Deep Learning-based Anomaly Detection in Network Traffic Patterns
Date
2024
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
university of bordj bou arreridj
Abstract
The anomaly in network traffic is a crucial issue that can cause significant losses
in network security and performance. This prompted us to undertake this work to detect
these anomalies accurately and promptly using deep learning techniques.
This thesis investigates the use of long short-term memory (LSTM) neural
networks, one of the deep learning methods, to detect anomalies in network data flows.
LSTMs are well suited to this task thanks to their ability to capture long-term temporal
dependencies. Our approach is distinguished by its ability to detect complex and varied
anomalies, thus improving the security and efficiency of computer networks. The results
show a significant improvement over traditional methods
Description
In conclusion, this thesis delves into the critical realm of anomaly detection
in network traffic, recognizing its pivotal role in safeguarding the integrity and
security of digital communication networks. As traditional methods falter in
coping with the escalating complexity and volume of network data, the
exploration of advanced techniques, particularly leveraging deep learning,
emerges as imperative. Through the lens of Long Short-Term Memory (LSTM)
networks, this research endeavors to overcome the limitations of conventional
approaches and enhance the accuracy and timeliness of anomaly detection.
By delineating the challenges, elucidating the fundamentals of deep
learning, describing the dataset, and presenting implementation details and
experimental findings, this thesis strives to contribute to the evolving landscape
of network security.
Looking forward, future efforts should focus on developing AI-driven
anomaly detection systems that not only identify anomalies but also provide
actionable insights and explanations to cybersecurity analysts. Moreover,
expanding hybrid methods that combine LSTM with complementary models
holds promise for achieving even greater performance gains. Ultimately, the
culmination of this endeavor aims to furnish insights, methodologies, and
advancements that fortify the resilience of network infrastructures against
evolving cyber threats
Keywords
Anomaly detection, Network traffic, Network security, Network performance, Deep learning, Long short-term memory (LSTM), Neural networks, Temporal dependencies, Complex anomalies, Computer networks