[Jindal, 1(8): Oct.

, 2012] ISSN: 2277-9655

2277

IJESRT
INTERNATIONAL JOURNA
JOURNAL OF ENGINEERING SCIENCES
ENCES & RESEARCH
TECHNOLOGY
Wavelength Division Multiplexing
Neeraj Jindal*1, Shweta Lakhara 2
*1
Department of Electronics
lectronics &
&Communication Engineering, Chartered Institute of Technology,
Abu Road , India
2
Department of Computer Science Engineering
Engineering, Chartered Institute of Technology,
Abu Road , India
[email protected]
Abstract
In Wavelength Division Multiplexing (WDM) networks, there is huge amount of available bandwidth
so each light path can carry a huge amount of traffic and if somehow any failure occurs. It may seriously damage the
end-user applications. Hence fault-tolerance
tolerance becomes an important issue on these types of networks. The light
path which carries traffic during normal operation is called as primary path. In case any failure occurs, the same
traffic is rerouted on an alternative path that may be called as backup pat path.
h. This paper mainly presents the
routing approach for establishing primary and backup paths. In order to establish the primary path, this
approach uses load balancing in which link cost metrics are estimated based on the current load of the links.
l In
backup path setup, the source calculates the blocking probability through the received feedback from the destination
by sending a small fraction of probe packets along the existing paths. It then selects the optimal light path with
the lowestest blocking probability. This approach reduces the blocking probability and latency at the same time
increases the throughput and channel utilization.

Keywords- WDM, fault-tolerance,

tolerance, primary path, backup path

Introduction
Wavelength Division Multiplexing (WDM)
Network
The term wavelength-division
division multiplexing
is commonly applied to an optical carrier (which is
typically described by its wavelength), whereas
frequency-division
division multiplexing (FDM) typically
applies to a radio carrier (which is more often
described by frequency). Since wavelength and
frequency are tied together through a simple
relationship, the two terms actually describe the same
Fig1: Basic point to point
concept. WDM system uses a multiplexer at the
communication configuration using mux-
mux
transmitter to joinoin the signals together and a demux
demultiplexer at the receiver to split them apart [1]. WDM multiplexer (MUX) combines N independent
With the right type of fiber it is possible to have a data streams, each on a unique wavelength, and sends
device that that does both simultaneously, and can them on a fibre.. Demultiplexer (DEMUX) at the
function as an optical add-drop
drop multiplexer (OADM). fibre’s receiving end separates out these data streams.
Fig 2 shows use of star couplers thereby obviating the
need of multiplexer and Demultiplexer [2].

http: // www.ijesrt.com (C) International Journal of Engineering Sciences & Research Technology[466-470]
Technology
[Jindal, 1(8): Oct., 2012] ISSN: 2277-9655

on these networks in contradiction to the

conventional copper links.
Fault tolerance schemes can be broadly classified
into
Path Protection
Path
Restoration
1) Path protection: In path protection,
backup resources are reserved during
connection setup and both primary and
Fig 2: Basic point to point communication using star backup light path are computed before a
couplers. failure occurs. There are two types of
protection schemes: Dedicated and Shared
This type of system uses a star coupler to mix signals
protection.
of different wavelengths and wavelength tunable
filters to extract the information. Although the Dedicated-path protection: In dedicated-
power is decreased by a factor of 1/N, this loss can be path protection (also called 1:1 protection); the
offset with the use of an optical amplifier prior to the resources along a backup path are dedicated for
second star coupler. only one connection and are not shared with the
backup paths for other connections.
WDM systems are popular with telecommunications Shared-path protection: In shared-path
companies because they allow them to expand the protection, the resources along a backup path may
capacity of the network without laying additional be shared with other backup paths. As a result,
optical fiber cable. By using WDM and optical backup channels are multiplexed among different
amplifiers, they can accommodate several failure scenarios, and therefore, shared-path
generations of technology development in their protection is more capacity efficient when
optical infrastructure without overhauling the compared with dedicated-path protection.
backbone 2) Path Restoration: In path restoration,
the source and destination nodes of each
Fault Tolerance in WDM Networks connection traversing the failed
Since each light path can carry a huge mount of link participate in a
traffic, failures in such networks may distributed algorithm to dynamically
seriously damage end-user applications. According discover an end-to-end backup route. If no
to the scale of their effect, failures in all-optical routes are available for a broken
WDM networks can be classified into two connection, then the connection is
categories [3]. One category is a wavelength-level dropped.
failure which impacts the quality of transmission
of each individual light path. The other category is Related Work
a fiber-level failure which affects all the light paths Lei Guo [4] has studied the problem of multiple
on an individual fiber. Since each light path is failures in WDM networks. In order to improve the
expected to operate at a rate of several gigabytes survivable performance he proposed a heuristic
per second, a failure can lead to a severe data loss. algorithm called Shared Multi-sub backup paths
The ability of network to with-stand failures is called Reprovisioning (SMR). The survivable performance
as fault-tolerance. Failures arise due to the node of SMR in multiple failures was considerably
failure or link failure. When a link fails all its improved when compared with the
constituent fibers also fails. All previous algorithm.
the connections which use these fibers are to be Guido Maier, Achille Pattavina, Luigi Barbato,
rerouted and a wavelength will be assigned. The Francesca Cecini and Mario Martinelli [5] have
light path which carries traffic during normal investigated the issue of dynamic connections in
operation is called as primary path. The traffic is WDM networks. It is also loaded with the high-
rerouted on a backup path in case of a failure. priority protected static connections. They have
Optical networks which use the wavelength division compared various routing strategies by discrete event
multiplexing (WDM) and wavelength routing are simulation in terms of blocking probability. Based on
subjected to failures. Fault tolerance becomes an the occupancy cost function they have proposed a
important issue because of the large amount of traffic heuristic algorithm which takes several

http: // www.ijesrt.com (C) International Journal of Engineering Sciences & Research Technology[466-470]
[Jindal, 1(8): Oct., 2012] ISSN: 2277-9655

possible causes of blocking into account. The scheme that searches for disjoint paths, their methods
behavior of their algorithm was tested in well have considerably higher routing success rates which
known case study of mesh networks, with and are shown through the simulation results.
without wavelength conversion. Yufeng Xin, Jing Teng, Gigi Karmous-Edwards,
A.Rajkumar and N.S.Murthy Sharma [6] have George N.Rouskas and Daniel [11] have studied the
proposed a distributed priority important fault management issue which concentrates
based routing algorithm. In order to establish on the fast restoration mechanisms for Optical Burst
the primary and backup light paths they have Switched (OBS) networks. The OBS network
proposed a variety of traffic classes which uses operates under the JIT signaling protocol. The basic
the concept of load balancing. Based on the load routing mechanism is similar to the IP networks,
on the links, their algorithm estimates the cost metric. where every OBS node maintains a local forwarding
The routing of high priority traffic was performed table. The entries in the forwarding table consist of
over the lightly loaded links. Therefore while routing the next hop information for the bursts per
the primary and backup paths, the lightly loaded destination and per FEC (Forward Equivalent Class).
links are chosen instead of choosing the links with Based on looking up the next-hop information in
heavier loads. The load balancing will not reflect their forwarding tables, OBS nodes forwards the
the dynamic load changes because it is used in the coming burst control packets and set up the
routing metric. Michael T. Frederick and Arun K. connections. The connection set up process is
Somani [7] have presented an L+1 fault tolerance signified by the burst forwarding or burst routing.
which is used for the recovery of optical networks Jian Wang, Laxman Sahasrabuddhe and Biswanath
from single link failures without the allocation of Mukherjee [12] have considered the fault-monitoring
valuable system resources. While the approach in functions which are usually provided by the optical-
its simplest form performs well against the existing transmission systems. In order to measure the bit
schemes, the flexibility of L+1 leave many options to error rate in the wavelength channels using SONET
examine possible ways to further increase framing, the B1 bit in the SONET overhead can be
performance. used. Moreover, to detect certain failures like fiber
Muriel M’edard [8] has described that the protection cut in other formatted optical channels, the optical
routes are pre-computed at a single location and thus power loss can be used. Optical-Electrical-Optical
it is centralized. Before the restoration of the traffic, (OEO) conversion is used before each OXC port
some distributed reconfiguration of optical switches because most of the OXCs use electronic switching
is essential. On the other hand, restoration techniques fabric. Therefore, faults can be detected on link-by-
depend upon distributed signaling between nodes or link basis. Both the end nodes of the failed link can
on the allocation of a new path by a central manager. detect the fibre cut for all-optical switches. S
Hongsik Choi, Suresh Subramaniam and Hyeong-Ah
Choi [9] have considered the network survivability Routing Approach
which is a critical requirement in the high-speed In order to establish the primary path, this approach
optical networks. A failure model is considered so uses the concept of load balancing. Given a physical
that any two links in the network may fail in an network with the link costs and the traffic
random order. They have presented three loop back requirements between every source-destination pair,
methods of recovering from double-link failures. then finding a route of the light paths for the network
Only the first two methods require the identification with least congestion, is called as .load balancing. In
of the failed links. But pre-computing the backup this approach, based on the load of the links the cost
paths for the third method is more complex than the metric is estimated. The traffic is routed over the
first two methods. The double link failures are lightly loaded links. Therefore when routing the
tolerated by the heuristic algorithm which primary path, the links with the lighter loads are
pre- computes the backup paths for links. selected instead of links with the heavier loads. Using
Dong–won shin, Edwin K.P.Chong and Howard Jay path restoration backup paths are established. In
Siegel [10] have developed two heuristic multipath backup path setup, the source sends a small fraction
routing schemes for survivable multipath problem of probe packets along the existing paths. For a
called CPMR (Conditional Penalization Multipath higher burst arrival rate, the fraction of
Routing) and SPMR (Successive Penalization traffic probing will be lower. For a slow changing
Multipath Routing). traffic, the period of update will be higher resulting in
Their schemes use “link penalization” an even smaller fraction. The source edge can
methods to control (but not prohibit) link-sharing monitor and identify the requests that are rejected
to deal with the difficulties caused by the link at the network based on receiving the
sharing. When compared with the routing PACKS/NACKS from the destinations [13]. Thus the

http: // www.ijesrt.com (C) International Journal of Engineering Sciences & Research Technology[466-470]
[Jindal, 1(8): Oct., 2012] ISSN: 2277-9655

source can easily calculate the blocking

probability through the monitored results from the Plost is the number of packets dropped
probe packets. The ingress edge node selects the and
optimal light path with the lowest blocking Psent is the number of packets
probability based on the measured blocking sent.
probabilities and forwards the data through this
optimal light path. On the other hand, it keeps 4) Similarly for all the paths Rj, the source
probing the sub-optimal path for their current S calculates their blocking probabilities BP j
blocking probability. based on the PACK and NACK feedback
Computing Primary Path received from the destination D.
The link cost function for primary path
computation is designed based on the following 5) Now sort the paths {Rj, j=1, 2…k} in
steps: ascending order of BPj values.
• For each link Lj, j = 1, 2, 3… calculate
the load index of the link Lj as 6) The paths which are having less blocking
Load (LI) = Cf / Cn (1) probabilities BP1, BP2, BP3 … are selected as
Where Cf gives the number of free channels in backup paths. If there is any sudden or unexpected
that link and Cn is the total no. of channels in that failure occurs in the primary path, traffic can be
link. rerouted through these backup paths.
• The link cost function Cost (Lj) is then At the same time, for their blocking probability it
defined as keeps searching the sub optimal paths. Because of
Cost (Lj) = 1- Load (LI), if Load (LI) >LT this, we can able to jump quickly to a new path
= 1+ Load (LI), if Load (LI) > 0 when blocking probability of the current path
And Load (LI) < = increases. This occurrence is quite obvious in IP
LT networks where the traffic patterns may vary
=∞, if Load (LI) = 0 (2) significantly.
Where LT shows the load threshold. By sending a small fraction of traffic for probing,
After we assign each link a cost using the above the aggregated throughput is reduced. However by
formula, Dijkstra’s shortest path algorithm is then finding a new optimal path quickly this reduction is
used to compute the least-cost path as the primary compensated. The value of small fraction depends
path. If the least-cost path has a cost of infinity, then upon the sample size for accurately calculating the
the demand is blocked; otherwise a backup path is blocking probability. For a higher burst arrival rate
computed using the method given in the next the fraction of traffic for probing is very low.
subsection.
Computing Backup Path Conclusion
Let the number of paths between source S and In this paper reliable routing approach for
destination D are n. In this approach, small fractions establishing primary and backup paths in optical
of probe packets are sent by the non-optimal paths WDM networks has been studied. In order to
such that these paths are selected very rarely. The establish the primary path, this approach uses
probe packets contain sequence numbers to identify load balancing in which link cost metrics are
the packets. estimated based on the current load of the links. The
1) Let Pj, j=1, 2….k be the set of probe packets traffic is routed over the lightly loaded links.
sent on the paths Rj, j=1, 2…k. Therefore the links with the lighter loads are
2) On receiving the probes packets, the destination selected instead of links with the heavier loads. In
D for the path Rj, send an PACK packet to the backup path setup, the source sends a small fraction
source, for each packet correctly received. The of probe packets along the existing paths. It can
missing or dropped packets can be identified using monitor and identify the requests that are rejected at
the sequence numbers of the received packets. the network based on the received positive and
For each dropped packet, it sends a NACK packet negative feedback from the destinations. The source
to the source. then calculates the blocking probability from the
3) For the path Rj, the source calculates the received feedback and selects the optimal light path
blocking probability BPj such that with the lowest blocking probability.
BPj = Plost / Psent (3)
References

http: // www.ijesrt.com (C) International Journal of Engineering Sciences & Research Technology[466-470]
[Jindal, 1(8): Oct., 2012] ISSN: 2277-9655

[1] L. Kazovsky, S. Benedetto, and A. International Conference.

Wilner, Optical Fiber Communication [11] Yufeng Xin, Jing Teng
Systems, Artech House, ISBN 0-89006- Gigi Karmous-
756-2 Edwards,GeorgeN.Rouskas and
[2] A. Ghatak, and K. Thyagarajan, Daniel Stevenson, “Fault
Introduction to Fiber Optics, Cambridge Management with Fast Restoration for
Press, ISBN 0-521-57785-3 Optical Burst Switched Networks”,
[3] S.Ramamurthy, Laxman Sahasrabuddhe, IEEE,2004
and Biswanath Mukherjee, "Survivable [12] Jian Wang,Laxman Sahasrabuddhe and
WDM Mesh Networks", Journal of light Biswanath Mukherjee, “Fault Monitoring
wave technology, vol. 21, no. 4, April and Restoration in Optical WDM
2003. Networks”, National Fiber Optic Engineers
[4] Lei Guo, "Heuristic Survivable Routing Conference NFOEC 2002
Algorithm for Multiple Failures in WDM [13] G.Ramesh S.SundaraVadivelu,” A
Networks", in proc. of 2nd Reliable and Fault- Tolerant Routing
IEEE/IFIP International Workshop on for Optical WDM Networks”
Broadband Convergence Networks, pp: 1- (IJCSIS) International Journal of
5, 21 May 2007, Doi: Computer Science and Information
10.1109/BCN.2007.372750. Security, Vol 6, No 2, 2009
[5] Guido Maier, Achille Pattavina, Luigi
Barbato, Francesca Cecini and Mario
Martinelli, "Routing Algorithms in
WDM Networks under Mixed Static and
Dynamic Lambda-Traffic", Journal on
Photonic Network Communications, vol.
8, no. 1, pp: 69- 87,
June2004,Doi:10.1023/B:PNET.000003161
9.18955.4
[6] Rajkumar and Dr.N.S.Murthy Sharma, "A
Distributed Priority Based Routing
Algorithm for Dynamic Traffic in
Survivable WDM Networks", International
Journal of Computer Science and Network
Security, vol.8, no.11, November 2008.
[7] Michael T. Frederick and Arun K. Somani,
“A single-fault recovery strategy for
optical Networks using sub graph routing”,
The International Journal of Computer and
Telecommunications Networking, Volume
50, Issue 2 (February 2006)
[8] MurielM´edard, “Network Reliability and
Fault Tolerance”, IEEE Transactions on
Volume 50, Issue 1, Page(s):85 – 91, Mar
2001
[9] Hongsik Choi, Suresh Subramaniam, and
Hyeong-Ah Choi, “On Double-Link
Failure Recovery in WDM Optical
Networks”, INFOCOM Volume: 2, on
page(s): 808- 816 vol.2, 2002
[10] Dong-won Shin, Edwin K.P.Chong and
Howard Jay Siegel, “Survivable
Multipath Routing Using Link
Penalization”, Computing and
Communications, 2004 IEEE

http: // www.ijesrt.com (C) International Journal of Engineering Sciences & Research Technology[466-470]