Software Requirements

Specification (SRS)
for

Analysis of DDoS Attack

Using
Machine Learning Algorithm

Prepared by

S.No Name of the Student Roll Number, Specialization

1 Rasani Chukka Subramanyam 22N31A67E2, CSE(DS)
2 Sharvirala Sai Advika 22N31A67F6, CSE(DS)
3 Penkey Deekshitha Sai Priya 22N31A67D5, CSE(DS)

Supervisor : Mrs.N.Chandana

Designation: Assistant Professor

Department: Dept. Of Emerging Technologies

Batch ID: DS/2025/Application-Development-2/02

Date:
Supervisor Sign.&
Date
Department of CSE (Data Science)
Software Requirements Page
Specification ii
Title of the Project Analysis of DDoS Attack using Machine Learning Algorithm.

Contents
CONTENTS........................................................................................................................................... III

REVISIONS........................................................................................................................................... III

1 INTRODUCTION........................................................................................................................... 1
1.1 DOCUMENT PURPOSE................................................................................................ 1
1.2 PROJECT SCOPE....................................................................................................... 1
1.3 EXISTING SYSTEMS................................................................................................... 2
1.4 PROBLEMS WITH EXISTING SYSTEMS............................................................................2
1.5 PROPOSED SYSTEMS................................................................................................. 3
1.6 ADVANTAGES OF PROPOSED SYSTEMS.........................................................................4
2 OVERALL DESCRIPTION............................................................................................................ 5
2.1 FEASIBILITY STUDY................................................................................................... 5
2.2 PRODUCT FUNCTIONALITY.......................................................................................... 6
2.3 DESIGN AND IMPLEMENTATION CONSTRAINTS................................................................6
2.4 ASSUMPTIONS AND DEPENDENCIES..............................................................................7
3 FUNCTIONAL REQUIREMENTS.................................................................................................. 9
3.1 SOFTWARE REQUIREMENT SPECIFICATIONS....................................................................9
3.2 HARDWARE REQUIREMENTS SPECIFICATIONS..................................................................9
3.3 USE CASE MODEL.................................................................................................... 9
4 OTHER NON-FUNCTIONAL REQUIREMENTS..........................................................................12
4.1 PERFORMANCE REQUIREMENTS.................................................................................12
4.2 SAFETY AND SECURITY REQUIREMENTS…………………..………………………………………13
4.3 SOFTWARE QUALITY ATTRIBUTES……………………….………………………………………...14
5 REFERENCES............................................................................................................................ 15

Revisions
Version Primary Author(s) Description of Version Date Completed
1.0 Rasani Chukka Primary Revision giving an overall view 02/03/25
Subramanyam of the project and document.

Department of CSE (DS | CyS | IoT)| MRCET (A) Ca

mPpagues7
Software Requirements Specification Page 1

1 Introduction
This study explores the application of Machine Learning (ML) techniques, specifically
leveraging the Pycaret module, for analyzing Distributed Denial of Service (DDoS)
attacks. The focus is on predicting efficient methods for timely detection, mitigation, and
prevention of such malicious activities by scrutinizing incoming network packets. ML
algorithms, particularly those implemented through Pycaret, are employed for
classification based on influential data flow features. Compared to traditional ML
algorithms, such as XGBoost, Random Forest, Extra Trees, Decision Tree, Gradient
Boosting, and others, results show that Pycaret, especially with the XGBoost model,
exhibits heightened efficiency in DDoS attack detection. This approach aims to ensure
the availability, integrity, and confidentiality of cloud-based services, minimizing
financial and reputational risks associated with DDoS incidents, contributing
significantly to cybersecurity measures in the dynamic digital landscape.
1.1 Document Purpose
This Software Requirements Specification (SRS) document delineates the
requirements for
“Analysis of DDoS Attack using Machine Learning Algorithm”.This project focuses on
leveraging Machine Learning (ML) techniques, particularly through the PyCaret module,
to enhance defenses against Denial of Service (DoS) attacks. By analyzing network
traffic patterns, the system aims to detect, mitigate, and prevent potential DoS
incidents. It explores a range of ML algorithms such as XGBoost, Random Forest, and
Support Vector Machine to identify anomalous patterns indicative of attacks. The
proposed methodology employs a three-tier ML classification architecture combined
with PyCaret’s automation capabilities for model selection and evaluation. Results
indicate XGBoost as the top-performing model, showcasing exceptional accuracy and
effectiveness in DDoS attack detection. The project shows robustness and reliability in
differentiating between normal and attack traffic through extensive analysis and
visualization techniques like confusion matrices and precision-recall curves, adding to
the resilience of cloud-based services against cyber threats.
1.2 Project/Product Scope
The ubiquity of Distributed Denial of Service (DDoS) attacks in the age of digital
connectivity presents serious obstacles to cloud-based services’ availability and
integrity. These malevolent intrusions interfere with network operations, resulting in
monetary losses, harm to one’s reputation, and deteriorated user experiences. This
project’s main goal is to predict a complete system that uses machine learning (ML)
techniques to detect, mitigate, and prevent DDoS attacks against cloud infrastructures
in order to address this urgent problem.The methodologies, findings, and insights
obtained throughout the project will be comprehensively documented. Department of
Computer Science and Engineering 2 Research outcomes will be disseminated through
academic publications, conference presentations, and industry forums to contribute to
the broader cybersecurity community. By tackling these issues, the project hopes to
improve DDoS detection and mitigation techniques and increase cloud-based services’
resistance to online attacks. The interdisciplinary nature of the project, integrating
principles from cybersecurity, machine learning, and network engineering, will facilitate
a holistic approach towards combating DDoS attacks and safeguarding critical digital
infrastructure.

Department of CSE (DS | CyS | IoT)| MRCET (A) Campus

Software Requirements Specification Page 2

1.3 Existing System

Current framework for combating Distributed Denial of Service (DDoS) attacks
combines traditional security measures with machine learning (ML) techniques. DDoS
attacks overwhelm networks, disrupting access for legitimate users, making robust
defenses crucial as cyber threats evolve.
 Traditional Measures:
 Firewalls: Filter traffic based on predefined rules but struggle with
complex DDoS attacks.
 Intrusion Detection System (IDS): Monitor traffic for suspicious patterns.
Signature-based IDS detects known threats, while anomaly-based IDS
identifies deviations from normal behavior. Both may produce false
positives or miss new attack methods.
 The existing system may rely primarily on traditional DDoS mitigation techniques,
such as rate limiting, traffic filtering, or relying on the cloud service provider's built-
in security measures.
 Traditional methods may lack the adaptability to effectively counter new and
evolving DDoS attack tactics, potentially leading to increased vulnerability.
 While traditional defenses provide basic protection, ML offers advanced detection
and response capabilities. Ongoing research in ML algorithms, feature selection,
and dataset collection is essential to improve DDoS detection systems against
evolving threats.
1.4 Problems with Existing System
 The current system for identifying and countering Distributed Denial of Service
(DDoS) attacks has a number of shortcomings and difficulties despite utilizing
both machine learning (ML) and conventional security measures:
 A large number of DDoS detection systems that use machine learning currently in
use rely on static features that are taken from network traffic, like payload
attributes or packet headers. Nevertheless, in order to avoid detection, attackers
can simply obfuscate or modify these features, making the defenses useless.
 Some systems may lack adaptability to evolving DDoS attack techniques and
patterns. ML models trained on historical data may fail to generalize well to new
and emerging threats, leading to decreased detection accuracy and increased
false positive rates.
 ML-based DDoS detection systems may encounter scalability challenges,
particularly in large-scale cloud environments with high volumes of network traffic.
Large-scale real-time data processing and analysis can put a burden on
processing power and add latency, which affects the system’s overall
performance.
These problems highlight the need for a more innovative and user-centric
approach to “Analysis of DDoS Attack using Machine Learning Algorithm" aims to
provide. By addressing these challenges, the proposed system will enhance the
overall user experience and streamline the analysis process.

Department of CSE (DS | CyS | IoT)| MRCET (A) Campus

Software Requirements Specification Page 3
1.5 Proposed System
• The proposed system utilizes Machine Learning algorithms to analyze
network traffic patterns and identify DDoS attacks based on learned
behavior.
• Machine Learning models continuously learn from new data, adapting to
changing attack tactics, and improving detection accuracy over time.
• The primary objective is to enhance the accuracy of DDoS attack detection.
Achieve a high true positive rate while minimizing false positives.
• This approach can help prevent the damage and disruption caused by DDoS
attacks in large business firms and banking sectors, where service
availability is critical.
• Quick and accurate detection of DDoS attacks enables faster response and
minimizing downtime and service interruptions.
• Machine learning models can adapt to new and evolving DDoS attack
techniques, staying effective in the face of changing threats.
• This approach aims to ensure the availability, integrity, and confidentiality of
cloud-based services, minimizing financial and reputational risks associated with
DDoS incidents, contributing significantly to cybersecurity measures in the
dynamic digital landscape.

Department of CSE (DS | CyS | IoT)| MRCET (A) Campus

Software Requirements Specification Page 4

1.6 Advantages of Proposed Systems

The proposed system for detecting and mitigating DDoS attacks using machine
learning (ML) algorithms offers several advantages over traditional methods. These
benefits make it a robust, efficient, and scalable solution for combating evolving cyber
threats. Below are the key advantages:
 High Detection Accuracy: ML algorithms, such as Random Forest, XGBoost,
and Neural Networks, achieve high accuracy rates (often above 95%) in identifying
DDoS attacks. Reduces false positives and ensures reliable detection of malicious
traffic.
 Adaptability to Evolving Threats: ML models continuously learn from new data,
enabling them to adapt to new attack vectors and evolving DDoS techniques.
Ensures the system remains effective against zero-day attacks and sophisticated
threats.
 Real Time Detection and Response: The system can analyze network traffic
in real-time, enabling quick detection and immediate mitigation of DDoS attacks.
 Comprehensive Threat Detection: ML algorithms analyze multiple dimensions of
network traffic, including packet size, frequency, source IP addresses, and protocol
types. Provides a holistic view of potential threats, improving detection capabilities.
 Reduced False Positives: Advanced ML techniques, such as ensemble methods
and feature selection, improve detection accuracy and reduce false positives.
Prevents unnecessary disruptions to legitimate traffic.
 Interactive Visualizations: The use of interactive graphs, timelines, and visual
itineraries helps users better understand threat patterns , making it easier to visualize
mitigation strategies.
 Cost-Effective Solutions: By automating detection process, the system reduces the
need for extensive human resources and minimizes operational costs. Offers a
cost-effective solution for organizations of all sizes.
 Data Driven Insights: ML models generate actionable insights from network traffic
data, helping organisations understand attack patterns and improve their security
posture. Enables continuous improvement and informed decision-making.
 Enhanced Resilience: Ensemble methods, such as AdaBoost and XGBoost,
combine multiple ML models to create a robust system that is resilient to adversarial
attacks and data noise.
 Support for Diverse Environments: The system is adaptable to various network
environments, including IoT, cloud, and enterprise networks.
 User-Friendly Dashboards: The system includes intuitive dashboards and
visualizations to help network administrators monitor traffic, analyze threats, and take
informed actions. Improves usability and accessibility for security teams.
 Accessibility Features: The application will include accessibility options to cater to
users with disabilities, ensuring that everyone can benefit from the trip planning tools
provided.
 Sustainability Options: By preventing network downtime and resource wastage
caused by DDoS attacks, the system contributes to the sustainable operation of
onlineservices.

Department of CSE (DS | CyS | IoT)| MRCET (A) Campus

Software Requirements Specification Page 5

2 Overall Description
2.1 Feasibility Study
A comprehensive feasibility study has been conducted to assess the viability and
practicality of implementing a machine learning (ML)-based system for detecting Distributed
Denial of Service (DDoS) attacks. The study evaluates the project across four critical
dimensions: technical feasibility, economic feasibility, operational feasibility,
and scheduling feasibility.
Technical Feasibility:
 The proposed system leverages well-established ML algorithms such
as Random Forest, XGBoost, Support Vector Machines (SVM), and Neural
Networks, which are proven to be effective in detecting DDoS attacks with high
accuracy.
Economic Feasibility:
 The development costs are considered reasonable, The use of open-source ML
libraries and frameworks significantly reduces development costs. Additionally,
cloud platforms offer pay-as-you-go models, minimizing upfront investment.
Operational Feasibility:
 The system is designed to operate autonomously, requiring minimal human
intervention. Security teams can easily monitor and manage the system through
intuitive dashboards and alerts.

Scheduling Feasibility:
 A realistic timeline has been established, including phases for data
collection, model training, testing, and deployment. The use of pre-trained
models and open-source tools accelerates development. Sufficient time has
been allocated for rigorous testing and validation to ensure the system meets
accuracy and performance benchmarks. The system can be deployed
incrementally, starting with pilot testing in a controlled environment before full-
scale implementation.
Conclusion:
The feasibility study confirms the viability of the proposed ML-based DDoS detection
system. It is technically robust, leveraging proven algorithms and scalable infrastructure
for real-time detection. Economically, it is cost-effective with potential for revenue
generation. Operationally, it integrates seamlessly with existing systems and requires
minimal intervention. With a realistic timeline and adequate resources, the system offers a
scalable and efficient solution to combat DDoS attacks, ensuring the security and
availability of critical online services.

Department of CSE (DS | CyS | IoT)| MRCET (A) Campus

Software Requirements Specification Page 6

2.2 Product Functionality

The product designed for analyzing DDoS (Distributed Denial of Service) attacks using
machine learning algorithms typically has the following key functionalities:
 Data Collection & Preprocessing: Monitor network traffic and extract key
features (IP, packet size, protocol).Clean and normalize data for model training.
 Machine Learning Model Training: Use supervised and unsupervised learning
to train models (e.g., Decision Trees, SVM, Neural Networks). Optimize and tune
models for better performance.
 Visualization & Reporting: Provide dashboards with attack data visualizations.
Generate detailed reports on attack trends and model performance.
 Real-Time Detection: Detect DDoS attacks in real-time by analyzing traffic
patterns. Classify traffic as normal or attack and generate alerts.
 Interactive User Interface: Offer role-based access control, real-time alerts,
and custom notifications.
2.3 Design and Implementation Constraints
The development of " DDoS Attack Analysis using Machine Learning Algorithm" is
subject to several design and implementation constraints that will influence the
development process and final product:

 Data Quality and Availability:

 Labeled Data: Machine learning models require labeled data (normal vs. attack
traffic) for training. High-quality, comprehensive datasets may be hard to acquire,
especially for rare or emerging DDoS attacks.
 Imbalanced Data: DDoS attack datasets are often imbalanced (normal traffic is
more frequent than attack traffic), which can lead to biased models if not
addressed properly.
 Real-Time Processing Requirements:
 Low Latency: The system must analyze network traffic in real-time, requiring low-
latency processing. This can be challenging when dealing with large volumes of
data, as high computational power and fast algorithms are needed.
 Scalability: The system must scale to handle high network traffic volumes,
especially in large-scale environments or during massive DDoS attacks.
 Model Complexity:
 Algorithm Selection: The choice of machine learning algorithms impacts both the
accuracy and performance. Complex models (e.g., deep learning) may offer better
detection but require more computation and data.
 Overfitting: There is a risk of overfitting the model to specific attack patterns,
which can reduce its ability to generalize to new, unseen attacks.

Department of CSE (DS | CyS | IoT)| MRCET (A) Campus

 Software Requirements Specification Page 7

 Computational Resources:
 High Resource Usage: Training machine learning models on large
datasets or running real-time analysis may require significant computational
resources (CPU, GPU, memory), especially for advanced algorithms like
deep learning.
 Model Maintenance and Adaptation:
 Continuous Training: DDoS attack methods evolve rapidly. The system
needs continuous retraining with new data to adapt to emerging attack
patterns, which can be resource-intensive.
 Model Drift: The performance of machine learning models may degrade
over time as attack patterns change, requiring frequent updates to models.
 Integration with Existing Infrastructure:
 Compatibility: The system must be compatible with existing network
infrastructure, including firewalls, IDS/IPS, and other security tools.
Integration challenges may arise if these systems use different data
formats or protocols.
 API Integration: Seamless integration with third-party tools or SIEM
platforms requires well-designed APIs and may encounter issues related to
data synchronization or communication delays.
 Legal and Compliance Constraints:
 Regulatory Compliance: The system must adhere to regulatory
frameworks that govern data handling, security measures, and incident
reporting, which may vary by region or industry.
 Liability Concerns: Misclassifying traffic or failing to detect an attack could
result in legal consequences or a breach of service level agreements
(SLAs).

2.4 Assumptions and Dependencies

The successful analysis and execution of the "Analysis of DDoS Attack using Machine
Learning Algorithm" depend on several key assumptions and dependencies that must
be considered throughout the project analysis phase:
1. Availability of Network Traffic Data:
It is assumed that there is access to large volumes of labeled network traffic data
(both normal and attack traffic) for training and validating the machine learning
models. The system assumes that data can be captured from network interfaces
or through traffic monitoring tools.
2. DDoS Attack Types are Known:
The model assumes that the types of DDoS attacks are predefined and well-
understood (e.g., SYN flood, UDP flood, DNS amplification). The machine
learning models will focus on detecting these known attack types, although the
system may also adapt to novel attack patterns over time.

Department of CSE (DS | CyS | IoT)| MRCET (A) Campus

Software Requirements Specification Page 8
3. Access to Sufficient Computational Resources:
It is assumed that the system has access to adequate computational resources
(e.g., servers, cloud infrastructure) for processing high volumes of network data
and running machine learning models in a timely manner.
4. Model Retraining is Regularly Performed:
It is assumed that the system will support regular retraining and updates to
machine learning models to adapt to evolving attack patterns, ensuring long-term
effectiveness.
5. Machine Learning Libraries and Frameworks:
The system depends on popular machine learning frameworks and libraries (e.g.,
TensorFlow, Scikit-learn, Keras, XGBoost) for developing and training models.
The functionality of these libraries and their integration into the system is crucial to
the development process.
6. Cloud or On-Premises Infrastructure:
The system may depend on cloud computing resources (AWS, Google Cloud,
Microsoft Azure) or on-premises hardware for data processing, model training,
and storage. The availability and scalability of these resources are crucial for
handling large traffic volumes and running intensive machine learning algorithms.
7. Network Hardware and Software Integration:
The system relies on seamless integration with existing network hardware and
security infrastructure, such as firewalls, routers, and intrusion prevention systems
(IPS). Proper API integration and communication between the machine learning
model and these devices are necessary for implementing defensive measures.
8. API Rate Limiting and Data Access:
The system's ability to pull external data (such as traffic information from cloud
providers or threat intelligence services) may be dependent on rate limits, data
access policies, and subscription plans for third-party APIs. Any constraints here
could limit the system's real-time effectiveness.

Department of CSE (DS | CyS | IoT)| MRCET (A) Campus

Software Requirements Specification Page 9

3 Functional Requirements
3.1 Software Requirement Specifications
 Programming Language: Python
 Machine Learning Libraries: PyCaret, scikit-learn
 Data Processing Libraries: Pandas, NumPy
 Visualization Libraries: Matplotlib, Seaborn

3.2 Hardware Requirements Specifications

 Processor: Multi-core processor with sufficient computing power for model
training and evaluation.
 RAM: 8GB or more
 Storage: 500 GB
 Network: Stable internet connection for accessing real-time data.

3.3 Use Case Model

Department of CSE (DS | CyS | IoT)| MRCET (A) Campus

 Software Requirements Specification Page
10

3.3.1 Use Case

Purpose -The Use Case Diagram provides a high-level overview of the primary
functionalities that different actors (Security Analyst, Data Scientist, System Admin)
can interact with in the context of analyzing DDoS attacks using machine learning
algorithms. The focus is solely on the analysis phase, including data collection,
model training, attack detection, and reporting, without involving mitigation
strategies.
Requirements Traceability –
 R1: Data Collection and Preprocessing
 R2: Machine Learning Model Training
 R3: DDoS Attack Detection
 R4: Reporting and Visualization
Priority - High.
Preconditions - The system for DDoS attack analysis is operational and accessible to
the respective actors.
Postconditions- The actors can effectively perform their designated tasks, such as
detecting DDoS attacks and generating analysis reports.
Actors – Security Analyst, Data Scientist, System Admin

Flow of Events
1. Data Scientist collects and preprocesses network data, then trains
machine learning models specifically designed for DDoS attack
detection.
2. Security Analyst uses the system to monitor and detect DDoS
attacks, reviews detailed analysis reports, and visualizes the findings.
3. System Admin ensures the system is properly configured and
maintained for accurate data collection and analysis.

Department of CSE (DS | CyS | IoT)| MRCET (A) Campus

Software Requirements Specification Page
11

3.3.2 Data Flow Diagram

Department of CSE (DS | CyS | IoT)| MRCET (A) Campus

 Software Requirements Specification Page
12

4 Other Non-functional Requirements

4.1 Performance Requirements

Performance requirements are essential to ensure that "VacAIgent" delivers a smooth

and responsive user experience. The following performance criteria have been
established:
 Real-Time Detection:
Requirement: The system must be capable of detecting DDoS attacks in real-
time, with minimal latency between traffic data collection and detection
Performance Metric: The system should have a maximum processing latency
of < 1 second per packet or request for real-time attack detection..
 Accuracy:
Requirement: The machine learning model should achieve a high detection
accuracy, minimizing false positives (incorrectly identifying legitimate traffic as
an attack) and false negatives (failing to identify actual attacks).
Performance Metric: The model should have a true positive rate (TPR) of >
95% and a false positive rate (FPR) of < 5% for known DDoS attack patterns.
 Scalability:
Requirement: The system must be scalable to handle high volumes of network
traffic, especially in environments with large-scale or distributed denial-of-service
attacks.
Performance Metric: The system should be capable of processing ≥ 10 Gbps
of network traffic in a production environment.
 Resource Utilization:
Requirement: The system should be efficient in terms of resource utilization,
especially for computational resources like CPU, memory, and storage, while
maintaining performance.
Performance Metric: The system should maintain < 75% CPU utilization and <
80% memory utilization during peak traffic analysis periods
 Response Time:
Requirement: Once an attack is detected, the system must trigger defensive
actions (e.g., IP blocking, rate limiting) immediately to mitigate the attack.
Performance Metric: The defense response should be triggered within < 2
seconds of detection.

Department of CSE (DS | CyS | IoT)| MRCET (A) Campus

Software Requirements Specification Page
13

4.2 Safety and Security Requirements

 Data Protection and Encryption:

All network traffic data and sensitive information should be encrypted during
storage and transmission to protect it from unauthorized access.
 Access Control:
The system must restrict access to sensitive functions, configuration settings,
and data to authorized users only.
 Malicious Traffic Handling:
The system should be able to handle and analyze malicious traffic, including
DDoS attack data, without causing system instability or failure.
 Error Handling and Recovery:
The system must have error detection capabilities to prevent the analysis from
failing or producing incorrect results due to malformed or corrupted network
data.
 Intrusion Detection:
The system must include basic security features to detect unauthorized access
attempts or malicious data manipulation during the analysis process.
 Model Security:
The machine learning models used for the analysis should be secure from
tampering and manipulation to ensure the integrity of the detection process.
 Logging and Auditing:
The system should log all interactions and analyses for audit purposes, with
secure and tamper-resistant logs that can be used for forensic purposes if
needed.
 Real-Time Data Analysis:
The system should be capable of processing data in real-time, enabling it to
detect and analyze DDoS attacks as they occur, without causing delays or
backlogs in the system's operations.

Department of CSE (DS | CyS | IoT)| MRCET (A) Campus

Software Requirements Specification Page
14

4.3 Software Quality Attributes

Software quality attributes are essential for ensuring that "VacAIgent" meets user
expectations and performs reliably. The following quality attributes will guide the
development process:

4.3.1 Usability:
The system's interface, if applicable, should be intuitive for security analysts to easily
interpret analysis results, configure parameters, and monitor the analysis of DDoS
attacks..
4.3.2 Maintainability:
The system should be structured in a way that allows easy updates to the machine
learning models and data sources used for DDoS attack analysis, ensuring
continued performance over time.
4.3.3 Adaptability:
The system should ensure continuous operation, providing real-time analysis
capabilities at all times, even during high traffic periods, without significant downtime
or loss of functionality.
4.3.4 Reliability:
The system must provide consistent and accurate results in detecting DDoS attacks.
It should ensure the integrity of the analysis process without frequent failures or
inaccuracies.
4.3.5 Performance:
The system must be capable of efficiently processing large volumes of network
traffic in real-time to accurately identify DDoS attack patterns without introducing
significant delays in analysis.
4.3.6 Security:
The system should ensure that sensitive network traffic data is securely analyzed
and protected from unauthorized access or tampering during the analysis phase.
4.3.7 Interoperability:
The system must be able to integrate and interact effectively with existing network
monitoring systems, security tools, and data sources to perform comprehensive
DDoS attack analysis.
4.3.8 Accessibility:
The system should ensure continuous operation, providing real-time analysis
capabilities at all times, even during high traffic periods, without significant downtime
or loss of functionality.

Department of CSE (DS | CyS | IoT)| MRCET (A) Campus

Software Requirements Specification Page
15

5 References
 Abdul Raoof Wani, Q.P. Rana, U. Saxena, Nitin Pandey. “Analysis and
Detection of DDoS Attacks on Cloud Computing Environment using Machine
Learning Techniques.” IEEE 2019

 Swathi Sambangi and Lakshmeeswari Gondi. “A Machine Learning Approach for

DDoS(Distributed Denial of Service) Attack DetectionUsing Multiple Linear
Regression.” 25 December 2020.

 Aman Rangapur, Tarun Kanakam and Ajith Jubilson. “DDOSDET: AN

APPROACH TO DETECT DDOS ATTACKS USING NEURAL NETWORKS.”
January 25, 2022.

 Myneni, S.; Chowdhary, A.; Huang, D.; Alshamrani, A. SmartDefense: A

distributed deep defense against DDoS attacks with edge computing. Comput.
Netw. 2022.

 Taherian-Fard, E.; Niknam, T.; Sahebi, R.; Javidsharifi, M.; KavousiFard, A.;
Aghaei, J. A Software Defined Networking Architecture for DDoS-Attack in the
Storage of Multimicrogrids. IEEE Access 2022, 10, 83802–83812

 Kaspersky Lab Team. DDoS Intelligence Report: Long-lasting Attacks,

Amplification Attacks and Old Botnets Make a Comeback Kaspersky,
Kaspersky Lab.2023.

Department of CSE (DS | CyS | IoT)| MRCET (A) Campus

 Software Requirements Specification Page
16

SRS DOCUMENT REVIEW

CERTIFICATION
This Software Requirement Specification (SRS) Document is reviewed and certified
to proceed for the project development by the Departmental Review Committee
(DRC).

Date of SRS Submitted:

Date of Review :

Supervisor Comments:

Supervisor Sign. & Date.

Coordinator Sign. & Date

HoD Sign. & Date

Dept. Stamp

Department of CSE (DS | CyS | IoT)| MRCET (A) Campus