This article has been accepted for publication in IEEE Photonics Journal.

This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/JPHOT.2022.3205412

An iterative optimization approach for routing,

modulation, and categorical spatial bandwidth block
allocation to improve network performance for
dynamic traffic in elastic optical networks
Akhtar Nawaz Khan?,1 , Hassan Yousif Ahmed2 , Abdellah Chehri?,3 , Medien Zeghid4 ,
Zawar H. Khan5 , Abid Iqbal5
1,5
Department of Electrical Engineering, University of Engineering & Technology, Peshawar, Jalozai
Campus, Pakistan.
2,4
Electrical Engineering Department, College of Engineering at Wadi Aldwaseer,
Prince Sattam Bin Abdulaziz University, KSA.
4
Electronics and Micro-Electronics Laboratory (E. µ. E. L), Faculty of Sciences,
University of Monastir, Tunisia.
3
Department of Applied Sciences, University of Quebec in Chicoutimi (UQAC), Chicoutimi, QC G7H
2B1, Canada.

Abstract— In this work, we have proposed an iterative I. I NTRODUCTION

optimization model for allocating spectral resources in
optical networks. The proposed model gives spatial routes
and spatial bandwidth allocations in optical networks with
variable data-rates, modulation schemes, and optical reach
adaptation. We have also proposed an algorithm which
allocates continuous and contiguous block of frequency
slots (FS) between transponders which forms bandwidth
partitions. The primary objective of the bandwidth par-
tition is to reduce spatial fragmentation. The integrated
approach includes the routing information from using the
optimization model and the categorical spectrum alloca-
tion from using the proposed algorithm. The integrated
approach has been used for dynamic traffic to improve
network performance in terms of bandwidth blocking, link
utilization, and fragmentation metrics. It has been shown
that the FS utilization (FSU) and link utilization (LU)
largely increase in the proposed integrated scheme with
80% LU compared to shortest path first (SPF) routing
with LU as low as 20%. Similarly, the standard deviation
between FSU in the proposed scheme is approximately 5%
compared to 25% in other schemes which shows that the
FSU sufficiently increases in the integrated approach.
Index Terms - Resource allocation, elastic op- its ability to allocate heterogeneous bandwidths to
tical networks, spatial fragmentation, network par- lightpath demands.
tition, iterative optimization.
Manuscript received June 6, 2022. Corresponding
authors: 1,3 Akhtar Nawaz Khan, Abdellah Chehri (Email: called frequency slots (FS). In EON, multiple FS’s
akhtar.nawaz@uetpeshawar.edu.pk; abdellah Chehri@uqac.ca)
Dense wavelength division multiplexing
(DWDM) based systems with a capacity of
40 Gbps were earlier deployed in optical networks
before 2012 to support high data rate traffic [1]. It
was extended to 100 Gbps with QPSK deployment
in 2013 and was again extended to 200 Gbps with
16QAM, 150 Gbps with 8QAM, and 100 Gbps
with QPSK in 2015 [1]. Since 2020, network
interfaces are available to support network traffic
at 400 Gbps with 16QAM, 300 Gbps with 8QAM,
and 200 Gbps with QPSK [2]–[4]. However, the
behavior of network traffic is rapidly changing
and it is becoming more dynamic [5]. Networks
are facing new challenges with an objective to
achieving high efficiency, more flexibility to
adapt with dynamic data rates, and scalability to
maintain network performance [6]. Elastic optical
networking (EON) is becoming the most favorable
candidate to address the network challenges with
In EON, the entire spectrum is divided into a
large number of narrow bandwidth channels with
bandwidths equal to 12.5 GHz [7]–[9] which are
may be allocated to establish a ligthpath between

This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
 This article has been accepted for publication in IEEE Photonics Journal. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/JPHOT.2022.3205412
the two end nodes. However, these FS’s must be
closed neighbors with each other which is called
contiguity [10]. EON utilizes optical orthogonal fre-
quency division multiplexing (O-OFDM) in which
the frequency bands are orthogonal to each other.
It is flexible to adapt signal modulations including
quadrature phase shift keying (QPSK), 8/16/32/64
quadrature amplitude modulation (QAM), and bi-
nary PSK (BPSK) [6]. A higher modulation scheme
has more number of bits and requires smaller band-
widths to transmit information signal. However,
the optical reach of a higher modulation scheme
is lower due to lower optical signal-to-noise ratio
(OSNR) and higher bit-error-rate (BER) [11]. This
means that the higher order modulation schemes
will provide faster data rates at the cost of higher
BER, lower OSNR, and lower optical reach. In
EON, optical interfaces can dynamically adapt data
rates according to the modulation format, optical
reach, and the amount of data to be transmitted
between end nodes.
Routing and spectrum allocation (RSA) is a
popular research area in EON to allocate limited
resources in an optimal way [12], [13]. In RSA, a
lightpath is established between the two transpon-
ders through a selected route and allocating contigu-
ous and continuous FS’s to support traffic demand.
The later is termed as the continuity constraint
such that similar indexed FS’s must be assigned
during lightpath establishment on all links along the
fiber path. The other is termed as the contiguity
constraint such that allocated FS’s along the path
must be closed neighbors with each other. RSA is
mostly categorized in to the routing subproblem and
spectrum allocation subproblem.
Routing subproblem may either be static or dy-
namic to support static or dynamic lightpath re-
quests respectively. For dynamic traffic and net-
work operation, shortest path first routing with k
paths using the Dijkstra’s algorithm are mostly
utilized. However, these paths are computed offline
which are utilized for dynamic network operation.
Hence, k SPF routing may over-utilize resources
on some links while under-utilize other resources.
Therefore, it degrades network performance for dy-
namic traffic. There are algorithms [14], [15] which
are adaptive in nature and compute paths between
transponders at the time of lightpath establish-
ment to support dynamic network operation. Adap-
tive routing [16], [17] tries to establish lightpaths
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
between transponders using link-state information
and using all possible routes between transponders.
Therefore, it provides lower bandwidth blocking
probability compared to k SPF routing at the cost of
lightpaths set up times. These algorithms are highly
time consuming and require extensive support to
continuously update information of routing tables at
the transponders. Finally, adaptive routing suits well
centralized implementation with common network
information compared to the distributed approach.
Spectrum (i.e., FS) allocation (SA) subproblems
in EON are mostly dynamic and are utilized for
online traffic. A continuous and contiguous band
of FS’s are allocated using SA algorithms to light-
path requests after finding suitable routes between
transponders. These SA algorithms include first-fit,
last-fit, random-fit, first-last-fit, and exact-fit algo-
rithms [18]–[20]. In the first-fit, a contiguous band
of slots with minimum possible indexes and satisfy-
ing the continuity constraint are allocated to a light-
path request. The number of allocated slots depends
on the adapted modulation scheme. The last-fit is a
reversal to the first-fit scheme or algorithm in which
a contiguous band of slots with maximum possible
indexes and satisfying the continuity constraint are
allocated to a lightpath request. First-last-fit is a
mix strategy in which the spectrum is divided into
two subgroups. The sub-grouping is performed in
several ways. In [21], this sub-grouping is based
on disjoint path routing and non-disjointed path
routing. In [21], lightpaths are established using
disjointed path routing with first-fit SA scheme and
non-disjointed path routing is followed by last-fit
SA scheme to establish lightpaths. Similarly, in
random-fit algorithm, in case of multiple available
contiguous band of slots, one contiguous band of
FS’s are randomly selected which also satisfies the
continuity constraint. Finally in exact-fit algorithm,
the spectrum is searched to find availability of the
exact number of FS’s for lightpath establishment.
In conclusion, the objective of a best possible SA
algorithm is to allocate resources with minimize
possible spectrum fragmentation.
In EON, sufficient number of FS may be available
on a link or path with combined bandwidth equal
to or more than the requested bandwidth for a
lightpath establishment. However, these FS’s may
not be contiguous or closed neighbors with each
other. Hence, these dispersed FS’s without shared
boundaries cause fragmentation in a fiber link which
 This article has been accepted for publication in IEEE Photonics Journal. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/JPHOT.2022.3205412
greatly degrades the performance of EONs with
heterogeneous data rates and their adapted modu-
lation schemes. Generally speaking, there are two
categories of fragmentation in a network which are
called vertical and horizontal fragmentation. Vertical
fragmentation is called spectral fragmentation which
occurs in a single fiber link. Horizontal fragmen-
tation is called spatial fragmentation which occurs
along a selected path between transponders [22].
The spectral and spatial fragmentation in an EON
is demonstrated in [23]. Furthermore, it is reported
in [24] that partitioning may improve lightpath
establishment probability and hence, will reduce
the amount of bandwidth blocking in a network.
However, the number of partitions should be mini-
mized to improve the performance [22]. There are
different fragmentation measurement metrics which
include external fragmentation metric (EFM) and
access blocking probability metric (ABPM). The
time complexity of EFM is lowest as it focuses on
the maximum size of contiguous slots in spectral
fragmentation and spatial fragmentation. However,
the time complexity of ABPM is higher and it will
find whether the spectrum is completely fragmented
or not. Both metrics will return values between 0
and 1. A 0 indicates no fragmentation in a link
(spectral fragmentation) or a path (spatial fragmen-
tation) which shows the initial condition of a link
or a path. Similarly, a 1 indicates that a link or path
is completely fragmented which will highly reduce
network performance.
Spectral and spatial fragmentation management
schemes in a network include reactive fragmenta-
tion as well as proactive fragmentation [25], [26].
These are also called defragmentation approaches.
Reactive fragmentation schemes are used after light-
paths are established using hitless methods. Hop re-
turning, push-pull returning, and make-before-break
schemes are all hitless methods which are briefly
explained in [26]. Proactive defragmentation meth-
ods are utilized during lightpath establishment time
for dynamic traffic by taking into account resources
information on different paths between transpon-
ders. The information is processed before lightpath
admission and establishment on a path subject to
different constraints including EFM, ABPM, and
contiguity ratio.
Optimization problems are generally formulated
in literature for static lightpath establishment in
EON. However, traffic nature has highly changed
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
in the recent past with higher traffic growth driven
by dynamic cloud, video, and IoT applications.
Optimization models are formulated with an objec-
tive to establish maximum static lightpaths between
transponders which are constrained by the network
resources. There are few optimization models [12],
[13], [27], [28] which are formulated with static
traffic and their results are utilized for routing
purpose for dynamic traffic. However, the routing
information changes in these models with changes
in network traffic load (Erlang) as the lightpath
establishments are subject to traffic loads (Erlang).
For higher traffic loads (Erlang), resources must be
increased to make the models feasible to support
higher traffic [29]. However, increasing network
capacity with associated hardware and software
deployment is a costly and time consuming pro-
cess. Furthermore, it will require sufficient time
to deploy new capacity and associated resources.
Therefore, the incremental approach for scaling net-
works with deploying more resources with pace of
increasingly traffic demand will lead to unnecessary
infrastructure and cost. It is a network challenge
to optimize the available network resources before
the next network upgrade subject to the elastic
characteristics of the network, technology used in
EON, optimizing modulation schemes for adapting
higher data-rates based on transponders locations,
and the distance/hops between the transponders. We
therefore, proposed an iterative based optimization
model which is further explained in Section II.
In this work,
1) we have considered network partitioning
problem to optimize the network performance.
We have design an iterative based optimiza-
tion problem for network partitioning and
routing. According to our latest information,
there is no work available in the literature
related to the application of iterative optimiza-
tion in EON.
2) Furthermore, proactive defragmentation ap-
proach is used in this work such that routes
between transponders are precomputed with
adapted modulation scheme in the optimiza-
tion model. Therefore, different bandwidths
are allocated to end pairs which are based
on the adapted modulation scheme and their
optical reach.
3) The outcome of the proposed optimization
model is to minimize link loading by adapt-
 This article has been accepted for publication in IEEE Photonics Journal. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/JPHOT.2022.3205412

ing higher modulation scheme with higher nodes.

data-rate transmission and allocating mini- S: Set of all possible source-destination (s-d)
mum numbers of FS’s between transponders. pairs in a network with a number equal to
The outcome of the proposed optimization N (N − 1).
model includes the routing information and cl : The cost c of using a FS channel on a
the amount of bandwidth allocation between link l. In the hop based routing, it has a value
transponders. setted equal to 1. In the distance based routing,
4) The proposed algorithm processes the data- this represents the actual distance of a fiber
sets from the optimization model and categor- link l.
ically allocates spatial contiguous FS blocks.
This also specifies the indexes of the contigu- • Path parameters:
ous band of FS’s according to the allocated length of SPFs : Represents the distance of an
bandwidths along the optimized routes. s-d pair s which are pre-computed using k = 1
5) We utilize the set of integrated output infor- shortest path routing (SPF) with Dijkstra’s
mation from the optimization model and the algorithm.
algorithm for dynamic traffic and networking P: Set of all possible paths which are
to improve efficiency of EON in terms of precomputed using k = 3 SPF routing with
average bandwidth blocking probability, av- Dijkstra’s algorithm.
erage EFM, average ABPM, and link and FS P s : Set of paths ∀s ∈ S. However, we consider
utilization. k = 3 shortest paths which are arranged in an
The rest of the paper is organized as follows. Section ascending order for each s-d pair s in the set.
II include the proposed iterative based optimization Pl : Set of paths which utilizes link l ∀l ∈ L.
model along with all traffic, network, and decision k s : Set of modulation indexes for each s ∈ S
variables. It also include the proposed algorithm such that k s = 1 for BPSK modulation,
which arranges the optimization results into a set k s = 2 for QPSK, k s = 3 for 8QAM, and
of paths and contiguous band of FS’s to be utilized k s = 4 for 16QAM. k s is precomputed using
for dynamic networking. In Section II, we have also length of SPFs . Number of bits required
s
discussed the performance parameters. In Section for each scheme is 2k . Therefore, BPSK
III, we have carried out extensive simulations and use 2 bits, QPSK use 4 bits, and 8QAM
presented the findings of this work. Finally, section use 8 bits to carry information between end
IV concludes this work. pairs. However, QPSK requires twice power
compared to BPSK and 8QAM requires twice
II. O PTIMIZATION MODELS power compared to QPSK to achieve similar
bit-error probability. If transmission power
A. Stage 1: Preparing parameters for the optimiza- remains same, then the optical reach of QPSK
tion model should be reduced to half compared to BPSK
The network, traffic, and path parameters as well and the optical reach of 8QAM should be
as different variables are given below. reduced to half compared to QPSK to achieve
• Network parameters: similar bit-error probabilities. In this work,
N : Represents a set of wavelength selective we consider similar transmission power for all
switch (WSS) nodes geographically located in lightpaths.
a network.
C: Set of frequency slots (FS) per fiber such • Variables:
that C = (1, 2, 3, · · · , C). An FS represents a ∆f s : Numbers of FS allocation to an s-d pair
frequency channel having bandwidth equal to s.
δchannel = 12.5 GHz. |C| represents the size of Λl : Total number of FS allocated on link l.
the set C in which C is the maximum index
of a FS. • Decision variables:
L: Set of directed fiber links which carry αp : Represents bandwidth allocation in num-
traffic in one direction between neighbor bers of FS’s to an s-d pair s

This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
 This article has been accepted for publication in IEEE Photonics Journal. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/JPHOT.2022.3205412

∆fi : An iterative integer value with a lower adapting higher modulation scheme with more bits
bound equal to 1 and an upper bound equal to carry information and to allocate minimum FS’s
to |C|. The proposed optimization problem it- to each s-d pair s. This is explained as follows.
eratively starts with the lower bound. It stops Equation (2) gives the link cost in terms of
when the proposed integer linear problem (ILP) bandwidth allocation in numbers of FS’s. These
becomes feasible and gives an optimal solution. FS’s are allocated to an s-d pair s along a path p.
We have observed that its value, which makes Therefore, the link cost includes the sum of FS’s
the problem feasible, remains unchanged for allocation on all paths which is utilizing a link l.
the same network. Noting its value will reduce Equation (3) shows that the sum of FS allocation
the numbers of iterations to only one time on any link l must not exceed the available numbers
execution for the same network. of FS’s on that link. Moreover, it is important to note
that path based FS allocation is considered in Eq.
B. Stage 2: Programming the optimization model (3) so that same band of FS’s will be allocated to
each s-d pair s on all links along the path which
Objective function: satisfies the continuity.
Equation (4) represents numbers FS allocation to
X X
s
minimize cl Λ l + ∆f (1)
∀l∈L ∀s∈S
an s-d pair s. For a demand request equal to 100G,
BPSK with modulation index k s = 1 will require
Subject to: only 2 bits and 4 FS with δchannel = 12.5 GHz
for transmission. Similarly, QPSK with k s = 2 will
X
p
Λl = α , ∀l ∈ L (2)
∀p∈Pl
require 4 bits and 2 FS to carry the information
X between end nodes. Finally, 8QAM with k s = 3
αp ≤ |C|, ∀l ∈ L (3) will require 8 bits and only 1 FS to carry the
∀p∈Pl information between end nodes. However, in case of
giving preference to the lightpath distance instead
s
of number of hops in the shortest path calcu-

 4, if 2k = 2bits
∆f = s
2, if 2 = 4bits , ∀s ∈ S (4) lation/consideration/computation, the constraint in
ks
 s
1, if 2k = 8bits Eq. (5) will be considered instead of the constraint
in Eq. (4) which will give similar information of
 allocating FS’s for each modulation scheme. In Eq.
 1, if length of SPFs ≤ 1000km
(5), if length of SPFs <= 1000km, then 8QAM
∆f s = 2, if length of SPFs ≤ 2000km , ∀s ∈ S
 4, if length of SPFs ≤ 4000km with 8 bits and 1 FS will be adapted by an s-d pair s.
Similarly, if 1000km < length of SPFs ≤ 2000km,
(5)
then QPSK with 4 bits and requiring 2 FS’s will be
adapted by an s-d pair s. Finally, for a transmission
X |C|
p
α =d e∆f , ∀s ∈ S (6) distance between 2000 km and 4000 km, BPSK with
s

∀p∈P s
∆f i δchannel 2 bits and 8 FS’s will be adapted by an s-d pair s
to satisfy the bandwidth demand.
Finally, the integer constraints are given in Eq. The constraint in Eq. (6) is explained as follows.
(7)- Eq. (9). We have considered s-d pairs with uniform demand
αp ∈ Z+ , ∀p ∈ P (7) requests equal to 100G. Therefore, BPSK with 2
bits will require more FS compared to QPSK with
∆f s ∈ Z+ , ∀s ∈ S (8) 4 bits and 8QAM with 8 bits. In other words, for
Λl ∈ Z+ , ∀l ∈ L (9) 100 Gbps bandwidth demand, BPSK will require
4 FS, i.e., 4 FS × 12.5GHz (channel width) × 2
The objective function of the optimization prob- bits = 100 Gbps. Similarly, QPSK will require 2
lem in Eq. (1) is to minimize the cost of using FS’s FS, i.e., 2 FS × 12.5 GHz (channel width) × 4
on all links in a network as well as to optimize bits = 100 Gbps. Finally, 8QAM will require 1 FS,
FS allocation to each s-d pair s. This will help i.e., 1 FS × 12.5 GHz (channel width) × 8 bits
in optimizing a route between end WSS nodes by = 100 Gbps. In all cases, 12.5 GHz represents the

This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
 This article has been accepted for publication in IEEE Photonics Journal. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/JPHOT.2022.3205412

channel width of a FS. Similarly, BPSK has highest the optimization problem will be used for dynamic
optical reach compared to other schemes. Therefore, network operation.
100G bandwidth demand requires 1 FS using 100G
8QAM (i.e., 8 bits) transmission with lowest optical Input: link-optimal-values;
reach. A similar bandwidth demand requires 2 FS optimal-set-of-paths
using 50G QPSK (i.e., 4 bits) transmission with Output: path-slot-info
an optical reach up to 2000 km. Finally, 100G /* [ ] → list, {} → set */
bandwidth demand requires 4 FS using 25G BPSK sd-path-used = [ ]
(i.e., 2 bits) transmission with an optical reach up link-path-cluster-size = link-optimal-values
to 4000 km. sd-path-values = [((sd,path),value(αp )),
∀s ∈ S, ∀path ∈ optimal-set-of-paths[s]]
C. Stage 3: Programing an algorithm to extract/sort paths-through-links = {c : [ ∀l ∈ L | link-
useful information optimal-values[l] = c ], ∀c ∈
In this Section, we have proposed an algorithm link-path-cluster-size }
which takes into account the information from the links-from-opt-problem = {l : [path transiting
optimization model and allocates contiguous blocks link l; ∀ path ∈ optimal-set-of-paths | l ∈
of FS’s with continuity constraint. The procedure to path ] ∀ l ∈ L }
extract the necessary information and allocation of for link-size in link-path-cluster-size do
FS bands is given in Algorithm (1). In Algorithm for link in paths-through-links[link-size]
(1), sd-path-used is a list which will keep track do
of sd-paths such that resources will be allocated dbr-path-sort =
only once to the path in the list. link-optimal-values links-from-opt-problem[link]
represents numbers of FS slots allocation to optical for path in dbr-path-sort do
links from the optimization problem. This list will if path not in sd-path-used then
be sorted in descending order without repetition. sd-path-used.append(path)
The sorted list will be used to group links according Find continuous-FS-slots
to FS numbers. The set sd-path-values keeps the list Find sd =
of all tuples of sd pair ∀s ∈ S and a path for sd pair. (path[0][0],path[len(path)-
p
The value of α is obtained from the optimization 1][1])
problem which shows the number of FS’s allocated
on route p according to the modulation used. Find contiguous-FS-slots in
In the set paths-through-links, paths are grouped continuous-FS-slots with
according to the link-path-cluster-size. The links- values
from-opt-problem represents a set of lists of tuples ≥ sd-path-values[(sd,path)]
including link l ∀ l ∈ L and paths from the // Numbers of
optimization problem which are utilizing link l. The contiguous-FS-slots
dbr-path-sort keeps list of paths utilizing link l. will be equal to
The paths in the list are sorted in descending order numbers of FS
according to the number of hops. This will ensure slots allocated
allocating FS slots first to the longest routes to min- from the
imize contiguity and continuity issues of FS slots optimization
allocation due to horizontal or spatial fragmentation. problem.
Shortest routes with higher modulation formats, end
better data rates, and minimum FS requirements will path-slots-info =
face minimum horizontal fragmentation. Therefore, {(path, contiguous-FS-slots)}
FS contiguity requirement will be addressed more end
efficiently. Finally, path-slots-info is a set of tuples end
with path information and list of contiguous FS slots end
allocated along the path. The resources in the path- Algorithm 1: Algorithm
slots-info as well as the routing information from

This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
 This article has been accepted for publication in IEEE Photonics Journal. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/JPHOT.2022.3205412

D. Stage 4: Performance parameters

In this Section, we have discussed several per-
formance measurement metrics for evaluating the
performance of EON for dynamic traffic in this
work. These metrics include average network exter-
nal fragmentation metric (EFM), average network
ABPM, average network contiguity, average net-
work bandwidth blocking probability (BBP), av-
erage standard deviation of FS utilization for the
EON, and network link utilization. These are ex-
plained as follows.
1) Average network EFM: This parameters gives
the information about maximum available spatial
FS block size along a route [30]. It is given in
Figure 1. Spectral fragmentation in links L1,2 ,L2,3 , and L3,4
Eq. (10) in which xi and yi represent available block
of maximum contiguous spatial FS’s and available
number of all spatial FS’s along a route for arrival
i. ABPM as follows.
 
xi
b 22 c + b 42 c
P
∀ i∈arrivals 1 − yi path1,3
EFM = (10) ABPM =1−
total arrivals b 12
2
c
2) Average network ABPM: Average network 1,3 3
ABPM can be used to estimate spatial fragmentation ABPMpath = 1 − = 0.5
6
for dynamic traffic in a network. It is given by a
mathematical expression in Eq. (11) [12], [30]. The Similarly,
value of average network ABPM ranges from 0 to 1 b2c + b4c
1,4
and shows spatial fragmentation. A 0 indicates the ABPMpath = 1 − 4 12 4
initial condition of FS along a route between the b4c
transponders. Similarly, a 1 indicates that the FS’s
1,4 0+1
along a route between the transponders is mostly ABPMpath = 1 − = 0.667
fragmented and not available for the demanded band 3
of FS’s. This means that the ABPM values on the spatial
fragmented optical paths path1,3 and path1,4 are

P block sizei

∀available FS blocks b demanded slots c 
50% and 66.7% for a demand of 2 FS and 4
i
P
1 −
FS using QPSK and BPSK modulation schemes

∀ i∈arrivals  0
FS s on pathi

b c
demanded slotsi
ABPM =
total arrivals
(11)
respectively. In the second case, we split the spectral
bandwidth into two categories which is shown in
Consider three links L1,2 ,L2,3 , and L3,4 between
Figure 2 with different colors. FS1 to FS4 are
nodes n1 , n2 , n3 , and n4 . The spectral fragmen-
allocated on L1,2 and L2,3 to carry the information
tation in each link is given in Figure 1. Let we
over the path path1,3 using QPSK. Similarly, FS5
consider two incremental lightpath requests s1 and
to FS12 are allocated on L1,2 , L2,3 , and L3,4 to carry
s2 between n1 and n3 and between n1 and n4 respec-
the information over the path path1,4 using BPSK
tively. Let s1 use QPSK modulation and requires 2
scheme. In this case, we calculated the values of
FS’s. Similarly, let s2 use BPSK modulation and
ABPM as follows.
requires 4 FS’s to setup lightpath betweeen the end
nodes. The spatial fragmentation over the path for path1,3 b 22 c
1,3
s1 is represented by path . Similarly, the spatial ABPM =1− 4
2
b2c
fragmentation over the path for s is represented
by path1,4 . In the first case, we consider the whole 1,3 1
bandwidth of fiber links to calculate the values of ABPMpath = 1 − = 0.5
2

This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
 This article has been accepted for publication in IEEE Photonics Journal. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/JPHOT.2022.3205412

indicates that a FS is busy. The value of spatial

route contiguity ranges from 0 to 1. A value 1
indicates that all available FS’s are contiguous and
close neighbors with each other. Alternatively, a
value 0 indicates that the available FS’s on a spatial
route (if any) are not contiguous and close neighbors
with each other.
4) Average network BBP: The statement of av-
erage network BBP is adapted from the work in
[15]. It is given in Eq. (13) as a ratio between the
bandwidth of lightpath requests which could not be
established successfully to the total bandwidth of
lightpath requests. The later includes the bandwidths
of successfully established and blocked lightpath
Figure 2. Spectrum is divided into two categories with contiguous
requests.
FS’s on each link for s-d pairs (1, 3) and (1, 4) using QPSK and
i
P
BPSK modulation respectively. i∈ [Lightpaths blocked] ∆f
BBP = P i
(13)
i∈[All lightpath requests] ∆f

Similarly,
1,4 b 44 c
ABPMpath =1− 8
b4c
1,41
ABPMpath =1−
= 0.5
2 Λl
This means that the ABPM values on the spatial
fragmented optical paths path1,3 and path1,4 are
50% each for a demand of 2 FS and 4 FS using
QPSK and BPSK modulation schemes respectively.
This means that splitting the bandwidth of the
fiber links into different categories improve ABPM
values. This is due to the fact that FS’s are allocated
and released in pairs in the same category. Hence,
the probability of availability of contiguous FS’s
is good compared to allocating different sizes of
FS’s in a shared category. Therefore, resources in
a shared category makes the spatial band of FS’s
more fragmented.
3) Average network contiguity: The mathemat-
ical formulation for finding the average network
contiguity (ANC) is adapted from the work in
[31] which is given in Eq. (12). SFFSp means
spatial fragmented FS along a route p between the
transponders during lightpath establishment.
P PC p p
i=2 (SFFS [i]×SFFS [i−1])
PC
j∈ arrivals
i=1
ANC =
P
j∈arrivals
SFFSp has two states which is either 1 or 0. A
1 indicates the availability of a free FS and a 0
5) Average standard deviation of FS utilization:
An average value of standard deviation of FS uti-
lization δ(FSU) in a network is given in Eq. (14).

maxF SIl

δ(FSU) = δ ∀l∈L (14)
In Eq. (14), FSI denotes the maximum index of
a FS which is utilized on a link l. Similarly, Λl
represents the numbers of all FS’s on a link l.
Finally, δ(−) represents the standard deviation of
a list of values which has a value in the range from
0 to 1. A value δ(−) = 0 indicates that all values
of FS on links in a network are same. Similarly,
a value δ(−) = 1 indicates that FS utilization on
some links is extremely low while on other links,
FS utilization is extremely high. Finally, the value
δ(−) = 0.5 indicates that some links in a network
are frequently utilized compared to others.
6) Network link utilization: The value of network
link utilization (LU) in a network for link l is given
in Eq. (15). In Eq. (15), (l, maxF
Λl
SIl
) is a tuple with
link l and its utilized value. LU is a ratio of the
maximum index of a FS which is utilized on a link
l to the maximum value of FS’s on that link. The
values of LU are in the range 0 to 1. A value =
1 indicates maximum utilization and a value = 0
SFFSp [i]−1
(12)
indicates minimum utilization of a link l.
j
maxF SIl
LU = ∀l∈L {(l, )} (15)
Λl

This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
 This article has been accepted for publication in IEEE Photonics Journal. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/JPHOT.2022.3205412

III. R ESULTS AND DISCUSSION

In this section, we have analyzed the performance

of the proposed integrated model for dynamic net-
working. The integrated model includes the results
from the optimization model which is followed
by the proposed algorithm to allocate resources
between transponders in groups. The integrated
approach will give routing information, spectral
capacity allocation on each link in terms of FS’s,
and spatial capacity allocation between transponders
along the route with their modulation schemes. In
simulations, we have used first-fit spectrum allo-
cation scheme for FS allocation with the routing
information from the optimization model. We have
also considered lightpaths with variable data rates
and bandwidth requirement arriving to a network
with Poisson’s distribution. However, traffic is con-
sidered to be uniformly distributed. The mean hold-
ing time of lightpaths to stay in the network follows
exponential distribution with the mean holding time
equal to 1 time unit. In general, we have simulated Figure 3. COST 239 network topology (net1)
different scenarios in two different networks for 1
million lightpath requests to get steady state results.
We have used two networks (net1 and net2) to
evaluate the performance of the proposed approach.
Networks net1 and net2 have 11 and 14 nodes
each which are shown in Figure 3 and Figure 4
respectively. Moreover, net1 has 50 directed links 1
while net2 have 64 directed links.
4
2 3
In the first stage, we get all related information
sets from the optimization problem using GLPK
solver with Python Linear Program Module (PuLP) 5
7 8
[32]. Let us consider network net1 with all input
parameters for the optimization model which are 6
given in section II. After solving the optimization 9
model, we extracted the information related to the 10
link cost for the link l = (1, 9) in net1, paths 11
utilizing the link l, and capacity allocation on routes
utilizing link l. These values of different parameters
are given in Table I which are extracted using the 12
programming syntax in algorithm 2, algorithm 3,
14
and algorithm 4. Similarly, we use the proposed
algorithm to allocate block of contiguous spatial 13
FS’s along the optimized routes and using all other
information from the optimization problem. The Figure 4. Radnomly generated network topology with 14 switching
nodes (net2)
contiguous FS allocation on different routes in Table
I are given in Table II using the proposed algorithm
1.

This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
 This article has been accepted for publication in IEEE Photonics Journal. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/JPHOT.2022.3205412

TABLE I. TABLE II.

E XTRACTING USEFUL INFORMATION FROM THE OPTIMIZATION A LLOCATING CONTIGUOUS FS’ S ALONG DIFFERENT ROUTES IN
MODEL FOR LINK l = (1, 9) TABLE I FROM THE OPTIMIZATION MODEL PASSING THROUGH

Link Link Capacity LINK l = (1, 9) USING THE PROPOSED ALGORITHM .

(1,9) 160 FS’s s-d Modulation
path FS allocation on Path
Link path through link pair
(1,9) ((1, 9), (9, 8)) (1,11) BPSK ((1, 9), (9, [FSindex ∀ index ∈ [1, 2, · · · , 60]]
(1,9) ((1, 9),) 8), (8, 11))
(1,9) ((1, 9), (9, 8), (8, 11)) (1,11) BPSK ((1, 3), (3, [FSindex ∀ index ∈ [1, 2, · · · , 20]]
(1,9) ((3, 1), (1, 9)) 6), (6, 11))
(1,9) ((6, 2), (2, 1), (1, 9)) (6,9) BPSK ((6, 2), (2, [FSindex ∀ index ∈ [61, 62, · · · , 80]]
path capacity of path and 1), (1, 9))
adapted modulation (6,9) BPSK ((6, 7), (7, [FSindex ∀ index ∈ [1, 2, · · · , 60]]
((1, 9), (9, 8)) 40 FS’s / QPSK 8), (8, 9))
((1, 9),) 20 FS’s / 8QAM (1,8) QPSK ((1, 9), (9, [FSindex ∀ index ∈ [81, 82, · · · , 120]]
((1, 9), (9, 8), (8, 11)) 60 FS’s / BPSK 8))
((1, 3), (3, 6), (6, 11)) 20 FS’s / BPSK (3,9) QPSK ((3, 1), (1, [FSindex ∀ index ∈ [121, 122, · · · , 140]]
((3, 1), (1, 9)) 20 FS’s / QPSK 9))
((6, 2), (2, 1), (1, 9)) 20 FS’s / BPSK (1,9) 8QAM ((1, 9),) [FSindex ∀ index ∈ [141, 142, · · · , 160]]
((6, 7), (7, 8), (8, 9)) 60 FS’s / BPSK

blocking probabilities (BBP) for different network

Algorithm 2 using Python Programming for ex-
traffic loads (Erlang) in net1 and net2 using the
tracting information of paths through link
proposed scheme. The BBP values are compared
[1]: paths-through-link = {}
with the existing SPF (k=1) routing and the alternate
[2]: ∀l ∈ L:
routing schemes. We have integrated the optimiza-
[3]: temp-link = []
tion information with the first-fit spectrum alloca-
[4]: for sd in sd-pairs:
tion subproblem for FS allocation in network. We
[5]: for path in optimized-paths[sd]:
adapted integration of the same spectrum allocation
[6]: if l ∈ path :
subproblem with the existing routing schemes also.
[7]: temp-link.append(path)
However, in both SPF (k=1) routing and alternate
[8]: paths-through-link[l] = temp-link
(k=2) routing, the complete spectrum is shared by
Algorithm 3 using Python Programming for ex- all s-d pairs in a network. It has been observed that
tracting information of bandwidth (capacity) over the proposed scheme highly reduces the amount of
a path: value(αpath ) is Python syntax using PuLP BBP values compared to the existing SPF routing
to give cost of a variable. and alternate routing. It should be noted that the
[1]: sd-path-value = dict(((s,path),value(αpath )) integrated approach is using approximately equal
∀s ∈ S and ∀ path ∈ optimized-paths[s]) number of routes compared to SPF routing and
half number of routes compared to alternate routing.
Algorithm 4 using Python Programming for ex- Similarly, the total link cost in the proposed scheme
tracting information of link cost values is minimum in both net1 and net2 compared to
[1]: link-cost = dict((l,[ ]) ∀l ∈ L) the existing schemes. The routing and s-d pair
[2]: ∀l ∈ L: informations for both net1 and net2 are compared
[3]: for temppath in paths-through-link[l]: in Table III and Table IV respectively.
[4]: link-cost[l].append(sd-path-value[((temp- In Figure 6, we have calculated the average
path[0][0],temp-path[len(temp-path)-1][1]),temp- ABPM values for different traffic loads (Erlang) in
path)]) both networks. The proposed integrated approach
[5]: link-values = dict((l, sum(link-cost[l])) ∀l ∈ has lower average ABPM values compared to SPF
L) routing and alternate routing. This shows that the
After getting related information from the op- spatial fragmentation decreases in splitting the avail-
timization model and the proposed algorithm, we able spectrum into a number of limited subgroups.
utilize these sets of information for dynamic traffic This increases the availability of idle spatial FS
in different networks. Figure 5 shows the bandwidth along a route. Therefore, reduces the amount of

This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
 This article has been accepted for publication in IEEE Photonics Journal. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/JPHOT.2022.3205412

TABLE III. 0.12

proposed
S - D PAIR AND ROUTING INFORMATION IN NET 1 SPF (k=1) routing (existing)
Algorithm s-d pairs Number of Network link 0.10 SPF (k=2) routing (existing)
routes cost (in num-
bers of FS) 0.08
SPF routing 110 110 8000
Alternate 110 220 8000

ABPM
routing 0.06
proposed 110 123 6440
scheme 0.04

0.02
TABLE IV.
S - D PAIR AND ROUTING INFORMATION IN NET 2
0.00
Algorithm s-d pairs Number of Network link 600 800 1000 1200 1400 1600
routes cost (in num- traffic load (Erlang)
bers of FS) 0.12
SPF routing 182 182 10204 proposed
Alternate 182 364 10204 SPF (k=1) routing (existing)
0.10 SPF (k=2) routing (existing)
routing
proposed 182 204 8368
scheme 0.08
ABPM
0.06
0.12
proposed
SPF (k=1) routing (existing) 0.04
0.10 SPF (k=2) routing (existing)

0.02
Bandwidth blocking

0.08
0.00
600 700 800 900 1000 1100 1200 1300
0.06 traffic load (Erlang)

0.04
Figure 6. ABPM verses traffic load (Erlang) using different routing
schemes in net1 and net2
0.02

0.00
600 800 1000 1200 1400 1600
traffic load (Erlang)
BBP values compared to SPF routing and alternate
routing. In both SPF routing and alternate rout-
0.10
proposed ing, the complete spectral FS on links are shared
SPF (k=1) routing (existing)
SPF (k=2) routing (existing)
among the s-d pairs which sufficiently increases
0.08 the values of ABPM. Therefore, it also increases
spatial fragmentation along the routes. In Figure
Bandwidth blocking

0.06 6, the values of average ABPM increases with the

network traffic load (Erlang) in all cases. Hence,
0.04
spatial fragmentation increases with an increase in
the network traffic load (Erlang).
0.02
In Figure 7, we have calculated the average EFM
values for different traffic loads (Erlang) in both net-
works. We have observed that the external fragmen-
0.00
600 700 800 900 1000 1100 1200 1300 tation in the proposed scheme is also less compared
traffic load (Erlang) to the existing schemes. It can be observed that the
EFM values in alternate routing is more compared
Figure 5. Bandwidth blocking probability verses traffic load (Erlang) to SPF routing and the proposed scheme. This is
using different routing schemes in net1 and net2
due to the fact that traffic will be diverted to the

This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
 This article has been accepted for publication in IEEE Photonics Journal. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/JPHOT.2022.3205412

0.8 1.0
proposed proposed
SPF (k=1) routing (existing) SPF (k=1) routing (existing)
0.7 SPF (k=2) routing (existing) SPF (k=2) routing (existing)
0.9
External Fragmentation

0.6

Average Contiguity
0.8
0.5
0.7
0.4

0.6
0.3

0.2
0.5

0.1 0.4
600 800 1000 1200 1400 1600 600 800 1000 1200 1400 1600
traffic load (Erlang) traffic load (Erlang)
0.8 1.0
proposed proposed
SPF (k=1) routing (existing) SPF (k=1) routing (existing)
0.7 SPF (k=2) routing (existing) SPF (k=2) routing (existing)
0.9
External Fragmentation

0.6

Average Contiguity
0.8
0.5
0.7
0.4

0.6
0.3

0.2
0.5

0.1 0.4
600 700 800 900 1000 1100 1200 1300 600 700 800 900 1000 1100 1200 1300
traffic load (Erlang) traffic load (Erlang)

Figure 7. External fragmentation verses traffic load (Erlang) using Figure 8. Average contiguity verses traffic load (Erlang) using
different routing schemes in net1 and net2 different routing schemes in net1 and net2

underutilized links on the second route between the it is lowly utilized. This trend continues for SPF
transponders to use the idle resources. Therefore, it (k=1) routing for all traffic load (Erlang) values as
will reduce the availability of a block of maximum BBP will increase sufficiently in the highly utilized
contiguous spatial FS’s to utilize idle resources on links for higher traffic. This will make a bottleneck
these routes. for other links to carry higher traffic. Hence, these
In Figure 8, the average contiguity in the pro- under-utilized links will not carry traffic to increase
posed scheme is less compared to other schemes. FSU. The proposed scheme has the lowest values
This is due to the fact that the complete spectrum of δ(FSU) for all traffic values compared to SPF
is divided into a number of limited subgroups. (k=1) routing and alternate (k=2) routing. This trend
Therefore, the availability of a maximum block shows that FSU on all links is approximately equal
of spatial FS’s along the network routes is lower as the standard deviation of FSU in both networks
compared to spectral bandwidth without partitions. for all traffic values remains within the range 5%.
In Figure 9, the values of δ(FSU) are shown for This further shows that link utilization in both net-
different traffic load (Erlang) values in net1 and works are balanced. For alternate (k=2) routing with
net2. It has been observed that for lower values lower traffic, the value of δ(FSU) is approximately
of network traffic load (Erlang), δ(FSU) is higher equal to the value of δ(FSU) for SPF (k=1) routing.
in SPF (k=1) routing which shows that FS’s on This is due to the fact that alternate routing may
some links are highly utilized and on other links, use only one route due to lower traffic. However,

This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
 This article has been accepted for publication in IEEE Photonics Journal. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/JPHOT.2022.3205412

0.30 compared to the others in a shared spectrum. Sim-

proposed
SPF (k=1) routing (existing)
ilarly, the FS with maximum indexes are utilized
0.25 SPF (k=2) routing (existing)
lowly. The high and low utilization of few bins in
SPF routing and alternate routing are reduced in the
STD of FS Utilization

0.20
proposed scheme in which the complete spectrum is
divided into limited numbers of spatial bandwidths.
0.15
These spatial categories of FS blocks are dedicated
on each path for each s-d pair. Therefore, first-fit
0.10
scheme will utilize the starting minimum index of
0.05
each subgroup. It is also increasing the frequency
of utilization of FS with higher indexes.
600 800 1000 1200 1400 1600
In Figure 12 and Figure 13, network links with
traffic load (Erlang) similar utilization are grouped together into bins.
0.30 It has been observed that the proposed scheme has
the highest link utilization. The lower bound of link
0.25 utilization in the proposed scheme is greater than
65% and 80% for net1 and net2 respectively. The
STD of FS Utilization

0.20 SPF routing has the lowest link utilization with few
proposed links under utilized. Alternate routing increases link
0.15 SPF (k=1) routing (existing) utilization compared to SPF routing. However, the
SPF (k=2) routing (existing)
highly utilized links in this case also becomes a
0.10 bottleneck for carrying additional traffic on other
paths.
0.05 Finally, we have plotted the simulation time in
Figure 14 in both networks for different traffic
600 700 800 900 1000 1100 1200 1300 values in all cases. It has been found that the
traffic load (Erlang)
proposed scheme has an improved lightpath setup
time compared to other schemes. This is due to
Figure 9. Standard deviation of FS utilization verses traffic load
(Erlang) using different routing schemes in net1 and net2
the dedicated bounded spatial blocks with fewer FS
compared to searching for idle FS in a completely
shared set of resources. It is expected that the pro-
posed scheme will also improve bandwidth blocking
as the network traffic increases, alternate routing due to delayed information sharing and signaling
utilizes the available resources on second route in between the network nodes.
case of unavailability of idle resources on primary
route. Therefore, δ(FSU) decreases slightly and
approaches to the values obtained in the proposed IV. C ONCLUSION
scheme. Hence, δ(FSU) decreases in alternate rout- In this work, we have proposed an integrated
ing for higher traffic loads (Erlang) due to increase scheme for dynamic traffic which works in three
in (FSU) over all links in the network. phases. In the first phase, an optimization formu-
In Figure 10 and Figure 11, FS with same lation is programmed which is solved using an
utilization in net1 and net2 are grouped together iterative procedure to minimize the network cost.
in different subgroups. These subgroups are called The optimization model gives routing information
bins. It has been observed in both SPF routing and according to the adapted modulation schemes. The
alternate routing that few groups are utilized very adapted modulation schemes between transponders
often compared to others. Similarly, few bins are are optimized according to optical reaches. The
utilized least frequently. This is possible as we have objective of the optimization problem is to reduce
integrated SPF routing and alternate routing with the network cost by adapting to higher modula-
first-fit spectrum allocation scheme. Therefore, the tion schemes with higher data-rates and minimum
FS with minimum indexes will be highly utilized bandwidth requirements between transponders. The

This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
 This article has been accepted for publication in IEEE Photonics Journal. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/JPHOT.2022.3205412

Bins (with listed sizes) of FS with different index

Bins (with listed sizes) of FS with different index

proposed 20.0 proposed
12
17.5

10 15.0

8 12.5

10.0
6
7.5
4
5.0
2
2.5

0 0.0
4000 6000 8000 10000 12000 14000 16000 18000 6000 8000 10000 12000 14000 16000 18000
Number of times FS utilized Number of times FS utilized
Bins (with listed sizes) of FS with different index

17.5
SPF (k=1) routing (existing)
Bins (with listed sizes) of FS with different index SPF (k=1) routing (existing)

25
15.0

12.5 20

10.0 15

7.5
10
5.0
5
2.5

0.0 0
0 5000 10000 15000 20000 25000 0 10000 20000 30000 40000
Number of times FS utilized Number of times FS utilized
Bins (with listed sizes) of FS with different index

Bins (with listed sizes) of FS with different index

SPF (k=2) routing (existing) 25 SPF (k=2) routing (existing)

16

14 20
12

10 15

8
10
6

4
5
2

0 0
5000 10000 15000 20000 25000 0 10000 20000 30000 40000
Number of times FS utilized Number of times FS utilized

Figure 10. Bins (of similar sizes) of FS with different indexes verses Figure 11. Bins (of similar sizes) of FS with different indexes verses
number of times FS is utilized using different routing schemes in net1 number of times FS is utilized using different routing schemes in net2

This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
 This article has been accepted for publication in IEEE Photonics Journal. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/JPHOT.2022.3205412

25
proposed proposed

Bins (with listed sizes) of different links

Bins (with listed sizes) of different links

20 40

15 30

10 20

5 10

0 0
0.65 0.70 0.75 0.80 0.85 0.90 0.95 1.00 0.800 0.825 0.850 0.875 0.900 0.925 0.950 0.975 1.000
link utilization link utilization

SPF (k=1) routing (existing) 25 SPF (k=1) routing (existing)

17.5
Bins (with listed sizes) of different links

Bins (with listed sizes) of different links

15.0 20

12.5
15
10.0

7.5 10

5.0
5
2.5

0.0 0
0.2 0.4 0.6 0.8 1.0 0.2 0.4 0.6 0.8 1.0
link utilization link utilization
35
SPF (k=2) routing (existing) SPF (k=2) routing (existing)
17.5
Bins (with listed sizes) of different links

Bins (with listed sizes) of different links

30
15.0
25
12.5

20
10.0

7.5 15

5.0 10

2.5 5

0.0 0
0.2 0.4 0.6 0.8 1.0 0.2 0.4 0.6 0.8 1.0
Link utilization Link utilization

Figure 12. Bins (with similar utilization) of different links verses Figure 13. Bins (with similar utilization) of different links verses
link utilization using different routing schemes in net1 link utilization using different routing schemes in net2

This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
 This article has been accepted for publication in IEEE Photonics Journal. This is the author's version which has not been fully edited and
300
250
Simulation Time (in seconds)
200
150
100
50
600 800
300
Simulation Time (in seconds)
250
200
150
100
50
600 700 800
Traffic load (Erlang)
Figure 14. Simulation time (in seconds) verses traffic load (Erlang)
using different routing schemes in net1 and net2
second phase utilizes the routing and spectral band-
width allocation such that spatial blocks of FS
are allocated between transponders with continuity
and contiguity constraints. These spatial FS blocks
along with the routing information are utilized for
dynamic traffic in the third phase. It has been
shown with the support of simulation results that
the spatial partitions improves the network perfor-
mance in EON. Average ABPM and EPM decreases
in the integrated proposed scheme compared to k
SPF routing in different networks which shows the
availability of FS blocks to support further traffic.
Therefore, the FS utilization and the link utilization
in different networks increase to support dynamic
traffic in the proposed scheme. The standard devi-
ation of FS utilization in different networks is ap-
proximately 5% compared to the existing schemes.
FS utilization is approximately 25% to 30% in SPF
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
content may change prior to final publication. Citation information: DOI 10.1109/JPHOT.2022.3205412
Proposed (k = 1) routing. For alternate routing, the standard
SPF routing
alternate routing deviation of FS utilization is 25% to 30% initially
for lower network traffic. However, as the network
traffic load (Erlang) increases, the FS utilization
approaches to the level of the proposed scheme
which shows that the alternate routing diverts traffic
to the under-utilized links to improve link and FS
utilization.
In practice, online routing is more time consum-
ing in searching a feasible or optimal route between
pair nodes with bounded constraints. Moreover, the
time utilization in practical networks will increase
1000 1200 1400 1600
Traffic load (Erlang) with taking more number of routes. There are
studies available on RSA/RMSA in dynamic EON
Proposed
SPF routing which allocates routes between end pairs during the
alternate routing lightpath setup [33]. However, in network operation,
the existing dynamic models will take more setup
time to search a route with given constraints during
lightpath establishment compared to the proposed
approach in this work. A comparison of simulation
time between the existing SPF routing, alternate
routing, and the proposed scheme in this work
is shown in Figure 14. The routing tables are
computed offline in all cases which are utilized
for lightpath establishment in network operation
900 1000 1100 1200 1300 stage. This means that the setup time will further
increased to search for an optimal route in online
routing compared to the reported schemes during
a lightpath establishment. Therefore, the existing
dynamic models are available and have been studied
in the literature. However, there may be practical
complications in implementing them for the dy-
namic network operation.
Furthermore, it is a known fact that the blocking
will also occur due to signaling protocols / outdated
information in online routing by utilizing shared
pool of network resources. Hence, the blocking does
not happens only due to unavailability of resources
to establish a lightpath. It also occurs due to out-
dated signaling between end pairs while utilizing
shared network resources. The matter is out of scope
of this work and we have not studied it. However,
the proposed scheme with dedicated resources allo-
cation between end pairs will further eliminate the
blocking probability with outdated signaling issue.
Unlike the proposed scheme, existing online routing
schemes will consider additional signaling proto-
cols in network operation at the cost of additional
connection establishment time. Alternatively, it may
not consider the signaling protocols in network
 This article has been accepted for publication in IEEE Photonics Journal. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/JPHOT.2022.3205412

0.10
operation at the cost of additional network blocking proposed (modified)
probability due to outdated signaling information SPF (k=1) routing (existing)
SPF (k=2) routing (existing)
between end pairs in a network. This shows the 0.08
novelty and uniqueness of our proposed work.

Bandwidth blocking
0.06
A PPENDIX
It is to note that a network may be operated under
0.04
2% to 5% blocking threshold level. In our case,
our proposed scheme maintains the threshold level
0.02
compared to the SPF blocking. However, unlike the
first case (COST239 network – net1), the proposed
scheme does not reduce the amount of bandwidth 0.00
600 700 800 900 1000 1100 1200 1300 1400
blocking probability in net2 in Figure 5 further to traffic load (Erlang)
achieve lower values compared to alternate routing
scheme. However, it achieves other gains in terms of Figure 15. Bandwidth blocking probability comparison of different
network cost, link blocking probabilities, fragmen- routing schemes in net2 with symmetric link capacities and equal
network cost in all cases
tation gain, and execution time compared to the ex-
isting SPF routing and alternate routing schemes. To
demonstrate this statement, we consider equal link Proposed
Proposed (modified)
costs to evaluate the performance of all algorithms. 300 SPF routing
alternate routing
This means that the number of FS’s on each link
Simulation Time (in seconds)

is symmetric and fixed to 160 FS’s in each case so 250

that the network cost becomes equal in all three
scenarios. Further more, the proposed algorithm 200
is modified such that it first uses the integrated
approach as explained in Section III. However, if 150
a lightpath fails to establish using the integrated
approach, it further utilizes the routing information 100

with first-fit heuristic to utilize the unused resources.

This means that the resources outside the established
tunnels are also utilized between end pairs. The
scenario is simulated with 1 Million connection
requests. This reduces the amount of bandwidth
blocking probability at the price of network cost
as shown in Figure 15. Further, we also plotted
the simulation time in Figure 16 which is slightly
increased compared to the proposed algorithm in
Section III. However, the simulation time in the pro-
posed algorithm remains largely below the existing
SPF routing and the alternate routing algorithms.
R EFERENCES
[1] https://www.cisco.com/c/dam/m/digital/elq-
cmcglobal/witb/CKN/0421 SP NETWORKING.pdf
Accessed: May.16.
[2] Hassan Yousif Ahmed, Medien Zeghid, Waqas A. Im-
tiaz, Teena Sharma, Akhtar Nawaz Khan, Kottakkaran
Sooppy Nisar, Adaptive Transceiver Architecture with
QoS provision for OCDMA network based on logic
gates, IEEE ACCESS, vol. 9, No. 21417120, pp.
151089 – 151109, 2021.
50
600 700 800 900 1000 1100 1200 1300 1400
Traffic load (Erlang)
Figure 16. Comparison of simulation time (in seconds) of different
routing schemes in net2 with equal link capacity in all cases
[3] Hassan Yousif Ahmed, Medien Zeghid, Teena Sharma,
Akhtar Nawaz Khan, Waqas A Imtiaz, Abd El-
Mottaleb, A Somia, An efficient algorithm to provide
triple play services in passive optical network (PON)-
OCDMA network, Optical and Quantum Electronics,
vol. 54, No. 6, pp. 1-21, 2022.
[4] Medien Zeghid, Kamal Amaseb, Hassan Yousif Ahmed,
Akhtar Nawaz Khan, Abdellah Chehri, Teena Sharma,
Kottakkaran Sooppy Nisar, Modified Optical Burst
Switching (OBS) Based Edge Node Architecture Using
Real-Time Scheduling Techniques, IEEE ACCESS, vol.
9, No. 21417120, pp. 167305-167321, 2020.
[5] T. Ahmed et al., ”Dynamic routing, spectrum, and
modulation-format allocation in mixed-grid optical net-
works,” in Journal of Optical Communications and
Networking, vol. 12, no. 5, pp. 79-88, May 2020, doi:

This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
 This article has been accepted for publication in IEEE Photonics Journal. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/JPHOT.2022.3205412
10.1364/JOCN.378370.
[6] L. R. Costa and A. C. Drummond, ”Dynamic
Multi-Modulation Allocation Scheme for
Elastic Optical Networks,” 2021 International
Conference on COMmunication Systems &
NETworkS (COMSNETS), 2021, pp. 596-604,
doi: 10.1109/COMSNETS51098.2021.9352827.
[7] ITU-T C1284, ”Proposal of subjects to be discussed
regarding flexible grids”, January 2011.
[8] ITU-T C1288, ”Extension of Rec. G.694.1 by a new
clause to address flexible frequency grids”, January
2011.
[9] S. Alam, M. U. Masood, I. Khan, A. Ahmad,
S. Ghafoor and V. Curri, ”Modeling Off-line
Routing and Spectrum Allocation Problem in
Elastic Optical Network,” 2021 International
Conference on Electrical, Communication, and
Computer Engineering (ICECCE), 2021, pp. 1-6, doi:
10.1109/ICECCE52056.2021.9514112.
[10] H. C. Le and T. L. Nguyen, ”Dynamic Control
Algorithm for Intra-node Group Routing Restricted
Elastic Optical Networks,” 2020 IEEE Eighth
International Conference on Communications
and Electronics (ICCE), 2021, pp. 129-134, doi:
10.1109/ICCE48956.2021.9352130.
[11] Y. Zhou, Q. Sun and S. Lin, ”Link State Aware
Dynamic Routing and Spectrum Allocation
Strategy in Elastic Optical Networks,” in IEEE
Access, vol. 8, pp. 45071-45083, 2020, doi:
10.1109/ACCESS.2020.2977612.
[12] A. N. Khan, H. Yousif Ahmed, M. Zeghid, W. A.
Imtiaz and Z. H. Khan, ”Link Congestion Aware
Proactive Routing for Dynamic Traffic in Elastic Op-
tical Networks,” in IEEE Photonics Journal, vol. 13,
no. 1, pp. 1-15, Feb. 2021, Art no. 7900815, doi:
10.1109/JPHOT.2021.3052650.
[13] A. N. Khan, H. Yousif Ahmed, M. Zeghid, S. Brahim
Belhaouari, W. A. Imtiaz and K. S. Khattak, ”Hy-
brid Optimized Approaches for Routing and Resource
Reservation Protocols in Optical Networks,” in IEEE
Photonics Journal, vol. 13, no. 4, pp. 1-13, Aug. 2021,
Art no. 7200813, doi: 10.1109/JPHOT.2021.3102871.
[14] Khan, A.N., ”A Constrained Maximum Available Fre-
quency Slots on Path Based Online Routing and Spec-
trum Allocation for Dynamic Traffic in Elastic Optical
Networks”, International Journal of Electronics and
Telecommunications, vol. 66, no. 4, 2020, pp. 707-714.
[15] A. N. Khan, ”Online routing, distance-adaptive modu-
lation, and spectrum allocation for dynamic traffic in
elastic optical networks,” Opt. Fiber Technol., vol. 53,
2019, Art. no. 102026.
[16] F. I. Calderon et al., ”BER-Adaptive RMLSA Algo-
rithm for Wide-Area Flexible Optical Networks,” in
IEEE Access, vol. 8, pp. 128018-128031, 2020, doi:
10.1109/ACCESS.2020.3008883.
[17] S. Zhang, Z. Xu, S. Du and J. Wu, ”Routing and Spec-
trum Assignment Algorithm with Traffic Prediction and
This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/
Periodic Rerouting in Elastic Optical Networks,” 2019
IEEE 11th International Conference on Communication
Software and Networks (ICCSN), 2019, pp. 17-21, doi:
10.1109/ICCSN.2019.8905247.
[18] X. Wang, K. Kuang, S. Wang, S. Xu, H. Liu and
G. N. Liu, ”Dynamic routing and spectrum allocation
in elastic optical networks with mixed line rates,” in
Journal of Optical Communications and Networking,
vol. 6, no. 12, pp. 1115-1127, Dec. 2014.
[19] R. Zhou et al., ”Spectrum Allocation Algorithm
for Satellite Elastic Optical Network Based on
Spectrum Resource Assessment Set,” 2021 19th
International Conference on Optical Communica-
tions and Networks (ICOCN), 2021, pp. 1-3, doi:
10.1109/ICOCN53177.2021.9563638.
[20] N. Kitsuwan, K. Akaki, P. Pavarangkoon and A. Nag,
”Spectrum allocation scheme considering spectrum slic-
ing in elastic optical networks,” in Journal of Optical
Communications and Networking, vol. 13, no. 7, pp.
169-181, July 2021, doi: 10.1364/JOCN.422915.
[21] B. C. Chatterjee, W. Fadini, and E. Oki, ”A spectrum
allocation scheme based on first-last-exact fit policy for
elastic optical networks”, J. Netw. Comput. Appl., vol.
68, pp. 164-172, June 2016.
[22] J. Comellas, J. Perelló, J. Sole-Pareta and G.
Junyent, ”Using Spatial Division Multiplexing to
Avoid Fragmentation in Flex-Grid Optical Networks,”
2020 22nd International Conference on Transpar-
ent Optical Networks (ICTON), 2020, pp. 1-4, doi:
10.1109/ICTON51198.2020.9203291.
[23] Y. Yin et al., ”Spectral and spatial 2D fragmentation-
aware routing and spectrum assignment algorithms in
elastic optical networks [invited],” in Journal of Optical
Communications and Networking, vol. 5, no. 10, pp.
A100-A106, Oct. 2013, doi: 10.1364/JOCN.5.00A100.
[24] B. C. Chatterjee, N. Sarma and E. Oki, ”Routing and
Spectrum Allocation in Elastic Optical Networks: A Tu-
torial,” in IEEE Communications Surveys & Tutorials,
vol. 17, no. 3, pp. 1776-1800, third quarter 2015, doi:
10.1109/COMST.2015.2431731.
[25] B. C. Chatterjee, A. Wadud and E. Oki, ”Proactive
Fragmentation Management Scheme Based on
Crosstalk-Avoided Batch Processing for Spectrally-
Spatially Elastic Optical Networks,” in IEEE
Journal on Selected Areas in Communications,
vol. 39, no. 9, pp. 2719-2733, Sept. 2021, doi:
10.1109/JSAC.2021.3064594.
[26] B. C. Chatterjee, S. Ba and E. Oki, ”Fragmentation
Problems and Management Approaches in
Elastic Optical Networks: A Survey,” in IEEE
Communications Surveys & Tutorials, vol. 20,
no. 1, pp. 183-210, Firstquarter 2018, doi:
10.1109/COMST.2017.2769102.
[27] M. Hadi and E. Agrell, ”Iterative Configuration in
Elastic Optical Networks : (Invited Paper),” 2020
International Conference on Optical Network De-
sign and Modeling (ONDM), 2020, pp. 1-3, doi:
 This article has been accepted for publication in IEEE Photonics Journal. This is the author's version which has not been fully edited and
content may change prior to final publication. Citation information: DOI 10.1109/JPHOT.2022.3205412

10.23919/ONDM48393.2020.9133016.
[28] A.N. Khan, Z.H. Khan, K.S. Khattak, A. Hafeez, ”Joint
routing, link capacity dimensioning, and switch port
optimization for dynamic traffic in optical networks”,
ETRI Journal, 43(5), pp. 799-811, 2021.
[29] P. Papanikolaou, K. Christodoulopoulos and E. Var-
varigos, ”Optimization techniques for incremental
planning of multilayer elastic optical networks,” in
Journal of Optical Communications and Network-
ing, vol. 10, no. 3, pp. 183-194, March 2018, doi:
10.1364/JOCN.10.000183.
[30] L. H. Bonani, M. Forghani-elahabad and M. L. F.
Abbade, ”Network Fragmentation Measure in Elastic
Optical Networks,” 2019 SBFoton International Optics
and Photonics Conference (SBFoton IOPC), 2019, pp.
1-5, doi: 10.1109/SBFoton-IOPC.2019.8910225.
[31] J. Yuan et al., ”A Spectrum Assignment Algorithm
in Elastic Optical Network With Minimum Sum
of Weighted Resource Reductions in all Associated
Paths,” in Journal of Lightwave Technology, vol.
37, no. 21, pp. 5583-5592, 1 Nov.1, 2019, doi:
10.1109/JLT.2019.2935506.
[32] Python, www.python.org.
[33] Z. Zhu, W. Lu, L. Zhang and N. Ansari, ”Dynamic Ser-
vice Provisioning in Elastic Optical Networks With Hy-
brid Single-/Multi-Path Routing,” in Journal of Light-
wave Technology, vol. 31, no. 1, pp. 15-22, Jan.1, 2013,
doi: 10.1109/JLT.2012.2227683.

This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/