switch via a fiber link. The combination of network user and its corresponding switch is referred to as a network node [2]. An access node employs a set of transmitters and receivers for transmitting signals fiber links and receives from it and an optical cross-

A wavelength converter is an optical device that transforms a signal from one wavelength to a different wavelength. If no wavelength converter is used, a lightpath established between source-destination node-pair always consists of the channels of the same wavelength. When this wavelength continuity cannot be satisfied, the lightpath is blocked [6]. All or limited number of nodes may employ wavelength converters to maintain all-optical transmission along the lightpath and hence to

ensure improved performance of WDM optical networks. The improved performance due to incorporation of wavelength converters is restricted and endorsed largely by network topology as well as traffic configuration. In a small network with sufficient wavelength channels and under light traffic, it does not need any wavelength converters at all, as the probability of wavelength conflict is small [6].

Fig. 5: A wavelength-routed (wide-area) optical WDM network.

A wavelength-routed WDM network can be classified in to the following different categories [1]:     Non-reconfigurable an reconfigurable Single hop and multiple hop Wavelength selective and wavelength convertible Single fiber and multiple fiber

Non-reconfigurable networks A non-reconfigurable network is also referred to as a static network , in which no switches are employed in the routing nodes. The network configuration does not change once the network is configured. The set of lightpaths established in the network are fixed and are determined at the time of network design. This type of network can consist entirely of passive optical devices, such as

passive couplers, multiplexers and demultiplexers (fig 6). Therefore, it is more reliable than reconfigurable networks. Reconfigurable networks A reconfigurable network employs switches in the routing nodes, which can route a wavelength on input port to a wavelength on different output port. Accordingly, set of lightpaths established in the network can be dynamically reconfigured by changing the statuses of the switches in the routing nodes. Single-hop networks A single-hop network is also referred to as an all-optical network, in which data traffic is transferred from one node to another throught a single lightpath without undergoing any intermediate optic-electric (O/E) and electrooptic (O/E) conversion. Accordingly, a signlehop network can provide the best quality of service to its end users. Because of the limit in network resources, a single-hop network may not be able to achieve the maximum network throughput. Multiple-hop networks In a multiple-ho network, an end-to-end connection consists of a tandem of lighpaths. At the end of each lightpath, data traffic must go through O/E and E/O conversion, which could introduce additional conversion delay. Because of the O/E and E/O conversion, congestion may also occur in a routing node, which may further introduce queuing delay. For this reason, a multiple-ho network may not be able to provide the best quality of service to its end users. However, it provides a higher-level flexibility and thus provides better network performance in terms of resource utilization, blocking probability, and network throughput.

Wavelength-selective networks In a wavelength-selective network, no wavelength conversion capability is provided at any routing node. A lightpath must use the same wavelength on all the links it traverses, which is referred to as the wavelengthcontinuity constraint. Because of this constraint, the network performance in therms of the wavelength utilization, blocking probability and network throughput would be largely affected. Wavelength-convertible networks Wavelength converters are deployed in the routing nodes to overcome the wavelength continuity constraint. A wavelength converter is an optical device that can convert one wavelength to another wavelength in the optical domain. With wavelength converters, the wavelength-continuity constraint can be alleviated or even eliminated. As result, a wavelength-convertible network can achieve better wavelength utilization and smaller blocking probability than a wavelengthselective network Single-fiber networks In a single fiber networks, there is only one fiber pair between each pair of network nodes. The number of available wavelengths on each fiber link limits network performance. Multiple-fiber networks A multiple fiber networks uses multiple fiber pairs between each pair of networks nodes to provide more wavelengths and larger transmission bandwidth. As result, network performance can be significantly improved. In addition, a multiple-fiber network with no wavelength conversion is equivalent to a single-fiber network with limited wavelength conversion.

IV. RECENT TRENDS IN WDM RESEARCH Recent research interests in WDM networks include network control and management, fault management, multicasting, physicallayer issues, IP over WDM, traffic grooming, and optical packet switching, just to name a few. We briefly examine these topics below. a) Network Control and Management

In a wavelength-routed WDM network, a control mechanism is needed to set up and take down all-optical connections (i.e., lightpaths) [2] Upon the arrival of a connection request, this mechanism must be able to select a route, assign a wavelength to the connection, and configure the appropriate optical switches in the network. The mechanism must also be able to provide updates to reflect which wavelengths are

Fig. 6: wavelength add/drop multiplexer There are two distributed network control management schemes, which have been examined in the literature. The first approach is proposed in [7], and it is referred to it as the link-state approach because it routes connections in a link-state fashion. It is proposed the second approach in [8], and they refer to it as the distributed routing approach because it utilizes the distributed BellmanFord routing algorithm. they describe the two approaches below. In the link-state approach (which may be implemented using the open shortest path first (OSPF)

currently being used on each fiber link so that nodes may make informed routing decisions. This control mechanism can either be centralized or distributed. Distributed systems are usually more robust than centralized systems; so they are generally more preferred. The objectives of various research efforts on this subject are to minimize 1) the blocking probability of connection requests, 2) the connection setup delays, and 3) the bandwidth used for control messages; as well as to maximize the scalability of such networks [2].

algorithm), each node maintains the complete network topology, including information on which wavelengths are in use on each fiber link. Upon the arrival of a connection request, a node utilizes the topology information to select a route and a wavelength. Once the route and wavelength are selected, the node attempts to reserve the selected wavelength along each fiber link on the route by sending reservation requests to each node in the route. If an intermediate node is able to reserve the wavelength on the appropriate link, it sends an acknowledgment directly back to the source node. If all of the reservations are successful, then the source sends a SETUP message to each of the nodes. The appropriate switches are then configured at each node, and the connection is established. If even one of the reservations is not successful, then the call is blocked and the source node sends a TAKEDOWN message to each node on the route in order to release the reserved resources. When a connection is established or torn down, each node involved in the connection broadcasts a topology-update message which indicates any changes in the status of wavelengths being used on the nodes outgoing links. In the distributed-routing approach, routes are selected in a distributed fashion without knowledge of the overall network topology. Each node maintains a routing table which specifies the next hop and the cost associated with the shortest path to each destination on a given wavelength. The cost may reflect hop counts or actual fiber-link distances. The routing table is established by employing a distributed BellmanFord algorithm [2]. In the distributed-routing approach, upon receiving a connection request, a node will choose the wavelength, which results in the shortest distance to the destination, and it will

forward the connection request to the next node in the path. Thus, the connection request is routed one hop at a time, with each node along the route independently selecting the next hop based on routing information, and reserving the appropriate wavelength on the selected link. Once the request reaches the destination node, the destination node sends an acknowledgment back to the source node along the reverse path (i.e., it sends the acknowledgment back to the node from which it received the connection request). Upon receiving the ACK, each node along the reverse path conFigs. Its wavelength-routing switch. The source node begins transmitting data after it receives the acknowledgment. If a node along the path is unable to reserve the desired wavelength on a link, it will send a negative acknowledgment back to the source along the reverse path. The nodes on the reverse path will release the reserved wavelengths as they receive the negative acknowledgment. The source node may then reattempt the connection on a different wavelength. If the source node is unable to establish the connection on any wavelength, the call is blocked. Once a connection is established, each node along the route sends to each of its neighbors an update message reflecting the status of the newly occupied link and wavelength. Each node receiving an update message may then update its routing table. Similar updates occur when a connection is taken down [2]. b) Fault Management In a wavelength-routed WDM network (as well as in other networks), the failure of a network element (e.g., fiber link, crossconnect, etc.) may cause the failure of several optical channels, thereby leading to large data (and revenue) losses.

There are several approaches to ensure fiber-network survivability. Survivable network architectures are based either on dedicating backup resources in advance, or on dynamic restoration. In dedicated-resource survivability (which includes automatic protection switching and self-healing rings), the disrupted network service is restored by utilizing the dedicated network resources. In dynamic restoration, the spare-capacity available within the network is utilized for restoring services affected by a failure. Generally, dynamic restoration schemes are more efficient in utilizing capacity due to the multiplexing of the spare-capacity requirements, and they provide resilience against different kinds of failures, while dedicated restoration schemes have a faster restoration time and provide guarantees on the restoration ability. There are three approaches to protect against fiber-link failures in an optical network: a dedicated-resource survivability approach called 1 1 protection, and two dynamic approaches called link restoration and path restoration [2]. c) Traffic Grooming in WDM ring networks

having an ADM on every wavelength at every node, it may be possible to have some nodes on some wavelength where no add/drop is needed on any time slot; thus, the total number of ADMs in the networks (and hence the network cost) can be reduced. Under the static traffic pattern, the savings can be maximized by carefully packing the virtual connections into wavelengths [2]. V. CONCLUSION This paper provides an overview of the research and development areas of optical networking. A brief discussion of WDM technologies and WDM network architecture is presented in this project.

REFERENCES

[1]

Zheng, J. & Mouftah, H. (2004). Optical WDM Netowks: Concepts and Design Principles. IEEE Press. B. Mukherjee, WDM Optical Communications networks: Progress and Challenges, IEEE journal on selected areas in communications, vol 18, No. 10 October 2000 K. W. Kim, D. J. ke, H. yae Design and Implementation of Performance Management Function in 1.6T WDM optical transmission system IEEE 2010 Laude, J. P. (1993) Wavelength-division Multiplexing Prentice Hall

[2]

Traffic grooming is a term used to describe how different (low-speed) traffic streams are packed into higher-speed streams [9]. In a WDM/SONET ring network, each wavelength can carry several lower-rate traffic streams in TDM fashion. The traffic demand, which is an integer multiple of the timeslot capacity, between any two nodes is established on several TDM virtual connections. A virtual connection needs to be added and dropped only at the two end nodes of the connection; as a result, the electronic add/drop multiplexors (ADMs) at intermediate nodes (if there are any) will electronically bypass this timeslot. Instead of

[3]

[4]

[6]

Roy, K.a simple approach for optimal allocation of wavelength converters in WDM optical networks IEEE 2007 R. Ramaswami and A. Segall.Distributed network control for optical networks

[7]

IEEE/ACM Trans. Networking, Vol 5, PP. 936-943,Dec, 2009 [8] H. Zang, Sahasrabudhe, J.P. Jue. S. Ramamurthy, and B. Mukherjee, Connection management for wavelengthouted WDM networks, in Proc. IEEE Globlecom99, Rio de Janeiro. Brazil, Dec 1999,pp. 1428-1423. C. Xin, Traffic Grooming in mesh WDM optical networks- performance analysis IEEE Journal on Selected areas in communications vol. 22, No.9, November 2004 S. Ramamurthy and B. Mukherjee, Fixed alternate routing and wavelength conversion in wavelength-routed optical networks, in Proc. IEEE Globecom98, Sydney, Australia, Nov. 2005

[9]

[10]