Improving Broadcast System of Integrated Satellite-Terrestrial Network-Based On Enhanced Ant Colony Optimization
Improving Broadcast System of Integrated Satellite-Terrestrial Network-Based On Enhanced Ant Colony Optimization
Improving Broadcast System of Integrated Satellite-Terrestrial Network-Based On Enhanced Ant Colony Optimization
Online ISSN 1841-9844, ISSN-L 1841-9836, Volume: 19, Issue: 1, Month: February, Year: 2024
Article Number: 5632, https://doi.org/10.15837/ijccc.2024.1.5632
CCC Publications
Deepa V, Sivakumar B
Deepa V
Department of Computing Technologies, School of Computing
SRM Institute of Science and Technology, Kattankulathur
Tamilnadu-603203, India
dv1018@srmist.edu.in
Sivakumar B*
Department of Computing Technologies, School of Computing
SRM Institute of Science and Technology, Kattankulathur
Tamilnadu-603203, India
*Corresponding author: sivakumb2@srmist.edu.in
Abstract
With the use of seamless high-speed worldwide network connectivity in future, it is anticipated
that the integrated satellite-terrestrial network (ISTN) will be a possible option. Due to the inad-
equacy of topological data and the close coupling of data and control planes, deploying optimized
rules on routers is difficult. The important factors in ISTN networks are routing rules and policies,
link failure, and high-bandwidth communication. Software-defined networking (SDN) is an open
innovation approach that enables programmability from a central location. The controller handles
the complexity of the network, whereas the infrastructure layer devices relay the packets. Thus, we
investigated the broadcast by dynamically adjusting routes for fault tolerance and energy efficiency
of the ISTN using distance between nodes. In addition, using dynamic source routing, this study
examines the competence functions of all managing nodes in a network. The routing design is dis-
tributed among all nodes to create numerous collective paths. For LEO satellite networks, the ant
colony optimization-based routing algorithm is an improved version that considers the probability
of faults and low-energy consumption. The proposed simulation ensures a seamless transition in
the event of failures and avoids the requirement for an additional coordination service.
Keywords: ant colony optimization, broadcast, dynamic source routing, energy efficiency, fault
tolerance, Integrated Satellite-Terrestrial Network (ISTN), Software-Defined Networking (SDN).
1 Introduction
In recent days, Integrated Satellite-Terrestrial Networks (ISTNs) [32] [31] have the ability to offer an
immediate wireless networking solution in situations where no pre-deployed infrastructure is present.
https://doi.org/10.15837/ijccc.2024.1.5632 2
The necessity for ISTNs arises from circumstances in which nodes, such as mobile phones and laptops,
must group together to form a network capable of supporting services like messaging, resource sharing,
and file-sharing. So, the main goal of ISTN routing is to build one (unicast) or more (multicast)
reliable end-to-end channels (also called "routes") between the nodes as soon as possible so that they
can communicate in a reliable or valid way.
Software-Defined Networking (SDN) [11] [13], with its centralized overview of the entire network,
presents significant opportunities in the context of broadcast. The data plane, the control plane, and
the application plane are the three regions in which broadcast can be considered in SDN. Failures of
links or switches make up data plane concerns, whereas controller connections to switches or controller
failures themselves fall within the control plane region. The application plane region is concerned with
the failure of an application that may impact the northbound API, which may subsequently affect
other applications over time. The southbound interface (SBI) is the communication between controller
and networking elements at data layer. In this study, we concentrate on the SDN data plane’s (LEO)
fault tolerance.
By using dynamic source routing, two key strategies rehabilitation and protection for failure recov-
ery in the LEO are made. To account for potential future failures, alternate rules for active flows are
set for protection. New routing rules for impacted flows were computed during rehabilitation while
considering the most recent network view. As shown in Figure 1, the GEO is the entity in charge of
this procedure. The corresponding switch first notices a failure in the LEO, and the GEO (controller)
[14] [28] then generates an event to start the path-calculating procedure. After the event has been
dealt with in the northbound interface (NBI) application, a new route is determined. Based on the
conduct of harvester ants looking for a connection between their territory and a foodstuff, an enhanced
ant colony optimization protocol was created to find the best path in a network.
The goal of incorporating Enhanced Ant Colony Optimization (Enhanced-ACO) into dynamic
source (DS) routing is to identify and preserve the optimum paths between the nodes. The Enhanced-
ACO can nevertheless provide an efficient route, despite requiring processing power [25] from individual
https://doi.org/10.15837/ijccc.2024.1.5632 3
nodes. This algorithm can reliably update the forwarding table according to the ACO’s degree of self-
organization and adaptation.
• We suggest and improve a method to assess a link’s and determine its congestion level.
• To provide a new route discovery technique that not only guarantees reduced end-to-end packet
delivery delays but also resolves a number of conflicts, such as the looping conflict of earlier
approaches.
• We detect congestion by determining the route based on the duration, traffic, and dependability
of the method. In addition, explore the Enhanced-ACO of the pheromone decay method for
maintaining routes and calculate the energy required for a node by analyzing the proposed
scheme.
• To this aim, we make a design of congestion avoidance with time frame, fault-tolerant controller
architecture and network lifetime is increasing by energy efficient.
1.2 Motivation
This study integrates technologies such as SDN, satellite networks, and virtualization to enhance
network programmability, congestion, interoperability, and control dynamic network changes. Adopt-
ing SDN in satellite-terrestrial networks can improve throughput, packet loss, and delay in networking
scenarios. With the help of SDN controllers, heterogeneous networks can use their resources together
and set up custom services in a way that changes as the network structure varies.
The paper is organized as Section II, which impersonates the literature surveys related to the
different methodologies incorporated for predicting fault tolerance. Section III includes data sources,
proposed enhanced architectures, and rerouting mechanisms. Section IV discusses the performance
metrics that are employed for testing the effectiveness and provides a comparative analysis between the
proposed model and the conclusion, along with the future direction of this study, which is presented
in Section V.
2 Related Study
The work related to the proposed broadcast and optimization-based routing algorithm is described
as follows.
The broadcast networks have gained significant attention in recent years owing to their ability to
optimize and manage complex networks. This review provides an overview of the different types of
broadcast systems for satellite systems with a focus on minimum energy consumption. The theoretical
exploration and development of ISTNs are currently in progress. Liu et al. [17] proposed a network
architecture called space-air-ground integrated network (SAGIN) for ISTN, in which concerns with
system implementation, protocol optimization, and resource management are also closely examined.
An evaluation of exemplary ISTN architectures and a summary of pertinent studies that concentrate
on wireless transmission, control and maintenance, resource allocation, and security can be found in
Cao. et.al., [2] Key network functionalities to ensure Quality of Service (QoS) in ISTNs are outlined
in [6] and is based on cross-section SDN controllers from the perspective of the varied requirements of
future applications and the lengthy transmission delay of satellite-to-earth links.
Three groups of fault diagnosis (FD) methods: model, signal, knowledge based — can be dis-
tinguished. Since the innovative research by Jones [10], observer-based techniques have been widely
studied and deployed in the field of well-known FD methods. Excellent references [18], [4], and
[8] include a number of survey studies on fault diagnosis and the three fundamental tasks of fault
identification, separation, and evaluation, which constitute fundamental fault isolation techniques.
https://doi.org/10.15837/ijccc.2024.1.5632 4
Nevertheless, there is still a concern regarding how to combine the three tasks (fault diagnosis, fault
isolation, and tolerance) into a single, well-thought-out plan.
where,
max Ii=1 kk=1 A(i, j, k) > l1 ⇐⇒
PI Pk
k=1 A(i, j, k) > l2 ∀ l1, ∀ l2, l1 ̸= l2
P P
i=1
r random ants assesses at iteration T. Each ant’s fitness in accordance with an even-handed func-
tion, and then uses formula (2) to update the pheromone concentration of each potential trail. The
pheromone evaporation technique, which involves evaporating the pheromone from every cell is the
initial step,
τi,j,k (r, T )
Pi,j,k (rT ) = PI (3)
i=1 τi,j,k (r, T )
Figure 3: Ant-colony improvised Network structure for achieving optimal path (with less pheromone
Now, the pheromone values are associated with an improvement phase for direct ants in subsequent
rounds to search in a more successful region. The convergence factor of iterative method (cfir ) is the
pheromone reward and K fitness value is the objective function defined by (4) and specifying the effect
of path and attractiveness, respectively, and taking values larger than 0,
PK
k=1 (τ i,j,k max∀ i,j ) − (τ i,j,k min∀ i,j )
cf ir = (4)
K
After the coverage range, following the broadcast step of wavelength, the transmitting power and
nearest node are known to the receiving node with an echo message. As a result, it is quite simple
to determine the threshold value of each nearest node with minimal computation time. It is evident
that various neighbors have different threshold values because they have varied specifications, and the
threshold value for a given neighbor node is fixed, as it is independent of the node placement.
The fault detection in equation (5) principles produced by ant colony fault diagnosis and classification
are used to determine if the network is defective and, if so, what sort of system fault occurred. When
a defect is detected, the system first updates the fault sign and determines whether it is capable
of in-orbit autonomous maintenance. If so, the system’s operating mode would change to enter the
fault-handling mode [19] in four distinct ways:
• The minor fault-mf system changes the working state to fault handling and converts the working
mode into entering fault handling mode.
• The general fault-gf system would spin to the NULL point; therefore, it would remain still for
return fault handling and awaits commands from the ground.
• The system would attain the serious fault-sf when changing the working state to fault handling
failure and turns the working mode into entering fault handling mode and wait for commands
from the ground
• Incorrect fault handling-if results, after experiencing a minor or general failure, if the servo
system is unable to rotate to a given position within a predetermined timeframe, it will change
the functioning status to fault handling failure. The mechanism would shut the engine off after
that and hold back for instructions from the ground.
is unable to discover any superior solutions. The occurrence of stagnation has the potential to result
in congestion at specific nodes within a given locality. In order to address this issue, we propose
the utilization of a congestion-avoidance time frame. This time frame aims to estimate the level of
congestion on a given link by taking into account the cache occupancy of said link. The variability in
length of intersatellite links necessitates consideration of the fluctuations in communication links (li ).
We design the associative learning property of time frame "A" (binary value: i, j, k) matrix of link (i,
j) at time t.
where, ERI = EDisseminate +EBroadcast +Ereply . As a result, the sender node broadcasts by Requisition
Ant are equivalent to number of neighbor nodes has around 0.9 RN , where RN is the transmission
satellite range of the node.
4 Performance Metrics
4.1 Datasets
It is required to take into account the ground stations (LEO), in order to determine end-to-end route
characteristics. The ground stations (LEO) include stationary facilities used by the constellation’s
controller and user terminals, such as ship trucks, airplanes, and automobiles. In order to determine
the suggested method performance, the process of solving a total of 7 benchmarks taken from [12].
Table 1 presents information on the characteristics of the benchmark dataset.
Assuming that satellites(i) and ground stations (LEO) can support a limitless number of con-
nections, the optimal architecture of a network would compute each link that is physically feasible
at each step time. When a satellite(i) is at an altitude that is greater than or equal to a certain
threshold over the horizon, it is possible for ground stations (LEO) to communicate with it. The
https://doi.org/10.15837/ijccc.2024.1.5632 9
inter-satellite connections are considered to be operational, provided that the achieve is known as a
minimum user-defined distance or communication altitude above the surface of the Earth. The num-
ber of connections in the network may rapidly reach millions when dealing with large constellations
located at high altitudes. In order to precisely take into account, the rotation of the Earth, the link-
ages between satellites and ground stations (LEO) must be computed at each and every step time.
The value of cf ir is considered by the algorithm when determining whether a pheromone should be
re-considered. Depending on the current condition of cf ir , the pheromone intensification approach will
implement one of the five available pheromone update processes. The intensify or reinforce weights W
for faults Fmf ,Fgf ,Fsf ,Fif , respectively, are denoted by Wmf Wgf ,Wsf ,Wif , (Table 2). The threshold
values for the partition of the five phases were as follows: cf ir = [0.2, 0.4, 0.6, 0.8, 0.85] for broadcast
requests r = 1,..., 5.
When cf ir is greater than or equal to cf i5 , the proposed algorithm performs the pheromone re-
consideration procedure. During this procedure, all pheromone are set to the initial values τi,j,k (T = 0).
This is immediately followed by the pheromone update procedure, which uses Wmf as the intensifica-
tion solution. The algorithm continues to search for the most important fields in subsequent iterations.
The iteration loop T continues until one of its termination criteria is fulfilled. These requirements
may include reaching the maximum number of possible iterations or finding a satisfactory solution.
4.2 Simulation
The simulator would simply generate massive amounts of location data and linkages if it did not
include the important and highly useful capability of creating a 3D picture of a satellite network
circling Earth. This is a crucial and extremely practical feature, and visualization is performed using
the Systems Tool Kit (STK) [30] library, which is an application programming interface for OpenGL.
The simulation environment parameters are shown in Table 3. Upon setup, the FaultSystem class
creates the STK service network pipelines. These network pipelines include the satellite point, inter-
satellite linkages, satellite-ground links, ground stations, Earth model, and shortest route.
As shown in Figure 5., each pipeline may have its own color, the other properties (no. of planes,
links, earth model), and various layers such as links, ground stations (LEO), satellites (i), and so on,
https://doi.org/10.15837/ijccc.2024.1.5632 10
can have their visibility activated or disabled independently. The FaultSystem class also uses a method
that exports a PNG file that shows the rendered model at each time step. This PNG file can then be
merged with other animations to produce a GIF or any other type of moving image. All evaluations
were simulated using the Systems Tool Kit (STK) of the space terrestrial platform to share the results
on the integrated Mininet SDN platform [9] and executed on an Intel Xeon processor at 3.6GHz.
The proposed enhanced ant colony optimization (Enhanced-ACO) into a dynamic source (DS)
routing algorithm in large LEO networks concentrates on a select group of source-destination pairings
that have been spatially mapped to the sites of major cities, as shown in Table 1. The simulator is
used to perform a fault analysis of link congestion and energy consumption increases by considering a
wide variety of constellation configurations and finding the shortest path from the hop count.
4.3 Analysis
In the proposed transitions, consider the extent of Ants = 80, ρ = 0.2, T (iterations) = 45, intact
configuration (the set of links recalculated at each time step), and S2SLink (the satellite-satellite
linkages remain static and are used only when necessary). Here, the experimental results show that
the intact network design provides the least number of faults feasible along all links, and the S2SLink
design for a network provides higher faults along all possible pathways, and there is often more
variation. Figure 6(a). illustrates the statistics for an LEO constellation (40, 40), which occurs at an
altitude of 500 km and an inclination of 65°. The connection may only travel between orbital planes
by relaying through a ground station, so the greater faults caused by S2Slink network are most evident
in east-west connections such as BM6 (Newport University – Los Angeles).
Figure 6(b) shows the percentage of fault detection accuracy, and it is evident that the intact
accuracy rate is 3.898% higher than the S2Slink. It represents the failure accuracy rate of existing
methods (+) grid [12] and Stochastic Diffusion search method hybridized with Differential Evolution
(SD − BDE) [27], which reported the lowest accuracy percentage and also failed with respect to time.
https://doi.org/10.15837/ijccc.2024.1.5632 11
The energy efficiency for the same set of network architectures is shown graphically in Figure
7. with a series of charts. The optimal architecture of a network would result in fewer hops, with
the exception of the direct route from Newport University to London. This is because the intact
and S2SLink network designs allow connections to be made between neighboring satellites. If the
architecture is perfect, the connections will be able to "skip" several nearby satellites, which will result
in reduced overall energy efficiency, especially for long-distance connections. In addition, periodic
variations in energy efficiency for the intact and S2SLink network designs occur for the same reason
as the periodic abrupt transitions in faults. These variations were caused by periodic changes in the
amount of time required for each hop. When ground stations (data plane) transition between the first
and last hop satellites, a drastic change in route length frequently results in a change in the hop count.
This is because the hop count is determined by the number of satellites passed. However, in an ideal
network architecture, periodic fluctuations in hop count are less noticeable than would otherwise be
because longer connections are better able to rapidly absorb changes in route length.
In BM5 (Newport Univ. to London), has energy efficiency and fewer hops than other BM, compar-
atively the other ideal network architecture suffers from poor route stability. In this regard, the intact
design provides noticeably improved stability in comparison to the S2SLink network designs with
an average of 35.6 seconds passing between route modifications. Experimental results show that the
optimum design of intact has a duration between change intervals that are 0.05 seconds on average,
whereas the S2SLink design has 0.45 seconds.
Funding
The APC was funded by R&D center "SRM Institute of Science and Technology, Research Lab"
of SRM University.
Author contributions
The authors contributed equally to this work.
Conflict of interest
The authors declare no conflict of interest.
References
[1] Al-Hraishawi, H., Minardi, M., Chougrani, H., Kodheli, O., Montoya, J. F. M., and Chatzinotas,
S. (2021). Multi-layer space information networks: Access design and softwarization. IEEE Access,
9, 158587-158598.
[2] Cao, X., Yang, P., Alzenad, M., Xi, X., Wu, D., and Yanikomeroglu, H. (2018). Airborne com-
munication networks: A survey. IEEE Journal on Selected Areas in Communications, 36(9),
1907-1926.
[3] Chatterjee, S., and Das, S. (2015). Ant colony optimization based enhanced dynamic source
routing algorithm for mobile Ad-hoc network. Information sciences, 295, 67-90.
https://doi.org/10.15837/ijccc.2024.1.5632 13
[4] Chen, X., Liu, C. Y., Proietti, R., Li, Z., Yoo, S. B. (2022). Automating optical network fault
management with machine learning. IEEE Communications Magazine, 60(12), 88-94.
[5] Coondu, S., Mitra, A., Chattopadhyay, S., Chattopadhyay, M., and Bhattacharya, M. (2014,
February). Network-coded broadcast incremental power algorithm for energy-efficient broadcast-
ing in wireless ad-hoc network. In 2014 Applications and Innovations in Mobile Computing
(AIMoC) (pp. 42-47). IEEE.
[6] Du, J., Jiang, C. (2022). Cooperative Beamforming for Secure Satellite-Terrestrial Transmis-
sion. In Cooperation and Integration in 6G Heterogeneous Networks: Resource Allocation and
Networking (pp. 129-164). Singapore: Springer Nature Singapore.
[7] Fakhar, U., Khan, H. Z., Tariq, Z., Ali, M., Akhtar, A. N., Naeem, M., Wakeel, A. (2023). Radio
resource allocation for energy efficiency maximization in satellite–terrestrial integrated networks.
Ad Hoc Networks, 138, 103001.
[8] Gu, R., Qin, J., Dong, T., Yin, J., Liu, Z. (2020). Recovery routing based on q-learning for
satellite network faults. Complexity, 2020, 1-13.
[10] Jones, H. L. (1973). Failure detection in linear systems (Doctoral dissertation, Massachusetts
Institute of Technology).
[11] Kazmi, S. H. A., Qamar, F., Hassan, R., Nisar, K. (2023). Routing-based interference mitigation
in SDN enabled beyond 5G communication networks: A comprehensive survey. IEEE Access.
[12] Kempton, B., and Riedl, A. (2021, June). Network simulator for large low earth orbit satellite
networks. In ICC 2021-IEEE International Conference on Communications (pp. 1-6). IEEE.
[13] Kianpisheh, S., Taleb, T. (2022). A survey on in-network computing: Programmable data plane
and technology specific applications. IEEE Communications Surveys and Tutorials.
[14] Li, T., Zhou, H., Luo, H., Yu, S. (2017). SERvICE: A software defined framework for integrated
space-terrestrial satellite communication. IEEE Transactions on Mobile Computing, 17(3), 703-
716.
[15] Liang, Y. C., Tan, J., Jia, H., Zhang, J., and Zhao, L. (2021). Realizing intelligent spectrum
management for integrated satellite and terrestrial networks. Journal of Communications and
Information Networks, 6(1), 32-43.
[16] Lin, Z., Lin, M., Champagne, B., Zhu, W. P., Al-Dhahir, N. (2021). Secrecy-energy efficient hybrid
beamforming for satellite-terrestrial integrated networks. IEEE Transactions on Communications,
69(9), 6345-6360.
[17] Liu, J., Shi, Y., Fadlullah, Z. M., and Kato, N. (2018). Space-air-ground integrated network: A
survey. IEEE Communications Surveys and Tutorials, 20(4), 2714-2741.
[18] Liu, J., Wei, Z., Zhao, B., Su, J., and Xin, Q. (2021). A probabilistic resilient routing scheme for
low-earth-orbit satellite constellations. In Wireless Algorithms, Systems, and Applications: 16th
International Conference, WASA 2021, Nanjing, China, June 25–27, 2021, Proceedings, Part III
16 (pp. 254-261). Springer International Publishing.
[19] Ma, Z., Zhao, Q., and Wang, S. (2022). Fault Diagnosis and Handling of the Two-Dimensional
Tracking Servo System for Space. Computational Intelligence and Neuroscience, 2022.
[20] Na, Z., Pan, Z., Liu, X., Deng, Z., Gao, Z.,Guo, Q. (2018). Distributed routing strategy based on
machine learning for LEO satellite network. Wireless Communications and Mobile Computing,
2018.
https://doi.org/10.15837/ijccc.2024.1.5632 14
[21] Ni, W., Xu, Z., Zou, J., Wan, Z., and Zhao, X. (2021). Neural network optimal routing algorithm
based on genetic ant colony in IPv6 environment. Computational Intelligence and Neuroscience,
2021.
[22] Qi, H., Guo, Y., Hou, D., Xing, Z., Ren, W., Cong, L., and Di, X. (2022). SDN-based dynamic
multi-path routing strategy for satellite networks. Future Generation Computer Systems, 133,
254-265.
[23] Qiu, C., Yao, H., Yu, F. R., Xu, F., Zhao, C. (2019). Deep Q-learning aided networking, caching,
and computing resources allocation in software-defined satellite-terrestrial networks. IEEE Trans-
actions on Vehicular Technology, 68(6), 5871-5883.25
[24] Rangisetti, A. K., and Sathya, V. (2020). QoS aware and fault tolerant handovers in software
defined LTE networks. Wireless Networks, 26, 4249-4267.
[25] Ruan, Y., Li, Y., Zhang, R., and Jiang, L. (2022). Energy efficient power control for cognitive
multibeam-satellite terrestrial networks with poisson distributed users. IEEE Transactions on
Cognitive Communications and Networking, 8(2), 964-974.
[26] Ruan, Y., Jiang, L., Li, Y., Zhang, R. (2020). Energy-efficient power control for cognitive satellite-
terrestrial networks with outdated CSI. IEEE Systems Journal, 15(1), 1329-1332.
[27] Salman, A. A., Ahmad, I., and Omran, M. G. (2015). A metaheuristic algorithm to solve satellite
broadcast scheduling problem. Information Sciences, 322, 72-91.
[28] Shi, Y., Cao, Y., Liu, J., and Kato, N. (2019). A cross-domain SDN architecture for multi-layered
space-terrestrial integrated networks. IEEE Network, 33(1), 29-35.
[29] Wieselthier, J. E., Nguyen, G. D., and Ephremides, A. (2001). Algorithms for energy-efficient
multicasting in static ad hoc wireless networks. Mobile Networks and Applications, 6, 251-263.
[31] Zhang, S., Zhu, D., and Wang, Y. (2020). A survey on space-aerial-terrestrial integrated 5G
networks. Computer Networks, 174, 107212.
[32] Zhu, X.; Jiang, C. (2022). Integrated Satellite-Terrestrial Networks Toward 6G: Architectures,
Applications, and Challenges, IEEE Internet of Things Journal, 9(1), 437-461, 2022.
https://doi.org/10.15837/ijccc.2024.1.5632 15