MMINIMUM RESOURCE CONSUMPTIONS ROUTING FOR OPTICAL NETWORKS

International Journal of Electronics and Communication Engineering Research and Development
(IJECERD), ISSN 2248-9525(Print), ISSN- 2248-9533 (Online) Volume 4, Number 2, April-June (2014)
64
MINIMUM RESOURCE CONSUMPTIONS ROUTING FOR OPTICAL
NETWORKS
1
Shravan S K, 2
Triveni C L, 3
Dr. P.C. Srikanth, 4
Taresh B.M
1,2,3,4
Department of ECE,
1,2,3,4
Malnad College of Engineering
1,2,3,4
Visvesvaraya Technological University (VTU), Belgaum, Karnataka, India
ABSTRACT
The problem of determining primary and backup paths for survivable optical WDM
networks is considered. Results of various available routing techniques that try to minimize
the combined cost of primary and the backup path are analyzed for the effects on network
parameters such as mean load, variance of the load on route, number of converters required
by the route and the length of the route. The route cost is modelled such a way that it is
extensible to include any new parameter and vary their relative importance. The efficiency of
such wavelength routed networks has been proved to improve for certain parameters, such as
reduction in blocking probability and number of converters required for desired performance.
The routing is enhanced to analyse effect on network parameters for all node full range
converters, limited number full converters, reserved primary and back up wavelengths and
with no such reservation.
Keywords: Wavelength Division Multiplexing (WDM), Back up path, Wavelength
convertor.
I. INTRODUCTION
Recent advances in optical switching and in particular wavelength division
multiplexing (WDM) have enabled next generation networks to be able to operate at several
Terabits per second. Wavelength routed networks consist of optical switching nodes
interconnected by one or more fibre links. A network failure (link or node failure) results in
huge data loss due to enormous bandwidth of fibre. Survivability of optical networks can be
realized either by pro-active or by reactive mechanism. In protection based scheme each
incoming request is provided with a primary path and a link disjoint back up path at set up
time where as in restoration based scheme an alternate path is determined only after the
IJECERD
© PRJ
PUBLICATION
International Journal of Electronics and Communication
Engineering Research and Development
(IJECERD)
ISSN 2248– 9525 (Print)
ISSN 2248 –9533 (Online)
Volume 4, Number 2, April- June (2014), pp. 64-72
© PRJ Publication, http://www.prjpublication.com/IJECERD.asp

65
failure. The key challenge is to devise a strategy to determine the primary and the backup
path such that the network throughput is maximized and the resource consumption is
minimized [1].
The above problem is addressed by either separate path selection (SPS) or joint path
selection (JPS). In SPS the algorithm first selects the path with minimum cost and then link
disjoint least cost backup path. Least loaded routing (LLR) selects path having more free
wavelengths, Conversion free primary routing (CFPR) minimizes the number of converters in
the primary path. In JPS algorithm tries to minimize combined cost of primary and the
backup paths [2]. Routing in WDM networks is accomplished by two approaches: 1) static
and 2) dynamic.
In static routing approach the routes between the node pairs are fixed, i.e.., the routes
do not change with the network status. Common mechanisms include fixed path routing and
alternate path routing. Fixed path routing is supposed to have a weaker performance as only
path is established between nodes. A connection requested is blocked if no wavelength is
available on that path. Alternate path routing, in which more than one candidate path are
provided between nodes for a connection request, improves the network performance
significantly. However the candidate path and their preferences are pre-determined without
considering the change in the current network status.
In dynamic routing, the routes are dynamically selected according to the current
network status as in Least Congestion Routing and WDM aware link weight functions. The
results show that the blocking probability of least congestion routing is one or two orders less
in magnitude than that of the alternate path routing in mesh-torus networks [1].
A fundamental property of light-path is its continuity. A connection must be assigned
the same wavelength on all hops of its path. This can lead to blocking of connection request
when same wavelength is not available on given path. Wavelength converters provide
solution to this problem by shifting the wavelength of an incoming signal to another
wavelength. This reduces the blocking due to wavelength unavailability on a hop, as any one
of the wavelengths being available suffices the cause [2]. All optical wavelength converters
are costly and the design should aim at minimizing the total number of converters while
achieving good blocking performance.
Wavelength-division multiplexing (WDM) is a method of combining multiple signals
of laser beams at various infrared (IR) wavelengths for transmission along fiber optic media.
Each laser is modulated by an independent set of signals. Wavelength-sensitive filters, the IR
analog of visible-light color filters, are used at the receiving end.
WDM is similar to frequency-division multiplexing (FDM). But instead of occupying
place at radio frequencies (RF), WDM is exploits the IR portion of the electromagnetic
spectrum. Each IR channel carries several RF signals combined by means of FDM or time-
division multiplexing (TDM). Using FDM or TDM in each IR channel in combination with
WDM or several IR channels, data in different formats and at different speeds can be
transmitted simultaneously on a single fiber.
II. BACKGROUND AND RELATED WORK
Survivability in optical networks can be achieved using protection (proactive) or
restoration (reactive) mechanisms [13] using two different approaches, namely:
1) Separate path selection (SPS)
2) Joint path selection (JPS)

66
1) Separate Path Selection
This is the typical approach where the algorithm first selects the path with the
minimum cost as the primary path and then selects a link-disjoint path with least cost as the
backup. The algorithms proposed in the literature differ in the way in which they model the
cost of the paths, as described in the subsequent sections.
a) Shortest path algorithm: This algorithm models the route cost as the number of hops
between the source node and the destination node. This is a static routing approach. The
backup for this shortest path is then computed as follows. The links that comprise the primary
path are removed and the new shortest path is found. This path will be link-disjoint with the
primary. The advantages of this method are its simplicity and its ease of implementation.
However, it is not the most efficient algorithm and many studies [12] have shown that a
dynamic algorithm that takes into account the network state performs better.
b) Least loaded routing (LLR) algorithm: In this dynamic routing approach, the link load
is taken into account when determining a new path. When a connection request arrives, the
path with the least average load is chosen as the primary path. The average load is calculated
as the average of the number of wavelengths that are currently reserved in each link of the
path. Once the primary path is chosen, the links corresponding to the primary path are
removed and the new least loaded path between the source and destination is chosen as the
backup path. This algorithm aims to distribute the load equally among all the links of the
network. The execution overhead of this approach is higher compared to the shortest path.
This algorithm is shown in [4] to improve the blocking performance. However, this algorithm
has not been studied in the presence of limited wavelength converters per node. Further, it
has been shown that the path produced using this mechanism tends to be longer thereby
utilizing more resources.
c) Conversion free primary routing (CFPR):
This method [6] uses a dynamic routing approach that tries to minimize the use of
converters in the primary path. The objective is to eliminate conversion delay, possible signal
degradation and also to reduce the number of converters needed in the network, thereby
reducing the cost. At the arrival of a connection request, the primary path is determined as the
path that requires no converters. The backup path is then determined as the shortest path that
is link-disjoint with this primary path.
2) Joint Path Selection: Here, the algorithm tries to optimize the combined cost of the
primary and the backup paths. SPS approaches that take into account the current network
state (e.g., LLR and CFPR) were seen to perform better than the basic hop count (HC)
scheme. However, a technique that tries to optimize the combined cost of the primary and
backup paths has been shown to perform even better [14], [15].
III. IMPLEMENT ALGORITHM
A Network Model
The network model that is used here is circuit switched network. There are N nodes
and L links, the number of wavelengths between two connected nodes is W. The dynamic
routing is made used. For every OD pair a primary path is established, which the shortest path
considered the route cost. The number of back up paths are chosen by the user for the purpose

67
an under selection or over selection of this number either increases blocking probability or
increases the complexity. A primary path or a back up path between certain pair could be
used as either primary path or a back up path for some other pair, making maximum use of
the route capacity to carry data over the channel where every signal is identified by a
wavelength.
B RWA algorithm
The objective of dynamic RWA is usually to minimize the blocking probability or in
other words, to maximize the number of connections that are established in the network at
any time. The difference in performance mainly lies in different routing strategies. Among all
wavelength assignment algorithms, First-Fit algorithm is the most well-known. Wavelength
continuity can cause the wastage of wavelength resources, resulting in low resource
utilization. One possible way to overcome this problem is to use wavelength converters at a
routing node. The care is taken to minimize the number of converters by first maintaining
light path continuity. When the continuity fails at some node the presence of converters at
that node establishes lightpath by assigning the signal the nearest possible wavelength. If
there is no converter at the node where continuity fails, the request is routed through the next
back up path till a success in routing is achieved or the request is dropped for no wavelength
availability over all the set of routes for the OD pair in concern.
C Route Cost Model
Traffic at each node is the measure of requests arrives at a node. The requests are
assumed to arrive following poisons variable with certain mean. The traffic matrix entries
TM(i,j) correspond to traffic demand on the link between node i and node j, directed from
node i to node j.
Link load: this represents the load on a link which is defined as follows
The load on link is the sum of the primary load on that link and part of load from
other OD paths using this same link in their back ups.
Hop counts Hp: this represents the number of hops between the OD pair throughout
the chosen path.
Mean load: this represents the mean load throughout the path.
Variance of the load on path= In case a decision between the two path selection is
trivial due to same mean over paths, the variance is considered. The one with lower variance
is selected.

68
D. MRCR Algorithm
When a connection request arrives at source s to destination d following steps are executed
1. Chose the primary path and check if the same wavelength as at s is free at next link of
the path.
2. If the continuity exists forward the request updating the status at the node until the
next node is d.
3. If same wavelength is not available then check for continuity in subsequent back up
paths and repeat step 2.
4. If the continuity for the wavelength is not found after all paths in step then chose e the
primary path and check if any wavelength is free at next link of the path.
5. If a free wavelength is available then convert the signal transmission to next nearest
wavelength (first fit). And forward the request to next link by updating the node status
until the next node is d.
6. If a free wavelength is not available then check in subsequent back up paths for a free
wavelength and repeat step5.
7. Failure to find a free wavelength in step 6 results in dropping the connection request.
IV. RESULTS AND PERFORMANCE
The simulation were performed using MATLAB 2011b, considering a 30 node
network and a 14 node NSFNET. The results are shown for a 30 node network in this paper.
Traffic is modelled to be poissions distribution event with respect to some mean. The mean
associated is the total session requests per link when the traffic matrix is generated for
N(N-1) OD pairs.
1. Optimal back up paths; Experiments are carried out with varying load to determine
the number of back up paths to be chosen, as lower number of back up increases the
blocking and higher number burdens the network with additional complexity and
lower resource usage. The results in fig 1 show that three back up paths give better
results considering blocking probability as important parameter. Fig2 shows the link
utilisation for the chosen back up paths.
2. Link utilization: this factor gives the measure if all the links of the fibre cable are
used and laying cable is not a financial waste. This factor is higher the better.

69
Fig 1: Blocking probability v/s back paths Fig 2: Link utilization v/s back up paths
3. Number of converters: wavelength converters form the major part of financial
restriction. Hence usage of minimal number of converters for acceptable blocking is
to be taken care of. The results are shown in fig 3 and fig4.
Fig 3: Blocking probability v/s converters Fig 4: Link utilization v/s converters
4. Reserving wavelengths for primary and back up paths will ensure the priority to
primary requests and continuity. But at the same time blocking and link utilization
should also be acceptable. Fig5 and 6 show the scenario.

70
Fig 5: blocking probability v/s reserved Fig 6: Link utilization v/s reserved
primary wavelengths primary wavelengths
5. The three wavelength routing algorithms, full range converter at every node, full
range converter at limited nodes, and wavelengths reserved primary and back up
routing are compared in fig7 giving the best individual performance scenarios.
Fig 7: Blocking probability v/s routing method

71
V. CONCLUSION
In this paper three RWA algorithms were implemented and experimented to find the
optimal number of converters, number of back up paths and number of wavelengths to be
reserved for acceptable performance given variable load. The project tries to minimize the
combined cost of the optical network under consideration also link utilization factor can
decide whether a cable link is useful if laid physically on field.
REFERENCES
[1] Manoj shivkumar and Krishna shivlingam. “Least resource consumption for survivable
optical networks”, 2005 IEEE.
[2] Manoj shivkumar and Krishna shivlingam. “A routing algorithm framework for
survivable optical networks based on resource consumption minimization,. Journal of
light wave technology”, vol 25 no.7, july2007.
[3] “Tabor Fabry-asztalos, Nilesh Bhide and Krishna shivlingam. Adaptive weight
functions for shortest path algorithms for multi-wavelength optical WDM networks”.,
2000 IEEE
[4] H. Zang, J. P. Jue, and B.Mukherjee, “A review of routing and wavelength assignment
approaches for wavelength-routed optical WDM networks,” Opt. Netw. Mag., vol. 1,
no. 1, pp. 47–60, Jan. 2000.
[5] A. Birman, “Computing approximate blocking probabilities for a class of all-optical
networks,” in Proc. IEEE INFOCOM, Apr. 1995, pp. 651–658.
[6] S. Gowda and K. M. Sivalingam, “Protection mechanisms for optical WDM networks
based on wavelength converter multiplexing and backup path relocation techniques,” in
Proc. IEEE INFOCOM, San Francisco, CA, Mar. 2003, pp. 12–21.
[7] N. M. Bhide, K. M. Sivalingam, and T. Fabry-Asztalos, “Routing mechanisms
employing adaptive weight functions for shortest path routing in multi-wavelength
optical WDM networks,” Springer J. Photon. Netw. Commun., vol. 3, no. 3, pp. 227–
236, Jul. 2001. 1692 JOURNAL OF LIGHTWAVE TECHNOLOGY, VOL. 25, NO. 7,
JULY 2007
[8] M. M. H. Harai and H. Miyahara, “Performance of alternate routing methods in all-
optical switching networks,” in Proc. IEEE INFOCOM, Apr. 1997, vol. 2, pp. 517–525.
[9] S. amamurthy and B. Mukherjee, “Fixed-alternate routing and wavelength conversion
in wavelength-routed optical networks,” in Proc. IEEE Globecom, Sydney, Australia,
Nov. 1998, vol. 4, pp. 2295–2303.
[10] L. Li and A. K. Somani, “Fiber requirement in multifiber WDM networks with
alternate-path routing,” in Proc. ICCCN, Boston, MA, 1999, pp. 338–343.
[11] A. Girard, Routing and Dimensioning in Circuit-Switched Networks. Reading,
MA:Addison-Wesley, 1990.
[12] L. Li and A. K. Somani, “Dynamic wavelength routing using congestion and
neighborhood information,” IEEE/ACM Trans. Netw., vol. 7, no. 5, pp. 779–786, Oct.
1999.
[13] W. Grover, Mesh-Based Survivable Networks: Options and Strategies for Optical,
MPLS, SONET and ATM Networking. Englewood Cliffs, NJ: Prentice-Hall, 2003.
[14] D. Xu, Y. Chen, Y. Xiong, C. Qiao, and X. He, “On finding disjoint paths in single and
dual link cost networks,” in Proc. IEEE INFOCOM, Hong Kong, Mar. 2004,
pp. 715–725.

72
[15] P.-H. Ho, J. Topolcai, and H. T. Mouftah, “On achieving optimal survivable routing for
shared protection in survivable next-generation internet,” IEEE Trans. Rel., vol. 53,
no. 2, pp. 216–225, Jun. 2004.
[16] C. Xin, Y. Ye, S. Dixit, and C. Qiao, “A joint working and protection path selection
approach in WDM optical networks,” in Proc. IEEE GLOBECOM, Nov. 2001, vol. 4,
pp. 2165–2168.
[17] R. He, H. Wen, G. Wang, and L. Li, “Joint optimal path selection algorithm in
survivable WDM networks,” in Proc. 4th Int. Conf. PDCAT, Aug. 2003, pp. 259–262.
[18] J. Zhang, K. Zhu, L. Sahasrabuddhe, J. B. Yoo, and B. Mukherjee, “On the study of
routing and wavelength assignment approaches for survivable wavelength-routed WDM
mesh networks,” Opt. Netw. Mag., vol. 4, no. 6, pp. 16–27, Oct. 2003.
[19] C. S. Ou, J. Zhang, H. Zang, L. Sahasrabuddhe, and B. Mukherjee, “Nearoptimal
approaches for shared-path protection in WDM mesh networks,” in Proc. ICC,
Anchorage, AK, May 2003, pp. 1320–1324.
[20] L. Guo, H. Yu, and L. Li, “Joint routing-selection algorithm for a shared path with
differentiated reliability in survivable wavelength-division multiplexing mesh
networks,” Opt. Express, vol. 12, no. 11, pp. 2327–2337, May 2004.

MMINIMUM RESOURCE CONSUMPTIONS ROUTING FOR OPTICAL NETWORKS

More Related Content

What's hot (16)

Viewers also liked (20)

Similar to MMINIMUM RESOURCE CONSUMPTIONS ROUTING FOR OPTICAL NETWORKS (20)

More from prj_publication (20)

Recently uploaded (20)

MMINIMUM RESOURCE CONSUMPTIONS ROUTING FOR OPTICAL NETWORKS