Routing and Wavelength Assignment in Optical

Networks
Olivier Brun, Sami Baraketi

To cite this version:

Olivier Brun, Sami Baraketi. Routing and Wavelength Assignment in Optical Networks. 2014. �hal-
01062321�

HAL Id: hal-01062321

https://hal.science/hal-01062321
Submitted on 11 Sep 2014

HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est

archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents
entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non,
lished or not. The documents may come from émanant des établissements d’enseignement et de
teaching and research institutions in France or recherche français ou étrangers, des laboratoires
abroad, or from public or private research centers. publics ou privés.
 ROUTING AND WAVELENGTH ASSIGNMENT IN OPTICAL
NETWORKS

O. BRUN, S. BARAKETI

1. Introduction
Transport technologies such as Synchronous Digital Hierarchy (SDH/SONET) and Asyn-
chronous Transfer Mode (ATM) becames increasingly speed-limited and can no longer re-
spond to the demand for high bandwidth services (HDTV, video conferencing, electronic
banking, multimedia applications, etc). They, through employing optical fiber, do not re-
alize the full potential of the optical medium. The speed of these technologies are limited
to few tens of Gbps du to the peak electronic speed of the network components, wheres a
single mode fiber can carry data at very highest speeds. In order to increase the bandwidth
of optical fiber, Wavelength division multiplexing (WDM) technology is evolved. It is a
promising technology to effectively utilize the enormous bandwidth of optical fiber.

In wavelength division multiplexing technology the transmission spectrum of a fiber link

can be divided into many protocol transparent channels. Multiple channels can be oper-
ated in a single fiber simultaneously at different wavelengths, providing for each channel
the bandwidth that is compatible with current electronic processing speeds. These chan-
nels can be independently modulated to accommodate dissimilar data formats at various
bit rates if necessary. By utilizing WDM in optical networks, we can achieve link capacities
on the order of Tbps.

WDM networks are rapidely evolved as a powerfull class of networks for use in wide
area networks. These networks consist of optical switches that route a signal based on
the identity of the input port (i.e. related overlying service) and the wavelength of the
incoming signal. A WDM network is called also wavelength routed network [29] [30] since
it employs wavelength routing technique. Access switches and end switches provide the
electronic-to-optical conversion and vice versa to interface the optical network with elec-
tronic stations. Wavelength routing provides the network with the ability to identify and
localize the traffic flow, thereby allowing the same wavelength to be reused in spatially
disjoint segments of the network. In order to carry data from one access node to another,
a connection needs to be set up at the optical layer similar to the case in a circuit-switched
networks. This operation is performed by determining a path in the network connecting
the source node to the destination node and by allocating a single free wavelength on all
of the fiber links in the path. Such an all-optical path is referred as lightpath [7] [30], each
lightpath can carry data at peak electronic speed. However, practical limitations on the
1
2 O. BRUN, S. BARAKETI

transmission technology and optical devices restrict the number of available wavelengths
per fiber link, it is unlikely that a lightpath can be established between every pair of access
nodes. The intermediate nodes in the path route the lightpath in the optical domain using
wavelength-sensitive switches. A fundamental constraint in a wavelength-routed optical
network is that two or more lightpaths traversing the same fiber link must be on different
wavelengths so that they do not interfere with one another. A wavelength-routed network,
which carries data from one access station to another without any intermediate optical-
to-electrical conversion is referred as an all-optical wavelength-routed network. All-optical
wavelength-routed networks will be the subject of our work. These networks have several
benifits like the potential to accommodate the rapidly increasing bandwidth, improved
network reliability, simpler network management, and are independent from modulation
format and bit rate [11] [30].

Since the lightpaths are the basic switched entities of a wavelength-routed WDM net-
work, their effective establishment and usage are crucial. Thus, it is important to propose
efficient algorithms to select the routes for the requested connections and to assign wave-
lengths on each of the links along these routes. This is known as the routing and wavelength
assignment problem. The routing and wavelength assignment problem (RWA) in optical
networks considers a network where requests (i.e. lightpaths) can be transported on dif-
ferent optical wavelengths through the network. Each accepted request is allocated a path
from its source to its sink, as well as a specic wavelength. Lightpaths routed over the
same link must be allocated to different wavelengths, while lightpaths whose paths are
link disjoint may use the same wavelength. Lightpaths that cannot be established due to
constraint on wavelengths availability are said to be blocked.

2. Litterature Survey
Several works have studied the RWA problem in all-optical WDM networks and various
contributions have been made through interesting algorithms. The problem consider a
directed network G = (V, E) where V is the set of nodes representing the switches of
the physical network and E is the set of edges representing the fiber links of the physical
network. Given a set of requests for all-optical connections or lightpaths between node-
pairs and a set of available wavelengths, the problem is to find routes from the source nodes
to their respective destination nodes and assign wavelengths to these routes.
The RWA problem in WDM networks can be categorized into two types based on traffic
arrivals.
(1) Static Lightpath Establishment (SLE) : The traffic is static and the set of connection
requests is known in advance. This kind of problem pertains to the planning phase of the
WDM network. The algorithms proposed for solving the static RWA problem are referred
to as Offline algorithms. Static RWA is known to be an NP-hard optimization problem [7]
since it is considered as a special case of the integer multicommodity flow (MCF) problem
[21] [13] with additional specific constraints and can be formulated as an integer linear
program (ILP).
 ROUTING AND WAVELENGTH ASSIGNMENT IN OPTICAL NETWORKS 3

(2) Dynamic Lightpath Establishment (DLE) in the case of dynamic traffic : The traffic
is dynamic and the connection requests arrive sequentially, one by one, at random times
over an infinite time horizon. The DLE problem is posed in the operational phase where
each network resource must be managed efficiently. The algorithms proposed for solving
the dynamic RWA problem are referred to as Online algorithms. A review on dynamic
RWA algorithms can be found in [40].

In our study we will focus on the static RWA problem and the offline algorithms that
proposed in litterature to solve this kind of problem. Offline RWA is more difficult than
online RWA since it aims at jointly optimizing the lightpaths used by the connections in the
same way that the multicommodity flow problem is more difficult than the shortest-path
problem in general networks. We classify the previous works in three classes. For each
class we discuss the already proposed algorithms and the adopted approaches.

2.1. Joint RWA ILP-based algorithms. Many ILP-based formulations have been pro-
posed in litterature for routing and wavelength assignment problems jointly. [15] provides
different ILP formulations (path-based, edge-based and arc-based formulations). The au-
thors proposed a synthesis of the mathematical models for symmetrical systems where
bandwidth requirements are similar on the downstream and upstream directions. Several
objectives have been considered: (a) minimizing the blocking rate or maximizing the num-
ber of accepted requests given a fixed number of available wavelengths, (b) minimizing the
number of used wavelengths assuming that all connection requests can be accepted, (c)
minimizing the congestion that is expressed through the minimization of the maximum
number of wavelengths on a given fiber link, (d) minimizing the network load defined as
the ratio of used wavelengths over the overall potential number of wavelengths. A contin-
uous relaxations have been also proposed. The authors of [5] developed an ILP model for
RWA problem using other optimization criteria like minimizing the number of wavelength
conversions or minimizing the hop count. [34] proposed two direct RWA ILP formulations
: basic formulation in which the aim is to minimize the maximum number of wavelengths
per fiber link, and extended formulataion that try to minimize the overall number of used
wavelengths. A source aggregation is considered to reduce the number of constraints and
improve the two mathematical models. In [14], the authors proposed a new formulation
that addresses the complete trafc grooming problem, including topology design as well
as routing and wavelength assignment (RWA) of lightpaths. In Other works, the authors
used some efficient techniques in order to scale their RWA ILP formulations to problem
instances encountered in practice. A link selection techniques were considered in [24] to
reduce the size of the link-based formulation in terms of both the number of variables and
the number of constraints. Column Generation technique was also used in [16] [20] to im-
prove the RWA formulation. In [23], the authors demonstrated that their new path-based
formulation achieves a decrease of up to two orders of magnitude in running time compared
to existing formulations. Since the majority of proposed ILP were very hard to solve, a
corresponding relaxed linear programs have been used to get bounds on the optimal value
that can be achieved. As an example, The LP-relaxation formulation proposed in [27], and
4 O. BRUN, S. BARAKETI

also considered in [9], is able to produce exact RWA solutions in many cases, despite the
absence of integrality constraints. In [10], the authors presented an algorithm for solving
the static RWA problem based on a LP-relaxation formulation that provides integer opti-
mal solutions despite the absence of integrality constraints for a large subset of RWA input
instances. The RWA formulation was then extended in order to take into consideration the
physical layer impairments and account for the interference among lightpaths. Iterative
fixing and Rounding techniques have been also used to provide an integer solution for the
relaxed problem.

2.2. RWA decomposition-based algorithms. Another known approach in litterature

is to break the RWA problem in the two constituent subproblems and solves them in-
dividually and sequentially. This approach consists on two steps, by first finding routes
for all requested connections and secondly searching for appropriate wavelength assign-
ment [39] [37]. Note that both subproblems are NP-hard: The routing subproblem for a
set of connections corresponds to a multicommodity flow problem, while wavelength as-
signment corresponds to a graph coloring problem. Several works have been adopted a
decomposition-based algorithms. In [34] [2] [9], the authors choose a decomposition tech-
nique which handles the first step with a ILP program which assigns paths to the demands
while minimizing the maximal number of demands routed over a link. The second step
is expressed as a graph coloring problem where the nodes are demands and disequality
constraints (links) are imposed between any two demands which are routed over the same
link. The final solution is an approximate solution of the original complete RWA problem
taht can be not optimal. Works [2] [9], formulated the routing sub-problem as a continuous
multicommodity flow problem and applied a randomized rounding technique to provide an
integer solution in which the objective function takes on a value close to the optimum of the
rational relaxation. A performance comparison between a RWA ILP-based algorithm and
a decomposition-based algorithm was made in [9]. However, [39] synthesized a lot of known
approaches for resolving the routing sub-problem. Fixed routing, fixed-alternate routing
and adaptative routing were described. Ten dynamic wavelength assignment heuristics
were also discussed. the authors said that these heuristics may also be applied to the static
wavelength assignment problem by ordering the lightpaths and then sequentially assigning
wavelengths to the ordered lightpaths. Note that all the proposed decomposition tech-
niques, previously cited, suffer from two major drawbacks: (1) The approximate solution,
obtained as a result of the problem decomposition, is often not optimal (2) The optimal
solution of the joint RWA might not be included in the solutions provided by the algorithms
used for the two subproblems.

2.3. RWA heuristics. For static lightpath establishment (SLE), several heuristic algo-
rithms have been proposed for establshing a maximum number of lightpaths from a given
set of request [12] [26] [42]. However, most of these old algorithms are based on traditional
circuit-switched networks where routing and wavelength assignment steps are decoupled.
Several recent studies have been focused on solving the joint RWA problem by sophisticated
heuristics. In [35, 33, 6, 36], a lot of proposed heuristics were presented and evaluated. In
 ROUTING AND WAVELENGTH ASSIGNMENT IN OPTICAL NETWORKS 5

the following, we summarize and describe the most known heuristics used for solving the
joint RWA problem.

• Shortest First Fixed Path (SFFP) [33]: M. shiva Kumar used wavelength-graph or
layered-graph to find routes and wavelengths for the given lightpath set. In the
layered graph model the physical network G = (V, E) is replicated |W| times, each
sub graph G = (Vλ , Eλ ), λ ∈ W corresponding to the given physical network on a
particular wavelength. The objective here was to maximize the network throught-
put. The heuristic algorithm finds the shortest path for all the given node-pairs
by using Dijkstra’s algorithm in the given physical network topology. Then the
requested connections are arranged in non-decreasing order of their path lengths.
Now heuristic algorithm routes lightpaths sequentially on the wavelength-graph in
a single layer with the shortest path. If a route is found with finite cost, then the
lightpath is established, and the wavelength-graph is updated by assigning infinite
cost to the edges along which the request is routed. If a route with minimum cost
is not found for a node pair then the request is skipped, and the next request from
the sorted list is considered. In the second phase the algorithm finds the short-
est available path on the residual wavelength-graph for the lightpath requests that
were skipped in the first phase. Thus by assigning a wavelength to the shortest
path first maximizes the number of lightpaths established, which is equivalent to
maximizing the network throughput.

• Longest First Fixed Path (LFFP) or Longest-path First [7] [35]: In this heuristic,
wavelengths are assigned to the longest lightpath first. A comparison with SFFP
was made in [33] based on the number of established lightpaths. The results showed
that The number of lightpaths established by SFFP is more after a certain size of
the lightpath request set. However, the average fiber utilization obtained by LFFP
is more when compared to SFFP. This is because, by establishing long lightpaths
first, would result better wavelength reuse.

• Minimum Number of Hops (MNH) [3] [35]: Baroni and Bayvel proposed an MNH
algorithm for minimizing the maximum load per link in arbitrarily connected net-
works. In MNH, each node-pair of the given set of connection requests is firstly
assigned one of its shortest paths. Then, alternate shortest paths are examined
for a possible better path and the previously assigned path will be replaced by the
alternate path if the load of the most congested link is reduced. This process is
repeated for all the node-pairs and stops when no subtitutions are possible. Results
in [35] showed that the MNH provided a more efficient routing with a minimum
number of used wavelengths than LFFP.

• Longest First Alternate Path (LFAP) [6] [35]: In LFAP, the RWA problem is for-
mulated as a knapsack problem. Wavelengths are treated as knapsacks, each of
which can hold more than one lightpath. Lightpaths are treated as items and more
6 O. BRUN, S. BARAKETI

than one lightpath can share the same wavelength on condition that no two light-
paths pass through the same link. The LFAP algorithm assigns a wavelength to
a longer lightpath with higher priority and attempts to maximize the number of
lightpaths per wavelength. More precisely, wavelengths are added one by one until
all the lightpaths for the given set of requests are established. For each newly added
wavelength, the longest lightpath among those of the given requests is established.
Then, the shorter lightpaths will be checked one by one. If no lightpath can be
established, alternate paths are searched. If no lightpath can be established any
more, a new wavelength will be added and the searching process is repeated. Af-
ter establishing each lightpath, the network topology is modified by removing the
links used by the newly established lightpath. The results in [35] showed that the
performance (i.e. number of wavelengths required) of LFAP is much better than
LFFP and MNH. However, LFAP spends more time than LFFP in order to provide
solution.

• Heaviest Path Load Deviation (HPLD) [35]: In HPLD, the RWA problem is for-
mulated as a routing problem where the link cost is determined based on the load
(utilization) of each link. The HPLD algorithm attempts to re-route some light-
paths that pass through the heaviest link in order to minimize the number of wave-
lengths. More precisely, HPLD algorithm search to deviate the load of the most
loaded link to other less loaded path so that the maximum number of wavelengths
used in the network is reduced. That is, the HPLD algorithm tries to re-route some
lightpaths that pass through the heaviest link. The HPLD algorithm employs the
shortest path routing technique to solve this problem based on the network graph.
The weight function of a link (link cost) is determined by the link load. The results
in [35] showed that HPLD is a bit more efficient than LFAP in term of the number
of used wavelengths. However, it needs more time than LFAP to find a feasible
solution.

Note that the four first following heuristics were developed by applying classical bin
packing algorithms [19]. Lightpath requests represented items, while copies of graph G
represented bins. Each copy of G, referred to as bin Gi , i = 1, . . . , |W |, corresponds to one
wavelength. The authors in [36] showed that their heuristics were tested on a series of large
random networks and compared with an efficient previous algorithm for the same prob-
lem. Results indicated that the proposed algorithms yield solutions signicantly superior in
quality, not only with respect to the number of wavelength used, but also with respect to
the physical length of the established lightpaths.

• First Fit (FF-RWA) [36], called also Greedy-EDP-RWA in [25] based on the bounded-
length greedy algorithm for Edge-Dijoint Paths problem: First, only one copy of G,
bin G1 , is created. Higher indexed bins are created as needed. Lightpath requests
(sj , dj ) are selected at random and routed on the lowest indexed copy of G in which
 ROUTING AND WAVELENGTH ASSIGNMENT IN OPTICAL NETWORKS 7

there is place (i.e. room). Bin Gi is considered to have room for lightpath (sj , dj ) if
the length of the shortest path from sj to dj in Gi , denoted as Pji , is less than H. If
a lightpath is routed in bin Gi, the lightpath is assigned wavelength i and the edges
along path Pji are deleted from Gi . If all the edges from bin Gi are deleted, the bin
no longer needs to be considered. If no existing bin can accommodate lightpath
request (sj , dj ), a new bin is created.

• Best Fit (BF-RWA) [36]: BF-RWA routed lightpaths in the bin into which they t
“best”. The algorithm considered the best bin to be the one in which the lightpath
can be routed on the shortest path. In other words, if at some point in running
the algorithm, there are B bins created, bin Gi ,1 ≤ i ≤ B , is considered to be the
best bin for lightpath (sj , dj ) if l(Pji ) ≤ l(Pjk ), for all k = 1, . . . , B, and k 6= i. This
is not necessarily the overall shortest path, SPj , since it is possible that none of
the existing bins have this path available. If there is no satisfactory path available
in any of the B bins (i.e. l(Pji ) > H, for i = 1, . . . , B), a new bin is created.
The motivation for the “best fit” approach described above, is not only to use less
wavelengths, but also to minimize the physical length of the established lightpaths

• First Fit Decreasing (FFD-RWA) [36]: FFD-RWA algorithm sorts the lightpath
requests in non-increasing order of the lengths of their shortest paths, SPj , in G.
Lightpaths with shortest paths of the same length are placed in random order. The
algorithm then proceeds as FF-RWA. The motivation for such an approach is as
follows. If the connection request with the longest shortest path is considered rst, it
will be routed in “empty” bin G1 . This means the lightpath will not only success-
fully be routed in G1 , but will be routed on its overall shortest path. After deleting
the corresponding edges from bin G1 , the remaining edges can be used to route
“shorter” lightpath requests which are easier to route on alternative routes that
are satisfactory (i.e. shorter than H). In other words, the FFD-RWA algorithm rst
routes “longer” lightpaths which are harder to route, and then lls up the remaining
space in each bin with “shorter” lightpaths. This may lead to fewer wavelengths
used.

• Best Fit Decreasing (BFD-RWA) [36]: BFD-RWA algorithm sorts the lightpath re-
quests in non-increasing order of the lengths of their shortest paths SPj in G, and
then proceeds as BF-RWA. The obtained results in [36] showed that BFD-RWA
provided more better solution (i.e. number of required wavelengths) than the three
other algorithms.

• A weight based Edge Dijoint Path (WEDP) algorithm has been also proposed in
[32] to solve the RWA problem. The authors showed that WEDP performed better
than a greedy MEDP algorithm used for solving the same problem.
8 O. BRUN, S. BARAKETI

Table 1. Summary of static RWA algorithms

Problems Approaches Comments References

Joint RWA ILP Formulation NP-complete [15, 5, 34, 14, 24, 23]
(including relaxed formulation) [27, 9, 10]
SFFP
LFFP
MNH Heuristics [7, 35, 33, 6, 3]
LFAP
HPLD Objective: Minimizing
FF-RWA the number of required
BF-RWA wavelengths [36, 25]
FFD-RWA
BFD-RWA
Routing ILP Formulation NP-complete [31, 34, 2, 9]
(including relaxed formulation)
Fixed routing [39, 22, 28]
Alternate routing
Wavelength Graph coloring NP-complete [39, 34, 2, 9]
Assignment (routes are known)
Random
LU (SPREAD) Heuristics
FF used with
MP (multi-fiber) Fixed Routing [39, 4, 17, 18, 38, 41]
MU (PACK) approach
LL (multi-fiber)
M
P

RCL

The Table 1 summarizes all the previous described works. Note that we illustrate only
static/offline RWA cases.

3. Minimization of the number of lighpaths

Given a set of traffic demands, our goal is to minimize the number of new lighpaths to
be created in the optical network for supporting these demands. We address this problem
by considering only the edge nodes of the Optical Transport Network (OTN). We view a
lighpath between nodes u and v as a directed edge between these nodes, with a certain
capacity representing the maximum amount of traffic that a lighpath can accomodate. We
assume that the number of lighpath between any two nodes is given by some large integer
value K. We show that the problem of minimizing the number of lighpaths can then be
formulated as that of routing demands in a multigraph with the goal of minimizing the
number of used links. We propose a simple heuristic to solve this problem.
 ROUTING AND WAVELENGTH ASSIGNMENT IN OPTICAL NETWORKS 9

3.1. Problem statement. We consider a network represented by a multidigraph G :=

(V, E, s, t), where
• V is a set of vertices or nodes,
• E is a set of edges or lines,
• s : E → V assigns to each edge its source node,
• t : E → V assigns to each edge its target node.
The network is such that there are K edges between any two pairs of nodes, that is the
cardinality of the set

{e ∈ E : s(e) = i, t(e) = j} ,
is K for all i, j 6= i in V . An example of such a network is shown in Figure 1. It is assumed
that the capacity of each edge e ∈ E is C units of bandwidth.

1 2

Figure 1. Simple example of network with 3 nodes and 2 directed edges

between each pair of nodes.

We are given a set D of traffic demands that have to be routed in the network. For each
demand d ∈ D, we let λd be the traffic volume of demand d, sd its source node and td its
destination node. Traffic demand d ships its flow by splitting its demand λd over a set of
paths Π(d). We assume that the set Π(d) contains all simple path between sd and td . Since
the network is symmetric, the number of paths is the same for all demands and it will be
denoted by S in the following. Let xd,π denote the amount of trafic sent by demand d over
path Pπ. A routing strategy for demand d ∈ D is a vector xd = (xd,π )π∈Π(d) in IRS+ such
that π∈Π(d) xd,π = λd . We let Xd denote the set of all routing strategies for demand d:
 
 X 
Xd = xd ∈ IRS+ : xd,π = λd .
 
π∈Π(d)

The vector x = (x
Nd )d∈D will be called a routing strategy. It belongs to the product
strategy space X = d∈D Xd . For each edege e ∈ E, we define the function ye : X → IR+
by
X X
ye (x) = δeπ xd,π .
d∈D π∈Π(d)
10 O. BRUN, S. BARAKETI

where δeπ = 1 if e ∈ π, and 0 otherwise. In other words, ye (x) represents the amount of
traffic flowing through link e under strategy x. A routing strategy is feasible if ye (x) ≤ C
for all links e. In the following, we say that link e is used under strategy x if and only if
ye (x) > 0.
We wish to find a feasible routing strategy which minimizes the number of used links.
This amounts to solving the following optimization problem:
X
(OPT) minimize v(x) = 1{ye (x)>0}
e∈E
subject to
x ∈ X,
ye (x) ≤ C, ∀e ∈ E.
It is immediate to see that problem (OPT) can be formulated as the following integer
programming problem:
X
(OPT-LP) minimize be
e∈E
subject to
ye ≤ C.be , ∀e ∈ E,
X X
ye = δeπ xd,π , ∀e ∈ E,
d∈D π∈Π(d)
X
xd,π = λd , ∀d ∈ D,
π∈Π(d)
xd,π ≥ 0, ∀π ∈ Π(d), ∀d ∈ D,
ye ≥ 0, ∀e ∈ E,
be ∈ {0, 1}, ∀e ∈ E.
In the following, we shall assume that K is sufficiently large for the above problem to
have a feasible solution. More precisely, we shall assume that
P 
d λd
K≥ .
C
We note that problem (OPT-LP) involves binary variables and is thus difficult to solve.
We propose in the following section a simple successive approximation heuristic to find an
approximate solution.
3.2. Successive approximation heuristic. Before describing the successive approxima-
tion algorithm in Section 3.2.2, we first establish in Section 3.2.1 some results regarding
the optimal routing strategy of a single demand when the routing strategies of the other
demands are given.
 ROUTING AND WAVELENGTH ASSIGNMENT IN OPTICAL NETWORKS 11

3.2.1. Optimal routing strategy for a single traffic demand. Denote by x−d the vector
(xf )f 6=d , that is the routing strategy obtained by routing all traffic demands but demand d.
It is assumed that this routing strategy is fixed and feasible. It can be viewed as a partial
solution to problem (OPT). Define E − as the set of links that are used under strategy
x−d , that is

E − = {e ∈ E : ye (x−d ) > 0} ,
and define E + as E + = E \ E − . We do not make explicit the dependance on d in order
to simplify notations. Let ce = C − ye (x−d ) be the residual capacity for demand d on link
e. The minimum-cost solution that can be obtained from the partial solution x−d is then
obtained by solving the following optimization problem:

X
(OPT-d) minimize vd (xd ) = 1{ye (xd )>0}
e∈E +
subject to
x d ∈ Xd ,
ye (xd ) ≤ ce , ∀e ∈ E.

As before, we note that the above problem can be formulated as a mixed linear program-
ming problem:

X
(OPT-d-LP) minimize βe
e∈E +
subject to
ze ≤ C.βe , ∀e ∈ E + ,
ze ≤ ce , ∀e ∈ E − ,
X
ze = δeπ xd,π , ∀e ∈ E,
π∈Π(d)
X
xd,π = λd ,
π∈Π(d)
xd,π ≥ 0, ∀π ∈ Π(d),
ze ≥ 0, ∀e ∈ E,
βe ∈ {0, 1}, ∀e ∈ E.

In the following, we denote by x∗d an optimal solution of problem (OPT-d). Proposition

1 gives an explicit expression for the cost of such an optimal routing strtategy.

Proposition 1. The optimal value of problem (OPT-d) is

12 O. BRUN, S. BARAKETI

λd − λ (E − )
 
vd (x∗d ) = ,
C

where λ (E − ) is the optimal value of the following linear program:

(Max-Flow) maximize λ
subject to
ze ≤ ce , ∀e ∈ E − ,
X
ze = δeπ xd,π , ∀e ∈ E − ,
π∈Π(d)
X
xd,π = λ,
π∈Π(d),π⊂E −
λ ≤ λd ,
xd,π ≥ 0, ∀π ∈ Π(d), π ⊂ E −
ze ≥ 0, ∀e ∈ E − ,

 
λd −λ(E − )
Proof. Let n = C . The proof is in two parts. We first show that for each routing
strategy xd ∈ Xd such that ye (xd ) ≤ ce , ∀e, we have vd (xd ) ≥ n. We then prove that there
exists a routing strategy for demand d whose cost is n. P
Consider a solution xd ∈ Xd such that ye (xd ) ≤ ce , ∀e. Let α = π∈Π(d),π⊂E − xd,π be
the amount of traffic routed on paths in the subgraph (V, E − ). By definition of λ (E − ), we
have α ≤ λ (E − ). Since ye (x
C
d)
≤ 1 for all e ∈ E + , we have

X
v(xd ) = 1{ye (xd )>0} ,
e∈E +
X  ye (xd ) 
= ,
+
C
e∈E
 
1 X
≥ 
C ye (xd ).
 + e∈E 

However
 ROUTING AND WAVELENGTH ASSIGNMENT IN OPTICAL NETWORKS 13

X X X
ye (xd ) = δeπ xd,π
e∈E + e∈E + π∈Π(d)
 
X X
=  δeπ  xd,π
π∈Π(d) e∈E +
X
≥ xd,π
π∈Π(d),π∩E + 6=∅
≥ λd − α,
from which we obtain

λd − λ (E − )
   
λd − α
v(xd ) ≥ ≥ = n.
C C
We thus conclude that any feasible solution to (OPT-d) has a cost greater than or equal
to n.

We now turn to the second part of the proof. Let x− d be the optimal solution of problem
(Max-Flow). If n = 0, i.e., if λ (E ) = λd , then the routing strategy x∗d defined by
−

(
x− −
d,π if π ⊂ E ,
x∗d,π =
0 otherwise.
is clearly an optimal solution to (OPT-d). Otherwise, choose arbitrarily n edges e1 , . . . , en ∈
E + such that s(ek ) = sd and t(ek ) = td for k = 1, . . . , n and let πk be the path πk = {ek }
for k = 1, . . . , n. Consider the routing strategy x∗d ∈ Xd defined as follows:
 −
xd,π
 if π ⊂ E − ,
λd −λ(E − )
x∗d,π = π = π 1 , . . . , πn
 n
0 otherwise.


We clearly have x∗d,π ≥ 0 for all paths π ∈ Π(d). Moreover,

X X n
X
x∗d,π = x∗d,π + x∗d,πk ,
π∈Π(d) π∈Π(d),π⊂E − k=1
−
+ λd − λ E − ,



= λ E
= λd ,
which proves that x∗ ∈ Xd . Note that by definition of x− ∗ −
d , we have ye (xd ) = ye (xd ) ≤ ce
λ −λ(E )
−
for all e ∈ E − . Moreover, since d n ≤ C, we have yek (x∗d ) ≤ C = ce for k = 1, . . . , n.
We thus conclude that ye (xd ) ≤ ce , ∀e ∈ E. Since the routing strategy x∗d uses n links in
∗
 14 O. BRUN, S. BARAKETI

E + , we conclude that vd (x∗d ) = n, and thus there exists at least one feasible solution to
(OPT-d) whose cost is n. 

We note that the second part of the proof of Proposition 1 provides an algorithm to find
an optimal solution to problem (OPT-d):
Step 1. Solve Problem (Max-Flow) in order to find the value of λ (E − ) and the amount
of traffic to be routed on each path π between sd and td in the subgraph (V, E − ).
We note that the structure of this problem is that of a standard maximum flow
problem, for which very efficient algorithms
 are known [1].
λd −λ(E − )
Step 2. Choose arbitrarily n = C edges in E + between sd and td , and split evenly
the traffic λd − λ (E − ) among these links.

3.2.2. Successive approximation algorithm. The successive approximation algorithm is de-

scribed in Figure 2. At each iteration, this algorithms routes optimally a single traffic
demand assuming the routing strategies of the other demands are fixed. The algorithm
stops when it is no more possible to decrease the number of used links by re-routing a
single traffic demand.
Lemma 1. The successive approximation algorithm converges in a finite number of steps.
Proof. Define a round to be a sequence of iterations of the algorithm in which each traffic
demand is rerouted exactly once (Steps 4-19 in Figure 2). Once an order is fixed in the
first round, it is assumed to be the same in each subsequent round (the order in which
the traffic demands are routed in the first-round can be arbitrary). Observe that if the
algoritm does not stop at the end of a round, then the number of used links at the end of
the round is lower by at least one than the number of used links at the end of the previous
round. Hence, the number of used links at the end of each round is a strictly decreasing
sequence. Since this number has to be positive, the algorithm converges in a finite number
of rounds. 

3.3. Experiments and results. In this section, several experiments on the optimization
problem for traffic routing were performed in order to evaluate and validate the effectiness
of the proposed solutions. The CPLEX solver, from IBM-ILOG society, is used for solving
the integer linear program (OPT-LP). The aim of the experiments here is to analyze the
performance and compare the proposed methods. The comparison is performed between
the ILP-based algorithm (OPT-LP) and the successive approximation algorithm illustrated
in Fig. 2.

In the following sections, the performance comparison between the algorithms is done
based on three criteria. The first and second ones are related to the computation time and
the memory consumption, respectively. The third one represents the routing strategy cost,
which reflects the number of used links for the routing of all demands.
 ROUTING AND WAVELENGTH ASSIGNMENT IN OPTICAL NETWORKS 15

Require: G := (V, E, s, t), C, D

1: x ← 0
2: Convergence = false
3: while Convergence = false do
4: Convergence = true
5: for d ∈ D do
6: E − ← {e ∈ E : ye (x−d ) > 0}, E + ← E \ E −
7: ce ← C − ye (x−d ), ∀e ∈ E
8: x∗d ← 0
9: Compute λ (E − ) and x∗d,π for all π ⊂ E − by solving (Max-Flow)
 
λ −λ(E − )
10: n= d C
11: for k = 1 . . . n do
12: Choose arbitrarily edge ℓ ∈ E + such that s(ℓ) = sd and t(ℓ) = td
13: x∗d,{ℓ} = (λd − λ (E − ))/n
14: E + ← E + \ {ℓ}, E − ← E − ∪ {ℓ}
15: end for
16: if n < v(x) − v(x−d ) then
17: Convergence = false
18: xd ← x∗d
19: end if
20: end for
21: end while

Figure 2. Successive approximation algorithm.

3.3.1. Computation Time. To evaluate the computation time for both ILP-based algorithm
and successive approximation algorithm, several topologies, with different sizes, are con-
sidered for simulations. In each simulation, the two algorithms are executed on the same
topology and for the same set of demands which are randomly generated. Table 2 illustrates
the obtained results for each topology size. For the last three simulations, the execution of
the ILP-based algorithm is stopped before its termination because lack of sufficient RAM
memory.

From these results, it can be seen clearly that the successive approximation algorithm
is much more faster than the ILP-based one. A routing strategy solution can be found in
a few seconds for large topology sizes.

3.3.2. Memory Consumption. For evaluating the memory consumption related to the two
algorithms, the same set of simulations as previous section is considered here. For the last
three simulations, the execution of the ILP-based algorithm is stopped before its termina-
tion because lack of sufficient RAM memory. The results of simulations are illustrated in
16 O. BRUN, S. BARAKETI

Table 2. Computation Time

Topology Size ILP-Based Algorithm Successive Approximation Algorithm

5 nodes 0.08s 0.008s
8 nodes 0.5s 0.01s
10 nodes 2m19s 0.014s
12 nodes 6m54s 0.2s
15 nodes 27m43s 0.3s
20 nodes 36m21s 0.6s
25 nodes 52m16s 0.8s
30 nodes >1h 0.9s
50 nodes >2h 1.2s
100 nodes >5h 3.4s

Table 3. Memory Consumption

Topology Size ILP-Based Algorithm Successive Approximation Algorithm

5 nodes 80M 2M
8 nodes 170M 4M
10 nodes 240M 6M
12 nodes 825M 8M
15 nodes 1.3G 9M
20 nodes 1.8G 11M
25 nodes 3.2G 14M
30 nodes >4G 18M
50 nodes >4G 26M
100 nodes >4G 40M

Table 3.

As mentionned in the Table 3, important memory consumptions are recorded when

solving the routing problem by the ILP-based algorithm. However, the successive approx-
imation algorithm gave a solution with a memory consumption of 40M for a 100-node
topology.

3.3.3. Routing Strategy Cost. The routing strategy cost is equal to the number of used
links when solving the routing problem. This value is returned by the objective function in
the linear program (OPT-LP) and deduced in the successive approximation algorithm after
routing all demands. To compare the quality of routing for the two algorithms, it suffices
to compare theirs routing strategies costs. For this comparison, a simplistic topology with
10 nodes is considered. Twenty five Simulations have been performed on this network.
For each simulation, both ILP-based algorithm and successive approximation algorithm
try to route the same set of demands, which are randomly generated. Figure 3 represents
 ROUTING AND WAVELENGTH ASSIGNMENT IN OPTICAL NETWORKS 17

Figure 3. Routing Strategy Cost

two curves illustrating the routing strategy costs, recorded for the two algorithms. And for
more precision, Figure 4 illustrates the relative error between the optimal and approximate
solutions.

The obtained results shows that the successive approximation algorithm provides a lower
quality of routing that the ILP-based algorithm, but globally acceptable. In fact, as this
heuristic iterates through the demands sequentially, it attempts to find optimal solution for
each demand without taking into account the impact on following ones in later iterations.
Therefore, it may not always reach the performance of the ILP-based algorithm which
solves the routing problem for all demands altogether. In this example shown in Figure 4,
the average gap, along 25 simulations, between the two solutions is around 7, 92%. Average
gaps of 7, 85%, 7, 82% and 7, 78% have been recorded, along 100 simulations, for 10-node,
15-node and 20-node topologies respectively. According to the obtained results, the routing
quality ensured by the successive approximation algorithm seems satisfactory compared
to the optimal ILP solution, especially if we consider the gain in time computation and
memory consumption.
The average gap is expressed by
P Solapprox −Solopt
simulations ( Solapprox )
nbsimulations
18 O. BRUN, S. BARAKETI

Figure 4. Relative error between optimal and approximate solutions

where Solapprox and Solopt represent the approximative solution and the optimal solution
respectively.

3.3.4. Convergence Statistics. An other type of statistics have been performed to evalu-
ate the convergence of the successive approximation algorithm. These statistics consists
on computing the number of iterations performed in order to converge. It represents the
convergence index of the algorithm. The same set of simulations shown in the sub-section
3.3.3 are considered here. The figure 5 illustrates the obtained results.

These results confirms that the successive approximation algorithm converges in a finite
number of iterations.

4. Routing of Lightpaths and Wavelength Assignment

The routing and wavelength assignment (RWA) problem is an optical networking prob-
lem. The general objective of the RWA problem is to maximize the number of established
connections.

4.1. Integer Linear Programming Formulation. We consider an optical transport

network represented by an directed graph G := (V, E), where V is a set of vertices or
nodes corresponding to the switches of the OTN and E is the set of optical fibers between
 ROUTING AND WAVELENGTH ASSIGNMENT IN OPTICAL NETWORKS 19

Figure 5. Successive Approximation Algorithm Convergence : Number of iterations

nodes. We denote by W = {1, . . . , W } the set of wavelengths (or colours) that can be
assigned to lighpaths. We are given a set K of lightpaths that have to be routed in the
network. Each lighpath request must be given a route and a wavelength. The wavelength
must be consistent for the entire path (it is assumed that no wavelength converter is used).
Two lighpaths can share the same optical link, provided a different wavelength is used. For
each lighpath k ∈ K, we let sk be its source node and tk be its destination node.
For each coulour w ∈ W, we define a layer of the network as a graph Gw = (Vw , Ew ),
where each node of Vw is obtained by duplicating the corresponding node of V , and each
edge of Ew is obtained by duplicating the corresponding directed edge in E. We thus have
as many network layers as there are possible colours. This is illustrated in Figure 6. We
view each network layer as a separate network where each link has capacity 1, so that
a single lighpath can be routed ont it. Then, the problem can be formulated as that of
routing each lighpath in one and only one network layer.
For all lighpaths k ∈ K and all colours w ∈ W, let us define the following binary decision
variables:
(
1 if lighpath k is assigned coulour w,
(1) ykw =
0 otherwise,
and
20 O. BRUN, S. BARAKETI

Layer 2
B2

A2 C2

A C

B1
B

A1 C1
A C Layer 1
(a) original network. (b) Three-layer network.

Figure 6. The original network and the three layers obtained by duplicat-
ing nodes and links in the case of two colours.

(
1 if lighpath k is routed on link e,
(2) xek =
0 otherwise,
for all e ∈ EW . Since each lighpath has to be routed in a single network layer, we clearly
have
X
(3) ykw = 1, ∀k ∈ K.
w∈W

Obviously, the links of network layer w are not used by lighpath k if this lighpath is not
assigned the colour w, so that

(4) xek ≤ ykw , ∀e ∈ Ew , ∀w ∈ W, ∀k ∈ K.

On the contrary, if lighpath k is assigned colour w, then it has to be routed in network
layer w. Defining, for each node n ∈ Vw , Ew+ (n) (resp. Ew− (n)) as the set of directed edges
of Ew that are incoming (resp. outgoing) at node n, this routing problem can be expressed
using a node-link formulation:
 ROUTING AND WAVELENGTH ASSIGNMENT IN OPTICAL NETWORKS 21


w
X X −yk
 n = sk
(5) xek − xek = ykw n = tk
+

e∈Ew (n) e∈Ew (n) 0 otherwise.
− 

In each network layer, each link can accomodate at most one lighpath, so that we also
have the following constraint:

X
(6) xek ≤ 1, ∀e ∈ Ew , ∀w ∈ W.
k∈K

Constraints (4)-(6) define feasible solutions to the RWA problem. The goal is to minimize
the number of used wavelengths, or, equivalently, the number of used network layers. For
each colour w, define

(
1 if colour w is used
(7) uw =
0 otherwise.

Since colour w is used if at least one lighpath is assigned to it, we clearly have

X
(8) ykw ≤ K uw , ∀w ∈ W,
k∈K

where K is the number of lighpaths. The RWA problem can now be formally defined as an
integer linear program (ILP), as shown below. It has been shown that the RWA problem
is NP-complete in [8]. The proof involves a reduction to the n-graph colorability problem.
In other words, solving the RWA problem is as complex as finding the chromatic number
of a general graph.
22 O. BRUN, S. BARAKETI

(RWA) minimize wmax

subject to
w uw ≤ wmax , ∀w ∈ W,
X
ykw ≤ K uw , ∀w ∈ W,
k∈K
X
xek ≤ 1, ∀e ∈ Ew , ∀w ∈ W,
k∈K
X X
xek − xek = −ykw , ∀k ∈ K, ∀w ∈ W,
+
e∈Ew (sk ) e∈Ew
−
(sk )
X X
xek − xek = ykw , ∀k ∈ K, ∀w ∈ W,
+
e∈Ew (tk ) e∈Ew
−
(tk )
X X
xek − xek = 0, ∀n 6= sk , tk , ∀k ∈ K, ∀w ∈ W,
+
e∈Ew (n) e∈Ew
−
(n)
xk ≤ ykw ,
e
∀e ∈ Ew , ∀w ∈ W, ∀k ∈ K,
X
ykw = 1, ∀k ∈ K,
w∈W
xek ∈ {0, 1}, ∀e ∈ Ew , ∀w ∈ W, ∀k ∈ K,
ykw ∈ {0, 1}, ∀w ∈ W, ∀k ∈ K,
uw ∈ {0, 1}, ∀w ∈ W,
wmax ≥ 0.
We note the following property of optimal solutions of the above problem.
Lemma 2. Let (x, y, u, wmax ) be an optimal solution of problem (RWA). Then

X
(9) ykw ≥ 1, ∀w ≤ wmax ,
k∈K

Proof. Assume on the contrary that there exists q < wmax such that k∈K ykq = 0. Then it
P
is easy to see that we can define another feasible solution (x̂, ŷ, û, q) such that all lighpaths
routed in network layer wmax are now routed in network layer q (with exactly the same
path), while the routes of the other lighpaths are the same in both solutions. The new
solution uses wmax − 1 network layers, which implies that (x, y, u, wmax ) is not an optimal
solution, i.e., a contradiction. 
We observe that, given a routing strategy x satisfying (5)-(6), we can easily obtain a
feasible solution to the above integer linear program. Indeed, if lighpath k is routed in
 ROUTING AND WAVELENGTH ASSIGNMENT IN OPTICAL NETWORKS 23

network layer w, then we can set ykw = 1 and uw = 1. This suggests that the RWA problem
can be formulated as a pure routing problem.

4.2. Formulation as an equivalent routing problem. We shall now show that the
RWA problem can be equivalently formulated as a pure routing problem. Define the
following constants:

(10) cw = K w−1 , w = 1, . . . , W.
Noting that

q
X q−1
X
(K − 1) cw = (K − 1) K n,
w=1 n=0
q
X q−1
X
= Kn − K n,
n=1 n=0
= K q − 1,
we see that these coefficients are such that

q
X q
X
(11) K cw < cw + cq+1 .
w=1 w=1

Let us now consider the following optimization problem:

X X
(EQ-RWA) minimize cw ykw
w∈W k∈K
subject to
X
xek ≤ 1, ∀e ∈ Ew , ∀w ∈ W,
k∈K
X X
xek − xek = hnk (ykw ), ∀n ∈ Vw , ∀k ∈ K, ∀w ∈ W,
+
e∈Ew (n) e∈Ew
−
(n)
xk ≤ ykw ,
e
∀e ∈ Ew , ∀w ∈ W, ∀k ∈ K,
X
ykw = 1, ∀k ∈ K,
w∈W
xek ∈ {0, 1}, ∀e ∈ Ew , ∀w ∈ W, ∀k ∈ K,
ykw ∈ {0, 1}, ∀w ∈ W, ∀k ∈ K,
where we have used the following notations for k ∈ K and n ∈ Vw
24 O. BRUN, S. BARAKETI


w
−yk
 n = sk
(12) hnk (ykw ) = ykw n = tk

0 otherwise.


We have the following result.

Theorem 1. Let (x, y) be any optimal solution of problem (EQ-RWA). Then (x, y, u, wmax )
is an optimal solution of problem (RWA), where
!
X
(13) uw = min 1, ykw , w = 1, . . . , W,
k∈K

and wmax = max (w uw ).

w∈W

Proof. It is clear from the definition of problem (EQ-RWA) that the vectors x and y satisfy
equations (4)-(6). From (13), constraints (8) are also satisfied. Moreover, the definition of
wmax implies that wmax ≥ 0 and wuw ≤ wmax for all w ∈ W. Hence, (x, y, u, wmax ) is a
feasible solution of problem (RWA).
Assume that this solution is not an optimal one, that is that there exists a feasible
solution (x̂, ŷ, û, q) of problem (RWA) such that q ≤ wmax − 1. This clearly implies that
the vector ŷ is such that
X
ŷkw = 0, ∀w > q.
k∈K
Hence,

X X q
X X
cw ŷkw = cw ŷkw ,
w∈W k∈K w=1 k∈K
q
X
≤ K cw ,
w=1
q
X
< cw + cq+1 ,
w=1
w
P
where the first inequality follows from ≤ K, ∀w ∈ W, and the second one follows
k∈K ŷk
from (11). Since q + 1 ≤ wmax and since from Lemma 2
X
ykw ≥ 1, ∀w ≤ wmax ,
k∈K
it yields
 ROUTING AND WAVELENGTH ASSIGNMENT IN OPTICAL NETWORKS 25

Require: G = (V, E), K and W

1: w ← 1
2: while K 6= ∅ do
3: Solve problem (MAX-FLOW-w) for network layer w.
4: K ← K \ {k ∈ K : ykw = 1}
5: w ←w+1
6: end while

Figure 7. Basic approximate algorithm (BA-RWA) to solve optimization

problem (EQ-RWA).

q
ykq+1 ,
X X X X X
cw ŷkw < cw ykw + cq+1
w∈W k∈K w=1 k∈K k∈K
X X
< cw ykw .
w∈W k∈K
We thus conclude that (x, y) is not an optimal solution of problem (EQ-RWA), i.e., a
contradiction. Therefore, (x, y, u, wmax ) is an optimal solution of problem (RWA). 
According to Theorem 1, we can easily obtain an optimal solution of the original problem
from an optimal solution of problem (EQ-RWA). Thus, in the following we shall study
problem (EQ-RWA) instead of studying problem (RWA).
4.3. Solution procedure. For a given network layer w, let us define the following opti-
mization problem:
X
(MAX-FLOW-w) maximize ykw
k∈K
subject to
X
xek ≤ 1, ∀e ∈ Ew ,
k∈K
X X
xek − xek = hnk (ykw ), ∀n ∈ Vw , ∀k ∈ K,
+
e∈Ew (n) e∈Ew
−
(n)
e
xk ∈ {0, 1}, ∀e ∈ Ew , ∀k ∈ K,
ykw ∈ {0, 1}, ∀k ∈ K,
In this section, we propose an heuristic for resolving the (EQ-RWA) problem. The
algorithm of the proposed heuristic is described in figure 7.
In this algorithm, we seek to maximize the use of each network layer before proceeding
to the next layer. The transistion to the next layer is done once the current layer is
saturated. Therefore, it’s clear that no demand of the w-layer can be rerouted or replaced
26 O. BRUN, S. BARAKETI

in the overlying layers. As against, demands placed in the w-layer can be rerouted over
the underlying layers, since these layers may be considered as under-used for some or all
of w-layer demands. Let us note Kw = {k ∈ K : ykw = 1} the set of demands placed on
the w-layer and LR = {λ ∈ W : λ < w} the set of used network layers after solving the
optimization problem (EQ-RWA) whith our proposed heuristic.
From this assumption, we proposed an improved algorithm based on successive approx-
imation method. After a first routing with the algorithm proposed in figure 7, we opt
to decrease the number of used network layers by trying to reroute all demands of each
layer in the others underlying layers. In each w-layer iteration, we seek to find a new rout-
ing strategy where w-layer is no longer used. This algorithm try to reroute each demand
k ∈ Kw over the shortest λ-path where λ ∈ [[w + 1...LR ]]. If there is no available paths for
k in all λ-layers, then the w-layer still used in the final routing strategy.
For a given network layer λ, let us define the following optimization problem for the
demand k:

(SHORTEST λ-PATH-k)
X
minimize ze
e∈Eλ
subject to

X X −1 if n = sk

ze − ze = 1 if n = tk , ∀n ∈ Vλ ,
+

e∈Eλ (n) e∈Eλ (n) 0 otherwise.
− 

ze ≤ δ e , ∀e ∈ Eλ ,
ze ∈ {0, 1}, ∀e ∈ Eλ .
where we have used the following notations for e ∈ Eλ
(
1 if e is λ-available
(14) δe =
0 otherwise.
The enhanced algorithm is described in figure 8
4.4. Experiments and results. In this section, several experiments on the optimization
problem for traffic routing were performed in order to evaluate and validate the effectiness
of the proposed solutions. The GUROBI solver is used for solving the two integer linear
programs: (RWA) and (EQ-RWA). The aim of the experiments here is to analyze the per-
formance of algorithms and compare the proposed methods. The comparison is performed
between the different proposed algorithms: ILP-based algorithm (RWA), equivalent ILP-
based algorithm (EQ-RWA), basic and enhanced algorithms to solve optimization problem
(EQ-RWA) which are illustrated in Fig. 7 and Fig. 8 repectively.
 ROUTING AND WAVELENGTH ASSIGNMENT IN OPTICAL NETWORKS 27

Require: G = (V, E), K and W

1: w ← 1
2: while K 6= ∅ do
3: Solve problem (MAX-FLOW-w) for network layer w.
4: K ← K \ {k ∈ K : ykw = 1}
5: w ←w+1
6: end while
7: LR ← {λ ∈ W : λ < w}
8: for w ∈ LR do
9: Kw ← {k ∈ K : ykw = 1}
10: Kw−tmp ← Kw
11: end for
12: LA ← ∅
13: amelioration ← 0
14: for w ∈ LR do
15: cpt ← 0
16: for k ∈ Kw do
17: rerouted ← f alse
18: for λ = w + 1 . . . |LR | do
19: Solve problem (SHORTEST λ-PATH-k) for network layer λ.
20: if {e ∈ Eλ : ze = 1} = 6 ∅ then
21: Kλ−tmp ← Kλ−tmp ∪ {k}, Kw−tmp ← Kw−tmp \ {k}
22: rerouted ← true, cpt ← cpt + 1
23: break
24: end if
25: end for
26: if rerouted = f alse then
27: for λ = w . . . |LR | do
28: Kλ−tmp ← Kλ
29: end for
30: break
31: end if
32: end for
33: if cpt = |Kw | then
34: LA ← LA ∪ {w}
35: for λ = w . . . |LR | do
36: Kλ ← Kλ−tmp
37: end for
38: end if
39: end for
40: LR ← LR \ LA
41: amelioration ← |LA |

Figure 8. Enhanced approximate algorithm (EA-RWA) to solve optimiza-

tion problem (EQ-RWA).
28 O. BRUN, S. BARAKETI

Table 4. Computation Time for RWA proposed algorithms

Topology Size (RWA) (EQ-RWA) (BA-RWA) (EA-RWA)

|V | = 5, |E| = 12, |W | = 20 0.5s 0.5s 0.2s 0.3s
|V | = 8, |E| = 20, |W | = 20 10s 1m5s 1.6s 1.8s
|V | = 10, |E| = 26, |W | = 30 2m19s 2m43s 4.2s 4.5s
|V | = 15, |E| = 40, |W | = 40 6m52s 8m24s 12.2s 12.8s
|V | = 20, |E| = 56, |W | = 40 13m12s 17m05s 52.2s 53s
|V | = 30, |E| = 86, |W | = 60 38m56s 49m21s 4m12s 4m14s
|V | = 40, |E| = 122, |W | = 60 >2h >2h 9m34s 9m38s
|V | = 50, |E| = 158, |W | = 80 >3h >3h 27m42s 27m53s

Table 5. Memory Consumption for RWA proposed algorithms

Topology Size (RWA) (EQ-RWA) (BA-RWA) (EA-RWA)

|V | = 5, |E| = 12, |W | = 20 5M 6M 5M 6M
|V | = 8, |E| = 20, |W | = 20 80M 85M 9M 11M
|V | = 10, |E| = 26, |W | = 30 280M 295M 14M 16M
|V | = 15, |E| = 40, |W | = 40 520M 545M 18M 20M
|V | = 20, |E| = 56, |W | = 40 940M 980M 26M 28M
|V | = 30, |E| = 86, |W | = 60 1.9G 2G 40M 43M
|V | = 40, |E| = 122, |W | = 60 2.8G >3G 75M 79M
|V | = 50, |E| = 158, |W | = 80 >3G >3G 200M 206M

The performance comparison between the algorithms is done based on two criteria.
The first is related to the execution performance including computation time and memory
consumption. The second one concerns the routing strategy cost which represents in this
case the number of used wavelengths (i.e. colours) for satisfying all routed lightpaths.

4.4.1. Execution performance. In this part, we summarize the run performance of the pro-
posed algorithms on several topologies, with different sizes. The results related to com-
putation time and memory consumption are illustrated in Tables 4 and 5, respectively.
For each simulation, the differents algorithms are executed on the same topology and for
the same set of demands which are randomly generated. Topology size is defined by three
terms |V |, |E| and |W | which represents the number of nodes, the number of fibers and
the number of available wavelengths in each fiber, respectively.
From these results, it’s clear that the approximate algorithms are much more faster, and
consume less memory than the ILP-based ones. A routing strategy solution can be found
in a few seconds for large topology sizes.

4.4.2. Routing strategy cost. In RWA problem, the routing strategy cost represents the
number of used layers (i.e. wavelengths) when routing the lightpaths matrix. This opti-
mization criterion is the same in all our proposed algorithms. In this part, we compare the
 ROUTING AND WAVELENGTH ASSIGNMENT IN OPTICAL NETWORKS 29

Table 6. Routing strategy cost

Matrix size (EQ-RWA) (BA-RWA) (EA-RWA)

89 13λ 14λ 13λ
86 14λ 15λ 14λ
104 16λ 16λ 16λ
97 14λ 14λ 14λ
80 13λ 14λ 13λ
81 14λ 15λ 14λ
82 16λ 16λ 16λ
98 15λ 16λ 15λ
83 14λ 14λ 14λ
92 13λ 14λ 14λ
169 25λ 25λ 25λ
174 28λ 28λ 28λ
187 27λ 28λ 28λ
158 22λ 23λ 22λ
174 26λ 27λ 26λ
168 27λ 27λ 27λ
172 27λ 27λ 27λ
165 23λ 24λ 24λ
169 28λ 29λ 29λ
177 26λ 27λ 27λ

quality of routing ensured by the differents solutions. For that, it suffices to compare their
routing strategy costs. For this comparison, a simplistic topology with 10 nodes, 26 fibers
and 40 available wavelengths is considered. Twenty simulations have been performed on
this network. For each simulation, the differents algorithms try to route the same set of
demands, which are randomly generated. Table 6 illustrate the obtained results and figure
9 represents the curve of the relative error between the ILP-based algorithm (EQ-RWA)
solution (i.e. optimal solution) and the (EA-RWA) one (i.e. approximate solution), along
the 20 simulations. The matrix size here represents the total number of lightpath demands.

The obtained results clearly show the efficiency of the proposed enhanced heuristic.
Among the twenty simulations, there are only five times where we don’t reach the optimal
solution. In addition, the (EA-RWA) algorithm provides an improved solution compared to
the (BA-RWA) algorithm. In several simulations, we succeeded in enhancing the strategy
routing cost returned by the (BA-RWA) thanks to our enhanced algorithm.

4.5. New heuristic as an equivalent to ”Maximum Multicommodity Flow” to

solve the (MAX-FLOW-w) problem.
30 O. BRUN, S. BARAKETI

Figure 9. Relative error between approximate and optimal solutions

4.5.1. Adopted Approach. As shown in the previous sections, our approximation algorithm
(BA-RWA) is based on the linear program (MAX-FLOW-w). This program try to route
the maximum number of demands (i.e. lightpaths) over the w-layer. It ensures an optimal
solution per network w-layer, but it can be time-limited and memory-limited for resolv-
ing the RWA problem over real networks. For this, we have opted to finding an efficient
heuristic that solves the same (MAX-FLOW-w) problem.

Since we search to maximize the number of routed lightpaths on the w-layer given that
each link (i.e. fiber) can be used only once, the problem is equivalent to a Maximum
Multicommodity Problem with edges capacities equal to 1. Commodity pairs represents
the set of pairs (sk ,tk ) of lightpaths for k ∈ K.
A multicommodity flow problem is defined on a directed network G = (V, E) with
capacities u(e) for e ∈ E and p source-target pairs (sj ,tj ) for 1 ≤ j ≤ p. The problem
is to find flows fj from sj to tj that satisfy flow conservation constraints and capacity
constraints that ensures that the sum of flows on any edge does not exceed the capacity of
the edge. For the maximum
P multicommodity flow problem, the objective is to maximize
the sum of the flows j |fj |.
While there are many different algorithms known for this problem we discuss one which
guarantee an (1 + ω)-approximation to the maximum throughtput. The goal of the Garg-
Knemann algorithm is to find an ω-approximate solution for any error parameter ω > 0.
the ω-approximate solution is a flow that has value at least (1 − ω) times the maximum
 ROUTING AND WAVELENGTH ASSIGNMENT IN OPTICAL NETWORKS 31

Require: network G = (V, E), capacities u(e), commodity pairs (sj , tj ) for 1 ≤ j ≤ p,
accuracy ǫ
1: Initialize l(e) = φ ∀e, x = 0
P
2: while there is a P ∈ P with e⊂P l(e) < 1 do
3: Select the minimum length path P ∈ P
4: u ← mine⊂P u(e)
5: x(P ) ← x(P ) + u
ǫu
6: ∀e ∈ P, l(e) ← l(e)(1 + u(e) )
7: end while
8: Return (x, l).

Figure 10. Generic Garg-Knemann algorithm.

value. In addition, the running time of this type of algorithms is known depending on the
approximation value ω, the number of commodity pairs p and the number of edges m.
This generic algorithm for maximum multicommodity flow, that we adopted here, try
to find mutually two solutions for the primal and dual programs for the maximum multi-
commodity flow problem. Let Pj denote the set of paths from sj to tj , and P := ∪j Pj .
x(P ) equals the amount of flow sent along path P and represents the decision variable of
the primal program. The length of an edge l(e) represents the decision variable of the
dual program and reflects the marginal cost of using an additional unit of capacity of the
edge. The generic Garg-Knemann algorithm, described in figure 10, stops when the ratio
between the primal and the dual solutions is at most 1 + ω. The accuracy ǫ is a fixed
constant that choosen appropriately depending on ω. φ represents the inial length of all
network edges which depends on ǫ and m.
The Garg-Knemann algorithm solves the maximum multicommodity flow problem with
a guarantee of error approximation. The real flow can be deducted by scaling the final
flow obtained in the generic algorithm by log1+ǫ 1+ǫ φ . But the generated flow is generally
fractional which is not adapted in our study case where we search to route a set of lightpaths
(i.e. flows in the case of maximum multicommodity flow problem). Therefore, the solution
must be integer or at least quasi-integer. For this, we have proposed a modification in Garg-
Knemann algorithm in order to converge to a quasi-integer solution: Since in each iteration
we search the minimum length path and the algorithm augments flow along this path, so we
can use different paths in two successive iterations. Our idea consists on keeping the choosen
minimum length path and augmenting flow along it until the ”real” flow converges to 1.
Once this condition is satisfied, the algorithm can switch to another minimum length path
and repeat the same treatment. The modified Garg-Knemann algorithm is described in
figure 11. The proposed modification allows a primal quasi-integer solution. This solution
can be easily rounded to an integer solution.
The Garg-Knemann algorithm try to maximize the flow sent between the source and the
taget of each commodity pair. Hence, the calculated flow for a commodity j can exceed
the required amount of demand (i.e. number of lightpaths) between sj and tj . This does
32 O. BRUN, S. BARAKETI

Require: network G = (V, E), capacities u(e), commodity pairs (sj , tj ) for 1 ≤ j ≤ p,
amount of demands tj for 1 ≤ j ≤ p, accuracy ǫ, threshold α (close to 1)
1: Initialize l(e) = φ ∀e, x = 0
2: iter ← 1
P
3: while there are a commodity j ∈ [[1...p]] and a path P ∈ Pj : e⊂P l(e) <
P
1 and Q∈Pj x(Q) < t j do
P
4: if iter = 1 or x(P ) ≥ α or Q∈Pj x(Q) ≥ tj then
5: Select the
P minimum length path P ∈ P and the commodity j : P ∈
Pj and Q∈Pj x(Q) < tj
6: end if
7: u ← mine⊂P u(e)
8: x(P ) ← x(P ) + log u 1+ǫ
1+ǫ φ
ǫu
9: ∀e ∈ P, l(e) ← l(e)(1 + u(e) )
10: iter ← iter + 1
11: end while
12: Return (x, l).

Figure 11. Modified Garg-Knemann algorithm.

not match with our expectation. Consequently, we have added an additional capacity
constraints for commodity pairs. For each commodity j, we have assigned a constant tj
whichP is equal to the required demand between sj and tj . The total flow for commodity
j (= Q∈Pj x(Q))) must not exceed the value of tj . Adopting this modification, we can
guarantee that the maximum flow sent between sj and tj can not exceed the required
number of lightpaths for the commodity j.
After applying our updates, we can adopt this modified algorithm for solving the (MAX-FLOW-w)
problem.

4.5.2. Experiments and results. several experiments on the optimization problem (MAX-FLOW-w)
were performed in order to evaluate and validate the effectiness of the proposed solutions.
The GUROBI solver is used for solving the integer linear program (MAX-FLOW-w). The
aim of the experiments here is to analyze the performance of algorithms and compare the
proposed methods. The comparison is performed between the ILP-based algorithm and
the modified Garg-Knemann algorithm after its adaptation to our optimization case. We
compare her the optimal solution generated by the ILP-based algorithm and the approx-
imate solution deducted by the modified Garg-Knemann algorithm. For this comparison,
three simplistic topologies are considered. Ten simulations have been performed on each
network. For each simulation, the two algorithms try to route the same set of demands
(i.e. lightpaths), which are randomly generated. Tables 7, 8 and 9 illustrate the obtained
results for the three studied topologies. the fixed ω for the Garg-Knemann algorithm is
0, 05.
 ROUTING AND WAVELENGTH ASSIGNMENT IN OPTICAL NETWORKS 33

Table 7. Topology 1 : 6 nodes and 14 links

Number of Optimal Approximate solution Approximate solution Rounded solution

demands solution (shortest-path-first) (Modified Garg-Knemann) (Modified Garg-Knemann)
40 13 13 12,9247 13
0.4s 0.009s 0.04s
42 9 9 8,9894 9
0.4s 0.008s 0.02s
48 10 10 9,9893 10
0.5s 0.007s 0.02s
41 12 12 11,9707 12
0.4s 0.008s 0.03s
39 9 9 8,9986 9
0.4s 0.007s 0.02s
49 10 10 9,9985 10
0.6s 0.009s 0.03s
34 10 10 9,9618 10
0.4s 0.007s 0.02s
55 11 11 10,9892 11
0.5s 0.008s 0.02s
47 11 11 10,9617 11
0.4s 0.007s 0.02s
42 12 12 11,9707 12
0.4s 0.008s 0.03s

4.6. New heuristic as an equivalent to ”Maximum Edge-Dijoint Paths” to solve

the (MAX-FLOW-w) problem.

4.6.1. Adopted Algorithm. As shown in the previous section, we can solve the (MAX-FLOW-w)
problem effectively using a modified version of the Maximum Multicommodity Flow Garg-
Knemann algorithm. the adopted algorithm try to route the maximum amount of flows
over a network w-layer. The obtained solution is generally fractionnal, but we can easily
provide a rounded feasible solution.
Since a flow unit represents a lightpath and a flow unit path can be used only once
because the capacity of every edge in w-layer is 1, the problem can be distinguish as
a ”Maximum Edge-Dijoint Paths” (MEDP). The goal is to route as many requests as
possible along edge-dijoint paths. The input consists of a directed graph G = (V, E) and a
set containing k requests K = {(si , ti )|i = 1 . . . k}, where each request is a pair of vertices
in V . The request (si , ti ) asks for a directed path from si to ti in G. A feasible solution is
given by a subset A and an assignment of edge-disjoint paths to all requests in that subset.
More precisely, each (si , ti ) ∈ A must be assigned a directed path Pi from si to ti in G
such that no two paths Pi and Pj (for i, j ∈ A and i 6= jhave a directed edge of the graph
34 O. BRUN, S. BARAKETI

Table 8. Topoogy 2 : 15 nodes and 38 links

Number of Optimal Approximate solution Approximate solution Rounded solution

demands solution (shortest-path-first) (Modified Garg-Knemann) (Modified Garg-Knemann)
263 27 27 26,9701 27
1.5s 0.02s 0.1s
308 31 31 30,9555 31
1.3s 0.026s 0.2s
328 31 31 30,9492 31
1.3s 0.027s 0.2s
315 28 27 26,9618 27
1.3s 0.021s 0.1s
289 31 31 30,9683 31
1.4s 0.026s 0.2s
316 30 29 28,9971 29
1.3s 0.024s 0.1s
312 30 30 29,9349 30
1.3s 0.024s 0.1s
304 34 34 33,9031 34
1.3s 0.026s 0.2s
295 31 31 30,9459 31
1.5s 0.024s 0.1s
211 26 26 25,9602 26
1.3s 0.021s 0.1s

in common. The requests in A are called the accepted requests. We search to maximize
the cardinality of A.
The MEDP problem has been studied in many works and solved by several algorithms.
One of these algorithms is the Shortest-Path-First suggested by Kolliopoulos and Stein.
The algorithm is illustrated in figure 12

4.7. Comparison between (RWA) solution and approximate solution using the
new heuristic. In this section, we compare the performance of the three algorithms :
the (RWA) ILP-based algorithm, the (BA-RWA) algorithm and the (EA-RWA) algorithm.
For the two last algorithms, illustrated in figure 7 and 8, the (MAX-FLOW-w) ILP-based
algorithm is remplaced by our new heuristic ”Modified Garg-Knemann algorithm” shown
if figure 11. We adopt here the same network model (10 nodes, 26 links and 40 available
wavelengths) and the same set of simulations we used in the section 4.4.2. We recall that
we compare the routing strategy costs for the three algorithms. The routing stategy cost
represents the number of used layers (i.e. wavelengths) for routing all lightpath demands.
Table 10 represents the results among twenty simulations and figure 13 show the relative
error between the optimal solution and the approximate (EA-RWA) solution.
 ROUTING AND WAVELENGTH ASSIGNMENT IN OPTICAL NETWORKS 35

Table 9. Topoogy 3 : 34 nodes and 92 links

Number of Optimal Approximate solution Approximate solution Rounded solution

demands solution (shortest-path-first) (Modified Garg-Knemann) (Modified Garg-Knemann)
1128 76 76 75,8519 76
5.4s 0.208s 2.3s
1119 72 72 71,8396 72
6.2s 0.202s 2.1s
1091 70 70 69,8849 70
5.8s 0.193s 1.9s
1089 71 71 70,8561 71
6.3s 0.194s 1.9s
1143 73 72 71,8837 72
5.6s 0.209s 2.1s
1103 71 70 69,8503 70
6.1s 0.196s 1.9s
1126 70 70 69,8529 70
5.5s 0.194s 1.9s
1108 73 73 72,8785 73
6.4s 0.207s 2.1s
1116 67 66 65,8944 66
5.2s 0.186s 1.8s
1058 65 63 62,8864 63
5.1s 0.177s 1.5s

Require: network G = (V, E), K = {(si , ti )|i = 1 . . . k}

1: A ← ∅
2: while K contains a request that can be routed in G do
3: Select a request (si , ti ) in K such that the shortest path from si to ti in G has
minimum length among all requests in K
4: A ← A ∪ {(si , ti )}
5: K ← K \ {(si , ti )}
6: Pi ← a shortest path from si to ti in G
7: Remove all edges of Pi from G
8: end while
9: Return A and {Pi |(si , ti ) ∈ A}.

Figure 12. Shortest-Path-First algorithm.

Computation time and memory consumption are also compared for the three studied
algorithms. we summarize here the run performance of the proposed algorithms on several
topologies, with different sizes. The results related to computation time and memory
36 O. BRUN, S. BARAKETI

Table 10. Routing strategy cost : Comparison between optimal solution

and approximate ones

Number of (RWA) or (EQ-RWA) (BA-RWA) → (EA-RWA) (BA-RWA) → (EA-RWA)

lightpaths (based on Modified Garg-Knemann) (based on Shortest-path-first)
89 13λ 13λ → 13λ 13λ → 13λ
86 14λ 15λ → 15λ 15λ → 15λ
104 16λ 17λ → 16λ 17λ → 16λ
97 14λ 14λ → 14λ 14λ → 14λ
80 13λ 14λ → 13λ 14λ → 14λ
81 14λ 15λ → 14λ 15λ → 14λ
82 16λ 16λ → 16λ 16λ → 16λ
98 15λ 16λ → 16λ 16λ → 16λ
83 14λ 15λ → 14λ 15λ → 14λ
92 13λ 14λ → 14λ 14λ → 14λ
169 25λ 27λ → 25λ 27λ → 25λ
174 28λ 30λ → 29λ 30λ → 29λ
187 27λ 29λ → 29λ 29λ → 29λ
158 22λ 23λ → 23λ 24λ → 23λ
174 26λ 29λ → 28λ 29λ → 28λ
168 27λ 27λ → 27λ 27λ → 27λ
172 27λ 28λ → 28λ 28λ → 28λ
165 23λ 25λ → 24λ 25λ → 24λ
169 28λ 30λ → 30λ 30λ → 30λ
177 26λ 28λ → 27λ 28λ → 27λ

consumption are illustrated in Tables 11 and 12, respectively. For each simulation, the
differents algorithms are executed on the same topology and for the same set of demands
which are randomly generated. Topology size is defined by three terms |V |, |E| and |W |
which represents the number of nodes, the number of fibers and the number of available
wavelengths in each fiber, respectively.

References
[1] R. K. Ahuja, T. L. Magnanti, and J. B. Orlin. Network Flows. Prentice Hall, 1993.
[2] D. Banerjee and B. Mukherjee. A practical approach for routing and wavelength assignment in large
wavelength-routed optical networks. Selected Areas in Communications, IEEE Journal on, 14(5):903–
908, 1996.
[3] S. Baroni, P. Bayvel, and R. J. Gibbens. On the number of wavelengths in arbitrarily-connected
wavelength-routed optical networks. University of Cambridge, Statistical Laboratory Research Report
1998-7, http://www. statslab. cam. ac. uk/reports/1998/1998-7. pdf, 1998.
[4] R. Barry and S. Subramaniam. The max sum wavelength assignment algorithm for wdm ring networks.
In Optical Fiber Communication. OFC 97., Conference on, pages 121–122. IEEE, 1997.
 ROUTING AND WAVELENGTH ASSIGNMENT IN OPTICAL NETWORKS 37

Figure 13. Relative error between optimal and (EA-RWA) approximate

(based on Modified Garg-Knemann algorithm) solutions

Table 11. Computation Time : Comparison between ILP-based algorithm

and proposed heuristics

Topology Size (RWA) (BA-RWA) → (EA-RWA) (BA-RWA) → (EA-RWA)

(based on Modified Garg-Knemann) (based on Shortest-path-first)
|V | = 5, |E| = 12, |W | = 20 0.5s 0.04s → 0.06s 0.002s → 0.003s
|V | = 8, |E| = 20, |W | = 20 10s 0.2s → 0.23s 0.023s → 0.026s
|V | = 10, |E| = 26, |W | = 30 2m19s 0.4 → 0.45s 0.042s → 0.046s
|V | = 15, |E| = 40, |W | = 40 6m52s 0.9s → 1.05s 0.09s → 0.1s
|V | = 20, |E| = 56, |W | = 40 13m12s 2.1s → 2.3s 0.25s → 0.27s
|V | = 30, |E| = 86, |W | = 60 38m56s 8.5s → 8.8s 0.92s → 0.98s
|V | = 40, |E| = 122, |W | = 60 >2h 17.8s → 18.3s 1.86s → 1.92s
|V | = 50, |E| = 158, |W | = 80 >3h 39.4s → 40.2s 3.76s → 4.38s

[5] D. Cavendish, A. Kolarov, and B. Sengupta. Routing and wavelength assignment in wdm mesh net-
works. In Global Telecommunications Conference, 2004. GLOBECOM’04. IEEE, volume 2, pages 1016–
1022. IEEE, 2004.
[6] B. Chen, G. N. Rouskas, and R. Dutta. On hierarchical traffic grooming in wdm networks. IEEE/ACM
Transactions on Networking (TON), 16(5):1226–1238, 2008.
[7] I. Chlamtac, A. Ganz, and G. Karmi. Lightpath communications: An approach to high bandwidth
optical wan’s. Communications, IEEE Transactions on, 40(7):1171–1182, 1992.
38 O. BRUN, S. BARAKETI

Table 12. Memory Consumption : Comparison between ILP-based algo-

rithm and proposed heuristics

Topology Size (RWA) (BA-RWA) → (EA-RWA) (BA-RWA) → (EA-RWA)

(based on Modified Garg-Knemann) (based on Shortest-path-first)
|V | = 5, |E| = 12, |W | = 20 5M 3M → 3.4M 2.6M → 2.9M
|V | = 8, |E| = 20, |W | = 20 80M 5.2M → 5.7M 4.5M → 4.9M
|V | = 10, |E| = 26, |W | = 30 280M 8.6M → 9.3M 8M → 8.6M
|V | = 15, |E| = 40, |W | = 40 520M 10M → 11.5M 9M → 10.2M
|V | = 20, |E| = 56, |W | = 40 940M 14M → 16M 13M → 14.6M
|V | = 30, |E| = 86, |W | = 60 1.9G 24M → 26M 21M → 22.8M
|V | = 40, |E| = 122, |W | = 60 2.8G 42M → 44M 40M → 42M
|V | = 50, |E| = 158, |W | = 80 >3G 85M → 88M 81M → 84M

Figure 14. Relative error between optimal and (EA-RWA) approximate

(based on shortest-path-first algorithm) solutions

[8] I. Chlamtac, A. Ganz, and G. Karmi. Lightpath communications: an approach to high bandwidth
optical wan’s. IEEE Transactions on Communications, 40(7):1171–1182, July 1992.
[9] K. Christodoulopoulos, K. Manousakis, and E. Varvarigos. Comparison of routing and wavelength
assignment algorithms in wdm networks. In Global Telecommunications Conference, 2008. IEEE
GLOBECOM 2008. IEEE, pages 1–6. IEEE, 2008.
[10] K. Christodoulopoulos, K. Manousakis, and E. Varvarigos. Offline routing and wavelength assignment
in transparent wdm networks. Networking, IEEE/ACM Transactions on, 18(5):1557–1570, 2010.
 ROUTING AND WAVELENGTH ASSIGNMENT IN OPTICAL NETWORKS 39

[11] P. Cochrane, R. Heckingbottom, and D. Heatley. The hidden benefits of optical transparency. Com-
munications Magazine, IEEE, 32(9):90–97, 1994.
[12] A. Ganz and X. Wang. Efficient algorithm for virtual topology design in multihop lightwave networks.
Networking, IEEE/ACM Transactions on, 2(3):217–225, 1994.
[13] N. Garg and J. Koenemann. Faster and simpler algorithms for multicommodity flow and other fractional
packing problems. SIAM Journal on Computing, 37(2):630–652, 2007.
[14] A. Jaekel, A. Bari, Y. Chen, and S. Bandyopadhyay. New techniques for efficient traffic grooming in
wdm mesh networks. In Computer Communications and Networks, 2007. ICCCN 2007. Proceedings of
16th International Conference on, pages 303–308. IEEE, 2007.
[15] B. Jaumard, C. Meyer, and B. Thiongane. Ilp formulations for the routing and wavelength assignment
problem: Symmetric systems. In Handbook of optimization in telecommunications, pages 637–677.
Springer, 2006.
[16] B. Jaumard, C. Meyer, and B. Thiongane. On column generation formulations for the rwa problem.
Discrete Applied Mathematics, 157(6):1291–1308, 2009.
[17] G. Jeong and E. Ayanoglu. Comparison of wavelength-interchanging and wavelength-selective cross-
connects in multiwavelength all-optical networks. In INFOCOM’96. Fifteenth Annual Joint Conference
of the IEEE Computer Societies. Networking the Next Generation. Proceedings IEEE, volume 1, pages
156–163. IEEE, 1996.
[18] E. Karasan and E. Ayanoglu. Effects of wavelength routing and selection algorithms on wavelength
conversion gain in wdm optical networks. Networking, IEEE/ACM Transactions on, 6(2):186–196,
1998.
[19] N. Karmarkar and R. M. Karp. An efficient approximation scheme for the one-dimensional bin-packing
problem. In Foundations of Computer Science, 1982. SFCS’08. 23rd Annual Symposium on, pages 312–
320. IEEE, 1982.
[20] K. Lee, K. C. Kang, T. Lee, and S. Park. An optimization approach to routing and wavelength
assignment in wdm all-optical mesh networks without wavelength conversion. ETRI journal, 24(2):131–
141, 2002.
[21] T. Leighton, F. Makedon, S. Plotkin, C. Stein, É. Tardos, and S. Tragoudas. Fast approximation
algorithms for multicommodity flow problems. Journal of Computer and System Sciences, 50(2):228–
243, 1995.
[22] L. Li and A. K. Somani. Dynamic wavelength routing using congestion and neighborhood information.
IEEE/ACM Transactions on Networking (TON), 7(5):779–786, 1999.
[23] Z. Liu and G. N. Rouskas. A fast path-based ilp formulation for offline rwa in mesh optical networks.
In Global Communications Conference (GLOBECOM), 2012 IEEE, pages 2990–2995. IEEE, 2012.
[24] Z. Liu and G. N. Rouskas. Link selection algorithms for link-based ilps and applications to rwa in mesh
networks. In Optical Network Design and Modeling (ONDM), 2013 17th International Conference on,
pages 59–64. IEEE, 2013.
[25] P. Manohar, D. Manjunath, and R. Shevgaonkar. Routing and wavelength assignment in optical net-
works from edge disjoint path algorithms. Communications Letters, IEEE, 6(5):211–213, 2002.
[26] B. Mukherjee, D. Banerjee, and S. Ramamurthy. Some principles for designing a wide-area wdm optical
network. Networking, IEEE/ACM Transactions on, 4(5):684–696, 1996.
[27] A. E. Ozdaglar and D. P. Bertsekas. Routing and wavelength assignment in optical networks.
IEEE/ACM Transactions on Networking (TON), 11(2):259–272, 2003.
[28] S. Ramamurthy and B. Mukherjee. Fixed-alternate routing and wavelength conversion in wavelength-
routed optical networks. In Global Telecommunications Conference, 1998. GLOBECOM 1998. The
Bridge to Global Integration. IEEE, volume 4, pages 2295–2302. IEEE, 1998.
[29] R. Ramaswami. Multiwavelength lightwave networks for computer communication. Communications
Magazine, IEEE, 31(2):78–88, 1993.
[30] R. Ramaswami, K. Sivarajan, and G. Sasaki. Optical networks: a practical perspective. Morgan Kauf-
mann, 2009.
40 O. BRUN, S. BARAKETI

[31] R. Ramaswami and K. N. Sivarajan. Routing and wavelength assignment in all-optical networks.
IEEE/ACM Transactions on Networking (TON), 3(5):489–500, 1995.
[32] V. S. Shekhawat, D. K. Tyagi, and V. Chaubey. Weight based edge disjoint path routing and wavelength
assignment (wedp-rwa) algorithm for wdm networks. In Industrial and Information Systems, 2008.
ICIIS 2008. IEEE Region 10 and the Third international Conference on, pages 1–5. IEEE, 2008.
[33] M. Shiva Kumar and P. Sreenivasa Kumar. Static lightpath establishment in wdm networksnew ilp
formulations and heuristic algorithms. Computer Communications, 25(1):109–114, 2002.
[34] H. Simonis. Solving the static design routing and wavelength assignment problem. In Recent Advances
in Constraints, pages 59–75. Springer, 2011.
[35] H. Siregar, H. Takagi, and Y. Zhang. Efficient routing and wavelength assignment in wavelength-routed
optical networks. In Proc. 7th Asia-Pacific Network Oper. and Mgmt Symposium, pages 116–127, 2003.
[36] N. Skorin-Kapov. Routing and wavelength assignment in optical networks using bin packing based
algorithms. European Journal of Operational Research, 177(2):1167–1179, 2007.
[37] T. Stern and K. Bala. Multiwavelength optical networks: A layered approach addison-wesley. Reading,
Mass, 1999.
[38] S. Subramaniam and R. A. Barry. Wavelength assignment in fixed routing wdm networks. In Com-
munications, 1997. ICC’97 Montreal, Towards the Knowledge Millennium. 1997 IEEE International
Conference on, volume 1, pages 406–410. IEEE, 1997.
[39] H. Zang, J. P. Jue, B. Mukherjee, et al. A review of routing and wavelength assignment approaches
for wavelength-routed optical wdm networks. Optical Networks Magazine, 1(1):47–60, 2000.
[40] H. Zang, J. P. Jue, L. Sahasrabuddhe, R. Ramamurthy, and B. Mukherjee. Dynamic lightpath es-
tablishment in wavelength routed wdm networks. Communications Magazine, IEEE, 39(9):100–108,
2001.
[41] X. Zhang and C. Qiao. Wavelength assignment for dynamic traffic in multi-fiber wdm networks. In
Computer Communications and Networks, 1998. Proceedings. 7th International Conference on, pages
479–485. IEEE, 1998.
[42] Z. Zhang and A. S. Acampora. A heuristic wavelength assignment algorithm for multihop wdm networks
with wavelength routing and wavelength re-use. Networking, IEEE/ACM Transactions on, 3(3):281–
288, 1995.