International Journal of Computer Applications |
Foundation of Computer Science (FCS), NY, USA |
Volume 187 - Number 19 |
Year of Publication: 2025 |
Authors: Neethu B., Sheena Mathew |
![]() |
Neethu B., Sheena Mathew . Temporal-Spatial Deep Learning Framework for Real-Time Intrusion Detection in Cloud Environments. International Journal of Computer Applications. 187, 19 ( Jul 2025), 38-42. DOI=10.5120/ijca2025925288
With the rapid advancement of cloud computing, security breaches and intrusion attempts have become increasingly sophisticated and real-time. Traditional intrusion detection systems often fall short in identifying and detecting the evolving threats within dynamic cloud environments because they lack adaptability and struggle with effective feature representation. To address this, this paper proposes a new Temporal-Spatial Deep Learning (TSDL) framework that combines Convolutional Neural Networks (CNNs) for capturing spatial features with Long Short-Term Memory (LSTM) networks to learn temporal patterns in cloud network traffic. The proposed model pre-processes sequential packet data while keeping track of data flow, which enables early and accurate intrusion detection of the system. The system is evaluated on benchmark datasets such as CICIDS2017 and UNSW-NB15, and it outperforms traditional machine learning models and regular deep learning networks in both detection accuracy and processing latency. This system is designed to operate in real-time, making it suitable for deployment in large-scale cloud infrastructures.