Welcome to Scribd, the world's digital library. Read, publish, and share books and documents. See more
Download
Standard view
Full view
of .
Save to My Library
Look up keyword
Like this
3Activity
0 of .
Results for:
No results containing your search query
P. 1
Wavelength Requirements for a Scalable Single-hop WDM Optical Network

Wavelength Requirements for a Scalable Single-hop WDM Optical Network

Ratings: (0)|Views: 164 |Likes:
Published by ijcsis
In this paper, we present a method for designing a passive optical based single-hop wavelength division multiplexing multicast architecture that can achieve a scalable structure and form the basis of a wavelength efficient single-hop WDM network. The proposed architecture minimizes the number of wavelengths required for efficient multicast service and also minimizes tunability requirement of the transceivers. The network size scalability is achieved by adding transmitters and receivers to the designated groups. We show that the proposed system can accommodate large tuning delays and keeps with suitable throughput when the number of wavelength is equal to the number of nodes. We also show that the design can lead to a scalable structure while minimizing the number of wavelengths and tunability of the transceivers required for an efficient multicast service resulting in an improved system throughput and delay performance.
In this paper, we present a method for designing a passive optical based single-hop wavelength division multiplexing multicast architecture that can achieve a scalable structure and form the basis of a wavelength efficient single-hop WDM network. The proposed architecture minimizes the number of wavelengths required for efficient multicast service and also minimizes tunability requirement of the transceivers. The network size scalability is achieved by adding transmitters and receivers to the designated groups. We show that the proposed system can accommodate large tuning delays and keeps with suitable throughput when the number of wavelength is equal to the number of nodes. We also show that the design can lead to a scalable structure while minimizing the number of wavelengths and tunability of the transceivers required for an efficient multicast service resulting in an improved system throughput and delay performance.

More info:

Published by: ijcsis on Sep 05, 2010
Copyright:Attribution Non-commercial

Availability:

Read on Scribd mobile: iPhone, iPad and Android.
download as PDF, TXT or read online from Scribd
See more
See less

03/27/2014

pdf

text

original

 
(IJCSIS) International Journal of Computer Science and Information Security,Vol. 8, No. 5, August 2010
Wavelength Requirements for a Scalable Single-hopWDM Optical Network 
Rabi W Yousif and Borhanuddin Mohd Ali
Faculty of Engineering, Universiti Putra MalaysiaSelangor, Malaysiarabi.habash@gmail.com
Mohd Khazani Abdullah
Significant Technologies Sdn. Bhd.Serdang, Malaysia
Kamaruzzaman Bin Seman
Faculty of Science and TechnologyUniversiti Sains Islam Malaysia, Nilai, Malaysia
Mohd Dani Baba
Faculty of Electrical EngineeringUniversity Teknologi Mara, Shah Alam, Malaysia
 Abstract
—In this paper, we present a method for designing apassive optical based single-hop wavelength division multiplexingmulticast architecture that can achieve a scalable structure andform the basis of a wavelength efficient single-hop WDMnetwork. The proposed architecture minimizes the number of wavelengths required for efficient multicast service and alsominimizes tunability requirement of the transceivers. Thenetwork size scalability is achieved by adding transmitters andreceivers to the designated groups. We show that the proposedsystem can accommodate large tuning delays and keeps withsuitable throughput when the number of wavelength is equal tothe number of nodes. We also show that the design can lead to ascalable structure while minimizing the number of wavelengthsand tunability of the transceivers required for an efficientmulticast service resulting in an improved system throughputand delay performance.
 Keywords-multicasting, wavelength-division multiplexing, single-hop passive optical network
I.
 
I
NTRODUCTION
 In recent years, the Internet traffic has increasedtremendously, because multimedia traffic such as videostreaming service, high resolution images, digital video andaudio conferencing, and business data distribution becomesprevalent in the Internet. Some multimedia applicationsrequire strict quality-of-service (QoS) or multicasting.Current state-of-the-art dense WDM systems are usingnarrow 50-GHz (0.4 nm) channel spacing. In such systems,functions traditionally performed by electronics, such asswitching, signal amplification, etc, are performed in theoptical domain, therefore achieve signal transparency. Thus,the capability for multicast transmission has become a veryimportant requirement for access networks [1, 2].WDM technology has the potential to satisfy the ever-increasing bandwidth needs of network users on a sustainedbasis. Today, optical backbones with a transmission speed of 40 gigabits per second are deployed. This technology isreliable and will meet bandwidth needs for the next few years.However, considering that traffic is growing by 40 percent ayear on average, even 40G networks will have to be expandedto 100G. WDM optical networks can efficiently supportmulticasting since splitting light is inherently easier thancopying data into an electronic buffer. Applications of multicasting include multimedia conferencing, distanceeducation, video distribution, distributed games and manyothers [3, 4]. For cost reasons each node in single–hop WDMnetworks deploys a rather small number of transceivers whichis typically smaller than the number of wavelengths availablefor data transmission/reception. To increase the network efficiency all wavelengths should be used at any given time.In single hop communication, the network must be able toestablish any possible connection in one hop, withoutintermediate relaying or routing. This in effect implies that thenetwork will have to change the connections it supports atdifferent times. Multi-hop networks have the ability tocircumvent the network capacity limitations. Each node isconnected to only a few other nodes, as such only fewwavelengths are required per node. This greatly reduces thewavelength bottleneck.II.
 
K
EY
D
ESIGN
R
EQUIREMENTS
 When designing a WDM network architecture andprotocol, the following key requirements and properties haveto be satisfied [5 - 8]:
 
 
Provide point–to–multipoint connections in order tosupport multicast applications such asvideoconferences and distributed games in aneconomical and bandwidth–efficient manner.
 
Add or remove network nodes in an easy andnondisruptive way without significantly degrading thenetwork performance.
 
Traffic should not have to traverse a large number of intermediate nodes to ensure smaller resourcerequirements and smaller propagation delays.
 
Provide some level of assurance that the servicerequirements for different types of traffic, e.g., fordelay–sensitive, real–time, and interactive applications,are satisfied.
20http://sites.google.com/site/ijcsis/ISSN 1947-5500
 
(IJCSIS) International Journal of Computer Science and Information Security,Vol. 8, No. 5, August 2010
 
Allocate network resources to all nodes which need tosend data. In networks with fair channel access controleach node ready to send data should have an equalopportunity to transmit.
 
To cope with the resulting increased local traffic, metronetworks have to be easily upgradeable. Advancedtechnologies, e.g., tunable transceivers with a widertuning range and a smaller tuning time, have to beincorporated without network service disruption andreconfiguration.III.
 
S
YSTEM
D
ESCRIPTION
 The system under study is based on a broadcast-and-selectWDM architecture consisting of 
 N 
network nodes connectedvia optical fibers to a passive star coupler (PSC) as shown inFigure 1. There are
wavelength channels, where
 
 
 N 
. Thebandwidth of a fiber is divided into
+1 channels, where
≤ 
 N 
. One of the channels,
 
0
, is used as a control channel whichis shared by all nodes. The rest of the channels,
 
1
 ,…,
 
, areused as data channels.
PSC
StationNetworksUsersApair of Optical FibersStationNetworksUsers
ProtocolProcessing
TTFTFR TR
ControlDataDataTo/FromUsers
0
1
-
W
 
(
i
)
0
StationNetworksUsers
1
1
2
PSC: passive star couplerFT/R: fixed transmitter/receiverTT/R: tunable transmitter/receiver
MulticastUnicast
A station possible architecture
 
Figure 1.
 
A broadcast-and-select star-based WDM optical system.
Each node in the network is connected to the PSC by atransmitting and receiving fiber, and each message isaddressed (multicast) to a number of receivers (destination setsize), randomly chosen from the
 N 
network nodes and eachreceiver tunes to one of the wavelengths that has a messageaddressed to it. Also, each node has one fixed transmitter andone fixed receiver in order to access the control channel.Moreover, in order to access data channels, each node has onetunable transmitter and one tunable receiver, so that fullconnectivity can be achieved by tuning transmitters to thedifferent wavelengths.Tuning times are not negligible with respect to the slottime. A centralized network controller allocates slots in aWDM frame according to (long-term) bandwidth requestsissued by users. When
, two or more nodes share onedata channel. Each node is equipped with a buffer in whicharriving data packets are stored. Deploying tunabletransmitters and receivers at each node allows for loadbalancing since traffic between a given pair of nodes can besent on any wavelength. In particular for nonuniform traffic,load balancing increases the channel utilization and improvesthe throughput–delay performance of the network.All stations can communicate with one another. Inaddition, a pair of fixed transceivers and control receiver bothare tuned to the control channel is dedicated for pre-transmission co-ordination. However, communication betweentwo nodes is possible only when the transmitter of the sourcenode and receiver of the destination node are tuned to thesame channel during the period of information transfer.IV.
 
SYSTEM
A
SSUMPTION
 The behavior of the system is characterized by thefollowing assumptions:
 
There are
 N 
network nodes and
wavelengthchannels in the system.
 
Each node has a single–packet buffer, i.e., each nodecan store at most one data packet at any given time.
 
After transmitting a data packet in a given frame thebuffer becomes empty at the end of that frame.
 
Each message is multicast to a set of 
l
receivers where
l
 
 
 
 
 N 
.
 
Whenever the receivers of a multicast group are readyto receive a data packet the source node's transmitter isready to transmit.
 
A packet that arrives at the start of a slot can betransmitted during that slot to any one of the other (
 N 
 
1) nodes with equal probability.
 
A node sends out its control packet in a frame withprobability
 p
, not only for retransmissions but also forfirst–time transmissions.
 
Random selection of a destination node among the (
 N 
 
1) nodes is renewed for each attempt of transmittinga control packet.V.
 
SYSTEM
M
ODEL
 
 A.
 
 Node Structure
The proposed architecture aims to define a minimum groupof network nodes for a local structure, assign a uniquewavelength to a transmitter, and identify, for each transmitter,the minimum set of additional wavelengths needed to achievecommunication with every other node in the local cluster andhence all the nodes in the network.Figure 2 shows the node structure of the system. Eachreceiver is able to tune to all the wavelengths assigned to thetransmitters having direct links to it. Each processor cantransmit data on a fixed number of wavelengths, but canreceive data on a range of wavelengths by dynamically tuningto the wavelength of a transmitting station. All the processorsare synchronized at the optical coupler. The use of the samestructure for both the transmitter and receiver is strategic [6 -9]. This will greatly simplify the coupling of the localstructure. Each node can switch channels (wavelengths) duringexecution by dynamically changing the injection current to thelaser.
21http://sites.google.com/site/ijcsis/ISSN 1947-5500
 
(IJCSIS) International Journal of Computer Science and Information Security,Vol. 8, No. 5, August 2010
N
000
N
010
N
100
N
110
N
001
N
011
N
101
N
111
 
ProtocolProcessing
T-TxF-TxF-Rx T-Rx
ControlDataData
0
1
-
W
 
(
i
)
 
0
Node Structure
 
Figure 2.
 
A network node structure.
Additionally, transmission and reception can be performedon different channels. The single star topology consists of 
n
 inputs, to which one transmitter is connected, and
n
outputs, towhich one receiver is connected.
 B.
 
Connection Establishment and Partitioning
There are basically two ways to achieve connectionsbetween nodes in an optical network, path multiplexing andlink multiplexing. In the first the same wavelength has to beassigned all to the links between source and destination, whilein the second, different wavelengths can be assigned ondifferent links. To achieve single-hop connectivity, awavelength allocation mechanism needed to determine a pathfor a new request. Each transmitter group can have direct linksto exactly two receiver groups. For any transmitter group, thetwo receiver groups that do not have a direct link to it consistof those that contain receivers with the same index notation asone of the transmitters in the transmitter group. For anytransmitter/receiver group, there are two receiver/transmittergroups that can have direct links to it and two others that donot have direct links to it [9]. The transmitter/receiver groupsnot having direct links to the same receiver/transmitter groupsare mutually exclusive. Finally, half of the number of transmitters/receivers can be directly connected to half of thenumber of receivers/transmitters simultaneously. Figure 3shows the connection establishment procedures among thenetwork nodes in order to achieve single-hop communication.
T-Rx
000
T-Rx
010
T-Rx
100
T-Rx
110
T-Rx
001
T-Rx
011
T-Rx
101
T-Rx
111
T-Tx
000
T-Tx
010
T-Tx
100
T-Tx
110
T-Tx
001
T-Tx
011
T-Tx
101
T-Tx
111
T-Tx
111
T-Rx
011
T-Tx
011
T-Tx
001
 
Figure 3.
 
Connection establishment procedures among the nodes.
Partitioning is desirable to design a hierarchical structureusing a cluster-based approach. This becomes quite obvioussince it is intended that the network should be scalable, flexibleand implemented with wavelength division multiplexingtechniques. Each local unit should be autonomous, so thatwavelength reuse can be achieved.
C.
 
Scalable Approach
If there are more than
m
access nodes, where
m
is thedesired number of access nodes representing the regular localstructure, a partition can then be accomplished by defining aminimum set of access nodes as the local structure andapplying the partition mechanism that is explained above toachieve the partition set. The transmitter in a group needs to beplaced according to the partitioning mechanism and same alsofor the receiver. Scalability here has two aspects. First, thetransmitter and receiver of the new access node must bephysically connected to the optical medium and second, theadded access node must be incorporated into the MACmechanism that controls the single hop connections [8-10].To incorporate the added access nodes in the MACmechanism, it requires only modifications for the controlchannel. This means that the number of added nodes mustincrease the number of control slots. To correctly reach eachadded node, all transmitters must be informed about itsreceiver configuration and its address.
 D.
 
Wavelength Allocation
We assign wavelengths such that the tunability for thetransmitters is minimum and optimal while the receivers mustbe able to tune to the maximum number of wavelengths usedin the entire network. Higher number of tuning disrupts thenetwork due to the retuning time. Our goal is to minimize thenumber of tuning so that the reconfiguration does not suspendthe operation of the network.Since computer networks traffic changes rapidly, there is aneed for a good mechanism to change the current situation of the network in terms of wavelength allocation (i.e., the currentwavelength assignment into a new wavelength assignment).However, the number of channels is a limiting factor in aWDM network, and is typically less than the number of nodesin the network. Therefore, more than one receiver is assignedto one channel. This problem is called wavelength assignmentproblem.Wavelength blocking is a major problem with pathmultiplexing. One obvious disadvantage with the link multiplexing is the use of wavelength converters atintermediate nodes to eliminate blocking. This however,increases the cost and complexity of the system.In single-hop communication, all the nodes can reach anyother node directly. This means that the transmitted data arenot passed through any intermediate routing stages and remainin optical form all the way from the source node to thedestination node. In such mode of communication, a lightpathis established before a communication starts and the datatransmission is carried out in a pure circuit-switched manner.
22http://sites.google.com/site/ijcsis/ISSN 1947-5500

Activity (3)

You've already reviewed this. Edit your review.
1 thousand reads
1 hundred reads
jshripati liked this

You're Reading a Free Preview

Download
/*********** DO NOT ALTER ANYTHING BELOW THIS LINE ! ************/ var s_code=s.t();if(s_code)document.write(s_code)//-->