Volume 8, Issue 10, October – 2023 International Journal of Innovative Science and Research Technology

ISSN No:-2456-2165

A Framework for Monitoring Network Node

Failure using Mobile Agents
M. O Lawal*1; K. G Akintola2; O. K Boyinbode3; N.C Onyeka4
1
Department of Computer Science, Federal Polytechnic Ede
2
Department of Computer Science, Federal University of Technology, Akure
3
Department of Computer Science, Federal University of Technology, Akure
4
Department of Computer Science, Federal Polytechnic Ede

Abstract:- Fault detection is an essential aspect of information overwhelming the network becomes especially
conducting fault diagnosis for computer networks. It severe, particularly since a quick solution is imperative.
comprises of two phases: fault detection and fault Swift diagnosis and resolution of the problem either through
localization. The use of mobile agents for detecting faulty automated means or by informing and guiding a human
nodes on a network is a concept aimed at ensuring the operator on the appropriate course of action becomes
proper functioning of networks. This research aims to crucial. Devices such as routers, hubs, servers, and more are
design a fault detection framework for a network system monitored by the manager and when there are faults within
using a mobile agent. Light Weight Agent (LWA) travels the network, the application manager within the network
within the nodes to detect nodes that are down on the notifies the network manager in real-time.
network and returns true or false along with other
information as the status of each node visited. The Operators working with large networks must remotely
system is designed using software agents. This interact with numerous devices from their management
subsystems of the system include the Agent Controller, workstation. To cater for the diverse range of network
Server Agent, Client Agent, Check Status and the components, management applications feature a plethora of
database. The Agent Controller allocates and determines interfaces and tools. However, network management
the agent functions using a unique identification number. systems are often designed as large monoliths, making them
The server agent controls the activities of the client agent challenging to maintain.
by monitoring the migration of each of the probing
agents to each node on the network. The system is Automatic discovery is a crucial aspect of network
implemented using the Java Application Development management systems, with various objectives depending on
Environment (JADE) platform. It was tested on a the scope of the system. At its most basic level, discovery
network with twenty nodes, for five hours per day for aims to locate all devices present within the network.
twenty days. The system achieved a reliability rate of However, an expanded version of this function involves
100% for the highest and 47% for the lowest. This constructing detailed views that encompass additional
research work will be beneficial for testing the reliability information, such as the services offered by each devices
of a networking system to ensure optimal functioning. that meet specific criteria. As the process of identifying the
Future research will focus on using mobile agents to problem becomes more complex, it becomes harder to
diagnose faulty nodes on a network. implement using traditional client/server methods.

Keywords:- Mobile Agent, System Reliability, Computer This research emerged from the exigency to use an
Network, JADE, Fault Detection. agent to detect network faults/failures using intelligent
decision-making agents. It also came from the reading
I. INTRODUCTION literature reviews of previous researchers such as [6] on how
to solve the problem of a complete recovery mechanism in
Nowadays computer networks is becoming very large, case of fault/ failure within a network without simulation.
covering the vast majority of geographical locations. The study by Jian Hu et al. (2008) enables users to define
Network Management Applications (NMA) designed to their own Management Information Base (MIB) tables, but
manage network tasks such as, maintenance and this also results in increased system complexity, as mobile
administration of the network were also designed to manage agents must communicate directly with the managed system,
traditional client/server networks. However, as computer which may impact system compatibility. The primary
networks expand, the size and complexity of client/server objective of this research is to leverage mobile agents in
models are faced with the problem of scaling and flexibility managing today's large and diverse networks. Mobile agent
[6]. software objects are autonomous and can move from one
node to another, carrying logic and data to perform tasks on
Researchers in the field of software mobile agents are behalf of the user. The network management software
now focusing their attention on Network management objects based on mobile agents will be equipped with agents
systems. However, if there is a malfunction, the issue of possessing network management capabilities that will enable

IJISRT23OCT1389 www.ijisrt.com 1232

Volume 8, Issue 10, October – 2023 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
them to issue requests to managed devices or nodes once separation, identification, and appraisal will also be
they migrate to these nodes. ineffective.

II. REVIEW OF RELATED WORK III. MOBILE AGENT

In this section, reviewed literature related to network Mobile agents are programs designed to function
faults, network fault detection, Mobile Agent, system automatically moving from node to node. They can perform
reliability and Network reliability are as follows: Mobile a task on behalf of users and allow difficult tasks to be
Agents in [1], [2], [3], [4], and [40]. Network fault detection shared amongst the agents [1], [2], [3], [4]. The primary goal
in [9], [10], [11] [12], [34], [40], [43], [49]. System of using mobile agents in the management of
reliability [11], [17], [45], [48]. Characteristics of Mobile telecommunication networks is reducing network traffic by
Agents are as presented in [4], [6], and [30]. Network using load balancing and building scalable and reliable
reliability in [12], [15], [28], [29], [31], [33]. Network distributed network management systems. Some of the
management and monitoring in [33], [34], [35]. advantages of using agent technology in telecommunication
networks are as follows:
 Fault Identification
Fault identification is used to understand the elemental  Addresses the handling of a large volume of data that
failure mode, ascertain the margin of the fault, and find the agents can explore, gather, and filter.
core cause. Fault identification methods may differ, but the  Facilitates the utilization of more intelligent techniques
strides to follow are mainly identical. to manage a network, integrate different services into
value-added services, and negotiate quality of service.
 A physical fault is a type of network failure that is  Promotes the development of higher-level
related to hardware issues. communication and organization within a network.
 Port faults typically fall into two categories: unstable  Demonstrates reactivity, as agents can promptly respond
ports and port failures.. to local events, such as link failures.
 When switches or routers break down, it's often due to  Exhibits robustness, as agents can perform their duties to
equipment damage resulting in abnormal network some extent, even when parts of the network are
behavior. temporarily inaccessible. This is particularly crucial in
 Network card faults are considered to be a type of host mobile computing, where links can be expensive and
hardware failure and are a frequent reason for network unstable.
problems.  Distributes management code to Simple Network
Management Protocol (SNMP) agents to reduce
 Fault Detection bandwidth consumption in a wireless network.
Fault detection is the process of locating the existence  Decentralizes network management functions by
of a fault in a network before it presents itself in the form of allowing mobile agents to autonomously and proactively
network failure and breakdown. It is the most important carry out administrative tasks, thereby reducing the
stage of network fault detection (NFD) as all of the amount of management traffic required.
subsequent processes depend on its accuracy. If the  Dynamically adjusts network policies, as mobile agents
equipment is unable to identify the proper failure mode (or if can modify the underlying rules of network management
detection is incorrect and triggers false alarms), then the periodically.

Fig 1 Mobile Agents Model (Singh1 et al., 2012)

IJISRT23OCT1389 www.ijisrt.com 1233

Volume 8, Issue 10, October – 2023 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
 Mobile Agent System Architecture monitor particular network nodes at each stage. By the
Network failure can be prevented if the concept of conclusion of numerous detection stages, every node in the
fault-detection design strategies is used. Fault tolerance is a network can be inspected. This approach guarantees that the
way to head off a failure before it has a chance to happen. traffic generated by probe agents during each detection stage
With the fault-detection technique, a problem such as a is considerably less than conventional methods, although it
software defect in safety-network machines can be identified may take more time to cover the entire network. The idea of
and prevented. The preemptive detection of node failure the strategy to ensure this aforementioned is established in
using mobile agents is the concern of this project work. If the following attributes:
this fault is left uncared for, it could result in network failure
and consequently network machine downtime.  There will be an Agent Controller manager that
identifies the agent type (Server/Client Agent).
In this research work, a fault-detection architecture  Each node should have an individualized fault detection
which is based on a fault identification procedure is used mechanism (client agent) to ensure that its service is not
and includes the following two steps: Fault detection and impaired by any hardware failure or software fault.
fault localization [7]. The initial step in fault diagnosis for  There is reliable and timely delivery of reliable messages
computer networks is detecting the presence of any faults, from nodes to the Server Agent on the shared network.
which involves using detection tools. If any faults are  The individual node transmits at the appointed time slot
detected, fault localization is then initiated to identify the at all times of the network machine’s operation.
location of the fault and the affected node. Therefore, fault
detection is a crucial first step towards ensuring the normal The essence of this work is to foresee the occurrence of
operation of networks, and it is essential to employ fast and a masqueraded fault and prevent it by providing a solution
precise fault detection techniques. The proposed method of before the network fails using mobile agents. In this view, a
fault detection used in this research work is a non- proactive strategy to prevent faults resulting from software-
deterministic environment [45], [49], [51]. The goal is to defect or hardware defects is presented in this work [51] and
partition the detection process into several stages, with a [44].
small number of lightweight agents (LWAs) assigned to

Fig. 2 Mobile Agent System Architecture

IJISRT23OCT1389 www.ijisrt.com 1234

Volume 8, Issue 10, October – 2023 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
IV. THE MATHEMATICAL MODEL RF(t) is a probability function that represents the
likelihood that a system will operate without a fault of class
The main performance metric considered in this F occurring within a given time interval [t_init, t_f) for all f
research work is reliability, which is a crucial aspect of in the set F. In other words, RF(t) measures the probability
engineering design and development. The field of reliability that the system will survive without experiencing any faults
engineering encompasses all stages of a system's lifecycle, of class F during the time interval [t_init, t_f) for all possible
from design to fabrication, with the goal of minimizing the faults F that may occur. RF (t) can be calculated using the
risk of equipment failure. Neglecting reliability can result in reliability function R(t) as:
severe consequences, including the loss of critical
information or the erosion of trust in the system. Moreover, RF (t) = R(tf | tinit ≤ t)
acceptable levels of reliability may differ depending on the
application environment [45] and [48]. Where R(tf | tinit ≤ t) is the conditional reliability of
the system at time tf given that it has operated successfully
The relationship between reliability parameters and until time t. It can be calculated as:
probability theory can be expressed as follows: Suppose a
fixed number N0 of identical items is being tested, and Ns is R(tf | tinit ≤ t) = R(tf)/R(t)
the number of items that survived after a certain time period
t, while Nf is the number of items that failed during the Where R(tf) is the reliability of the system at time tf
same period. Then, for all t, and R(t) is the reliability of the system at time t. Failure
Probability Qf (t), is complementary to Rf (t)
N0 = Ns + Nf. (1)
Rf (t) + Qf (t) = 1 (5)
If N0 is sufficiently large, the reliability R(t) of an item
can be calculated as Ns divided by N0. We can remove the subscript 'f' and write the equation
as R(t) + Q(t) = 1, where R(t) is the reliability function and
The failure rate function lambda (λ(t)) is defined as the Q(t) is the probability of failure function.
instantaneous rate of failure at time t, which can be
mathematically expressed as: If the lifetime of the system is exponentially
distributed, the probability of no failure occurring in the
λ(t) = -1/R(t) * dR(t)/dt (2) time interval [tinit, t] is given by:

Where R(t) is the reliability function of the system. RF (t) = e^(-λ(t-tinit))

The negative sign in front of the fraction indicates that
lambda λ(t) is a decreasing function of time t, as the R (t) = e−λ`t (6)
reliability function R(t) decreases over time. The failure rate
function λ(t) is an important concept in reliability Where ‘λ’ is called the failure rate.
engineering and is used to estimate the probability of failure
of a system over a given time interval. Since this research work is employing TTP/C (Time-
Triggered Protocol/Clock-Synchronized) which makes use
R(t) = ℮ - λ(t)dt (3) of the Time Distributed Media Access mechanism TDMA.
The research hereby assumes that the operation of each node
The survival probability function, as defined in successively takes place as they take part in the schedule
equation (2), is commonly known as the reliability function. round at their allotted timestamp, except in the case of the
This function represents the probability of an item not fault being currently detected. This makes the serial
failing during the time interval [0, t]. When discussing the reliability mathematical model suitable to abstract the
reliability of a system, it is often referred to as the networked embedded systems machine. The reliability for
probability of no occurrence of faults belonging to class F serial networked embedded systems is given as:
(i.e., the system survives) during time t.
Rk(t) is the reliability of a single component k:
The probability that a system will continue to operate
without experiencing a fault of class F within a given time Rk(t) = e−λkt (7)
interval, t, is denoted by RF(t). This is also referred to as the
system's reliability. It is defined as the probability that the The overall system reliability Rser(t)
time to the first failure, tf, is greater than t given that the
system has operated successfully until time tinit. Rser(t) = R1(t) × R2(t) × R3(t) × . . . × Rn(t) n (8)
Mathematically, it can be expressed as:
Rser(t) = ∏ Ri(t) I=1 n (9)
RF (t) = P(tinit ≤ t < tf ∀ f ∈ F) (4)
The serial failure rate is given as λser = ∑ λi (10)

IJISRT23OCT1389 www.ijisrt.com 1235

Volume 8, Issue 10, October – 2023 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
Assuming that the failure rates of individual  Mobile Agent Migration Process
components are statistically independent, the overall The Server Agent (SV) consists of Servercontroller,
reliability of a system can be calculated using equations (8) Threads (TD), Resourcebundle (RB), Sqlconnection (SC),
and (9). Equation (8) states that the system reliability Rser(t) Messages (MSG), and Agents table (AT). SV communicates
at time t is equal to the product of the reliability of each with all the nodes on the network using LWA and monitors
individual component, denoted by R1(t), R2(t), R3(t),..., the agent communication between the Server Agent and
Rn(t), raised to the power of the number of components, n. Client Agents. It creates threads for each client agent probe,
Equation (3.9) provides a compact notation for this product monitors the agent thread, and collects responses in the form
using the product symbol, Π. of messages which carries all the information representing
the status of each client agent that migrated to each visited
The serial failure rate of the system, denoted by λser, node. It creates Agent tables AT to store the list of all the
can be obtained using equation (10), which states that λser is probed nodes on the network and sends all the listed nodes
equal to the sum of the failure rates of individual and their respective statuses to the database using
components, denoted by λ1, λ2, λ3,..., λn, summed over all n Servercontroller and Sqlconnection.
components.

Fig. 3 Server and Client Agent

IJISRT23OCT1389 www.ijisrt.com 1236

Volume 8, Issue 10, October – 2023 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
V. IMPLEMENTATION OF THE MODEL Up) while a value of 0 indicates 0% live components (node
down) on the LWA transmission paths.
The local area network (LAN) in the Federal
Polytechnic, Ede library was used for the implementation of As the reliability of this method heavily relies on the
the model. For testing of the framework on the LAN, proper functioning of both the LWA and the destination
twenty-day, five-hour-per-day test of the framework was nodes, it is assumed that these network components are
done on a network with twenty nodes, consisting of one always functioning correctly. In the active mobile agent
server node and nineteen client nodes. The operating technique that utilizes light weight agents for fault detection,
platform used for the test is Windows 8 OS, SQLite-3.12.2- the association between the agents and nodes must be taken
win64 for database, and Eclipse IDE for Java Developers into account. Researchers in fault detection have
4.23. Eclipse allows the integration of JADE through traditionally employed deterministic dependency
plugins and it allows agent platform integration. Agent information to model the network, assuming that the
creation, starting, lunching, activation, and killing can be connections between the nodes and probe agents are well
achieved on the platform. Since agents require multiple understood. This approach was adopted in prior studies by
nodes, the start node and destination node must be [52]. As illustrated in Figure 1.2, the nodes used to transmit
compatible [42] and [43]. LWA to destination node 1 are not deterministic when the
Server Agent Controller (Manager) sends them. The reason
 Performance Evaluation of the Model for this is that any route from the Server Agent Controller
The research is composed of the Agent Controller, a (Manager) to destination node 1 can be chosen as the
single Server Agent, and several network nodes called the transmission path.
Client Agents (CA) and status checker. The Server Agent
employs Light Weight Agents (LWA), which are special The table 1 below shows the failure frequency, Failure
data packets sent to destination nodes (CA) to detect rate, Reliability, and Mean Time between Failures of each of
network faults. Each LWA can be identified by its the nodes. This figure shows the reliability rate of each of
destination node address, which includes the IP address and the nodes after the test period of four hours each day for
MAC address. Based on the results obtained from the LWA, twenty days. It also shows the mean time between failures
the status of the nodes is determined and valued between 0 for each of the nodes.
and 1. A value of 1 indicates 100% live components (Node

Table 1 Failure Rate, Reliability and MTBF for each day

Failure Rate, Reliability and MTBF for each day (T = 5 hours)
Nodes Failure Frequency (f) Failure rate (λ) Reliability MTBF T/f (per
(R=e-λt) per hours) hours)
1 0 0 1 0
2 2 0.021 0.979 2.5
3 1 0.011 0.99 5
4 1 0.011 0.99 5
5 1 0.011 0.99 5
6 0 0 1 0
7 1 0.011 0.99 5
8 4 0.047 0.791 1.25
9 1 0.011 0.99 5
10 1 0.011 0.99 5
11 9 0.15 0.472 0.5
12 1 0.011 0.99 5
13 0 0 1 0
14 1 0 0.99 5
15 1 0.011 0.99 5
16 0 0 1 0
17 1 0.011 0.99 5
18 5 0.061 0.731 1
19 3 0.033 0.846 1.6
20 6 0.08 0.67 0.833

This table 2 shows the cumulative figures for all the test done on all the twenty nodes on the network for twenty days. It
shows the total percentage reliability of all the nodes on the network. It also shows the reliability of the framework after all the test
has been conducted for the twenty days.

IJISRT23OCT1389 www.ijisrt.com 1237

Volume 8, Issue 10, October – 2023 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
Table 2 Cumulating (C) Failure Rate, Reliability, and MTBF
Cumulating (C) Failure Rate, Reliability, and MTBF
C = C/20 C*100
Failure rate (λ) 0.491 0.02455 2.455
Reliability(R=e-λt) per hours) 18.389 0.9195 91.945
MTBF T/f (per hour) 57.683 2.8841

Discussion: ServerAgent and the clientAgents are containing twenty nodes on a network. The failure
connected to the network which contains twenty nodes. The frequency (f) of each node per day (twenty days) of the test
serverAgent is loaded on the single node while the rest of is also recorded as the corresponding nodes that failed
the nodes are loaded with the clientAgents. The system during the test period. The node2 as 2 failures, node3,
works on Client/Server architecture and each of the client node4, node5, node7, node9, node10, node12, node14,
nodes receives probes from the serverAgent which node15, and node17 respectively have failed only one time
consistently monitors all the LWA sent to each of the client within the twenty days test period. Node8 failed four times,
nodes. Figure 4 is a chart representation of data gotten from node11 failed nine times, node18 failed five times, node19
table 3.2 which shows the test period of twenty days failed three times and node20 failed 6 times respectively.

Fig 4 Number of Failures with Corresponding Nodes for Twenty Days Test

Discussion: The failure rate of a system is the shows the failure rate of each of the nodes on the network.
frequency at which the system fails or malfunctions over a The system was tested using twenty nodes for twenty days,
given time and it is usually expressed as the number of and the data for the failure rate for each node was calculated
failures per unit of time. The measurement depends on the from the failure frequency date in table 3.2. Figure 4.8
type of system and the data available. The data is usually shows the Failure rate of each of the nodes, starting from (λ)
obtained by monitoring the system over some time and = 0.011 for nodes that have the lowest failure rate to (λ) =
recording the number of failures that occur. Figure 4 above 0.15 for node(s) that have the highest failure rate.

IJISRT23OCT1389 www.ijisrt.com 1238

Volume 8, Issue 10, October – 2023 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165

Fig 5 Failure rate of each of the Nodes on a Network

Discussion: measuring the reliability of a system is reliability rate of the nodes as node1 = 100%, node2 = 0.979
important for ensuring that it performs its intended function (97%), node3 = 0.99 (99%), node4 = 0.99 (99%), node5 =
consistently and identifying potential problems before they 0.99 (99%), node6 =1, node7 =0.99(99%), node8 =0.791
occur. Figure 5 shows the reliability of each node in the (79%), node9 =0.99 (99%), node10 =0.99 (99%), node11
system. The data used for the chart is from table 1 which =0.472 (47%). node12 =0.99 (99%), node13 =1 (100%),
shows the reliability rate of each of the nodes in the node14 =0.99 (99%), node15 =0.99 (99%), node16 =1
network. Each of the nodes tested for the twenty days with (100%), node17 =0.99 (99%), node18 =0.731 (73%),
the system is shown and the corresponding calculated node19 =0.846 (84%), node20 =0.67 (67%). The node with
reliability rates are also shown. Nodes without any failure the highest reliability shows a reliability rate of 100% while
have a reliability rate of 1 which is 100% and also specify the lowest reliability rate as indicated above is 47%.
the probability that the node will not fail. The test shows the

Fig 6 Reliability of each node for Twenty Days test

Discussion: The system testing was done for five hours can show the total test hours, the total number of failed
every day and for twenty days, the cumulating results of the nodes, and the total working nodes. Therefore, figure 5 can
nodes from all the days are summed together and the show how reliable the system is haven is gone through the
average of the result is found. This is shown in figure 4.10 five hours daily and twenty days test period.
and shows the total failure rate, reliability, and MTBF. This

IJISRT23OCT1389 www.ijisrt.com 1239

Volume 8, Issue 10, October – 2023 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165

Fig 7 Cumulating (C) Failure Rate, Reliability, and MTBF

Discussion: Three agents (Agent Controller, Server REFERENCES

Agent, and Client Agents) were used with twenty 20 nodes
during the twenty days test of the framework. The test was [1]. Zhang, J., Song, J., Hu, X., & Li, M. (2022). A
done five hours every day for twenty days with the mobile agent-based fault detection approach for
assumption that all the twenty nodes are in good condition. wireless sensor networks. Ad Hoc Networks, 126,
Figure 6, shows the total number of nodes used for the first 103907.
test and it also shows the nodes that are alive and the ones [2]. Anand, V., & Devaraj, R. (2021). Mobile Agent
that have failed during the five hours test. It shows the based Fault Detection in Wireless Sensor Networks.
number of nodes that were alive throughout the test and each Journal of King Saud University-Computer and
of the nodes that are still connected and stored in the Information Sciences.
database. The failure rate (λ) of the system for this test was [3]. Tripathi, D. S., & Gupta, S. (2020). Fault detection in
computed and it shows the cumulated failure rate of (λ) = wireless sensor networks using mobile agents.
0.02455 and the reliability of R (t) = 92%. This shows that Wireless Personal Communications, 114(2), 737-756.
the framework is reliable having computed the reliability [4]. Singh, M., & Dave, M. (2020). Mobile agent-based
rate of the system. fault detection and recovery mechanism for wireless
sensor networks. Wireless Personal Communications,
VI. CONCLUSION 114(3), 1319-1341.
[5]. Chen, Y., He, S., Li, L., & Li, Q. (2019). Mobile
This research work presents mobile agents as a agent-based distributed fault detection in wireless
solution for network fault detection systems. This work sensor networks. Journal of Ambient Intelligence and
provides a system testing and a prototype implementation of Humanized Computing, 10(8), 3247-3261.
a proactive fault detection solution using mobile agents. [6]. Choudhary, P., & Singh, A. K. (2018). A review on
Correspondingly, the results from the test period of the fault- mobile agent based fault detection and recovery
detection system model showed that the theory of reliability mechanisms in wireless sensor networks. In 2018 3rd
can be used to verify that this research can determine that International Conference on Internet of Things:
the use of mobile agents is viable if properly deployed in an Smart Innovation and Usages (IoT-SIU) (pp. 1-5).
embedded network system. It is capable of providing a IEEE.
reliable, dependable, and always-available network [7]. Maria Zubair and Umar Manzoor (2016). Mobile
detection system to organizations, industries, banks, and Agent based Network Management Applications and
every other social sector that hinges on computer network in Fault-Tolerance Mechanisms. The Sixth International
delivering their services. Conference on Innovative Computing Technology
(INTECH 2016)

IJISRT23OCT1389 www.ijisrt.com 1240

Volume 8, Issue 10, October – 2023 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
[8]. Ademaj, A. (2002). Slightly-Off-Specification [24]. Kopetz, H. (2001). REAL-TIME SYSTEMS Design
Failures in the Time-Triggered Architecture. In Principles for Distributed Embedded Applications.
Proceedings of the Seventh IEEE International Kluwer Academic Publishers, 1-5.
Workshop on High-Level Design Validation and Test [25]. Kopetz, H., & Bauer, G. (2002). The time-triggered
(HLDVT'02), Cannes, France, 7-12. architecture. In Proceedings of the IEEE Special
[9]. Ademaj, A., Sivencrona, H., Bauer, G., & Torin, J. ISSUE on Modeling and Design of Embedded
(2003). Evaluation of Fault Handling of the Time- Software, 6.
Triggered Architecture with Bus and Star Topology. [26]. Kopetz, H., & Bauer, G. (2003). The time-triggered
In Proceedings of the 2003 IEEE International architecture. In Proceedings of the IEEE, 91(1), 112-
Conference on Dependable Systems and Networks 126.
(DSN’03), San Francisco, California, 123-132. [27]. Lamport, L., Shostak, R., & Pease, M. (1982). The
[10]. Adeosun, O. O. (2011). Development of an Enhanced Byzantine Generals Problem. ACM Transactions on
Model for Internet System Availability Using Programming Language Systems, 4(3), 382-401.
Modular Redundancy (Unpublished doctoral thesis). [28]. Manju, N., & Jayanthi, J. (2013). An Effective
Department of Computer Science and Engineering, Verification and Validation Strategy for Safety-
Obafemi Awolowo University, Ile-Ife, 27-28. Critical Embedded Systems. International Journal of
[11]. Barborak, M., Dahbura, A., & Malek, M. (1993). The Software Engineering & Applications (IJSEA), 4(2).
Consensus Problem in Fault-Tolerant Computing. [29]. McCarthy, M. (2003). Fault-Tolerant. Tech Target,
ACM Computing Surveys, 25(2), 171-220. 3(1), 13-21.
[12]. Chandra, T. D., & Toueg, S. (1991). Unreliable [30]. Minsky, Y., Renesse, R., Schneider, F. B., & Stoller,
Failure Detectors for Asynchronous Systems S. D. (1996). Cryptographic Support for Fault-
(Preliminary version). In 10th Annual ACM Tolerant Distributed Computing. Proceedings of the
Symposium on Principles of Distributed Computing, 7th ACM SIGOPS European Workshop, 109-114.
325-340. [31]. Olubosi, F. (2005). Recovering from Information
[13]. Curtis, H., & France, R. (1999). Time-Triggered System Failures. Vanguard Newspapers Online
Protocol (TTP/C): A Safety-Critical System Protocol. Edition, August 12.
EE382C Literature Survey, 10-24. [32]. Payal, B., & Mukesh, K. (2013). A Detailed
[14]. Dilger, E., Uhrer, T. F., Muller, B. M., & Poledna, S. Anatomization of Mobile Agents. International
(1998). The X-by-Wire Concept: Time-Triggered Journal of Science and Research (IJSR), 2(11), 1-7.
Information Exchange and Fail Silence Support by [33]. Pfeifer, H. (2000). Formal Verification of the TTP
New System Services. SAE Conference. Group Membership Algorithm. In IFIP TC6/WG6
[15]. Dobel, B., Hartig, H., & Engel, M. (2012). Operating International Conference on Formal Description
System Support for Redundant Multithreading. In Techniques for Distributed Systems and
Proceedings of the Tenth ACM International Communication Protocols and Protocol
Conference on Embedded Software, 83-92. Specification, Testing, and Verification,
[16]. Elmenreich, W., & Ipp, R. (2001). Introduction to FORTE/PSTV 2000 (pp. 3-18). Pisa, Italy.
TTP/C and TTP/A. Vienna University of [34]. Raman, C. V., & Atul, N. (2005). A Hybrid Method
Technology, InstitutfürTechnischeInformatik, to Intrusion Detection System Using HMM.
Vienna, Austria. Proceedings of ICDCIT Workshop, 4-9.
[17]. Fischer, M. J., Lynch, N. A., & Paterson, P. S. [35]. Ross. C., Lee P. A. and Anderson T. (2003):
(1985). Impossibility of distributed consensus with “Providing Fault-Tolerant Call-Control in the IMS
one faulty process. Journal of the ACM, 32(2), 374- using the Rserpool Architecture.” In Proceedings of
382. the 8th International Workshop on Mobile
[18]. George, A. R. (2008). Software modulated fault Multimedia Communication, Munch, Germany, pp.
tolerance. (Doctoral dissertation, Princeton 67-72.
University). [36]. Schlichting R.D and Schneider F.B. (1983): ‘‘Fail-
[19]. George, A. P., & Barbara, J. P. (2003). Automotive stop processors: an approach to designing fault-
vehicle safety. Technology and Engineer CRC Press, tolerant computing systems,’’ ACM Transactions on
9-10. Computer Systems, Vol.1, No.3, pp.222-238, ACM
[20]. Helmer, G., Wong, J. S. K., Honavar, V., Miller, L., Press.
& Wang, Y. (2003). Lightweight agents for intrusion [37]. Seidel F. (2009): “X-By-Wire, Operation Systems”,
detection. The Journal of Systems and Software, 67, Presented at the Seminar on Transportation Systems,
109-122. Chemnitz University of Technology.
[21]. Kelvin, H. (2009). Introduction to TTP- Time- [38]. Sunil G., Harsh K. V. and Sangal A. L. (2012):
Triggered protocol (Seminar paper). Chemnitz “Analysis and Removal of Vulnerabilities in
University of Technology, pages 2-5. Masquerading Attack in Wireless Sensor Networks”,
[22]. Knoll, G., Suri, N., & Bradshaw, J. M. (2002). Path- International Journal of Computer Technology and
based security for mobile agents. Notes in Electronics Engineering (IJCTEE), Vol.2, Issue 3.
Theoretical Computer Science, 58(2), 16. [39]. Verissimo P. and Rodrigues L. (2001): “Distributed
[23]. Kopetz, H. (2001). A comparison of TTP/C and Systems for System Architects”, Kluwer Academic
FlexRay. Technische UniversitätWie, 5. Publishers, pp.1.

IJISRT23OCT1389 www.ijisrt.com 1241

Volume 8, Issue 10, October – 2023 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
[40]. Wensley J. H., Lamport L., Goldberg J., Green
M.W., Levitt K.N., Melliar-Smith P. M., Shostak
R.E., and Weinstock C. B. (1978): ‘‘SIFT: Design
and analysis of a fault-tolerant computer for aircraft
control,’’ Proceedings of the IEEE, Vol.66, No.10,
pp.1240-1255.
[41]. White T. and Pagurek B. (1998): “Towards Multi-
Agent Problem Solving in Networks”, Proceedings of
the Third International Conference on Multi-Agent
Systems (ICMAS ’98), pp.333-340.
[42]. Bathula, B. G., Sinha, R. K., Chiu, A. L., and
Woodward, S. L. (2018). Routing and regenerator
planning in a carrier’s core reconfigurable optical
network. Journal of Optical Communications and
Networking, 10(2), A196–A205.
[43]. Jhawar, R. and Piuri, V. (2017). Fault tolerance and
resilience in cloud computing environments. In
Computer and Information Security Handbook (Third
Edition), Elsevier, pp. 165–181.
[44]. Wang, Z. (1989). Model of network faults. Integrated
Network Management, Amsterdam: North-Holland.
[45]. Adetokunbo M and Ojieabu Clement E (2017).
Automated Fault Detection and Identification System
for Computer Networks.
[46]. Castro, M., & Liskov, B. (1999). Practical Byzantine
fault tolerance. In Proceedings of the Third
Symposium on Operating Systems Design and
Implementation (OSDI), pp. 173-186.
[47]. Ongaro, D., & Ousterhout, J. (2014). In search of an
understandable consensus algorithm. Proceedings of
the 2014 USENIX Annual Technical Conference, pp.
305-319.
[48]. Vukolić, M. (2015). The quest for scalable
blockchain fabric: Proof-of-work vs. BFT replication.
IEEE Communications Magazine, 53(9), 71-77.
[49]. Hopcroft, J. E., & Karp, R. M. (1973). An n5/2
algorithm for maximum matching in bipartite graphs.
SIAM Journal on Computing, 2(4), 225-231.
[50]. Liu, Y., & Li, B. (2011). A survey of network fault
diagnosis based on graph theory. Mathematical
Problems in Engineering, 2011, 1-19.
[51]. Coates, M., & Marbach, P. (2002). Network
tomography: Recent developments. Statistical
Science, 17(4), 465-480.
[52]. Paul Barford, Yan Chen, Anup Goyal, Zhichun Li,
Vern Paxson and Vinod Yegneswaran. “Employing
Honeynets For Network Situational Awareness”, In
Series: Advances in Information Security, Springer,
2009.
[53]. A. Dusia | A. S. Sethi (2018). Probe Generation for
Active Probing. Article in International Journal of
Network Management, Wiley, 2018.

IJISRT23OCT1389 www.ijisrt.com 1242