International Journal of Innovative Research in Engineering & Multidisciplinary Physical Sciences
E-ISSN: 2349-7300Impact Factor - 9.907

A Widely Indexed Open Access Peer Reviewed Online Scholarly International Journal

Call for Paper Volume 12 Issue 6 November-December 2024 Submit your research for publication

A GAPM-DLNN-based Attack Detection for Distributed Packet Switching in Local Computer Networks

Authors: Amaresan Venkatesan

DOI: https://doi.org/10.5281/zenodo.14183832

Short DOI: https://doi.org/g8rc48

Country: USA

Full-text Research PDF File:   View   |   Download


Abstract: In local computer networks, the packet transmission mechanism provided by Ethernet has been used. Switched packets are distributed in ether and transmitted to the destination in that transmission. None of the existing works weren’t focused on the buffer overflow attack type detection with mitigation. Here, Global Attention with Parametric Mish-based Deep Learning Neural Network (GAPM-DLNN)-based attack detection in packet switching is proposed in this paper. Primarily, the index is constructed using Indexed Divide and Conquer (IDC) if the packet transmission request is accepted by the main server. Next, paths are generated. The optimal path is selected using Guided Kookaburra Optimization (GKO) from the generated paths. The load is estimated and sent to the main server using the Transmission Control Protocol (TCP). Moreover, the packets are reordered in TCP. Then, by using the GAPM-DLNN approach, the attack is detected in the obtained packet. Next, by utilizing the Quantum Fuzzy Inference System (QFIS) approach, the attack is mitigated. The map-reduce function is used if the attack is stack-based overflow. If the attack is a heap overflow, then the attacked data is compared with the index table to reduce its size. If the attack is a format string, then the attacked data is blocked and sends an acknowledgment to the destination. The proposed method achieves a 97.5% Packet Delivery Ratio(PDR) as per experimental analysis.

Keywords: Global Attention with Parametric Mish based Deep Learning Neural Network (GAPM-DLNN), Guided Kookaburra Optimization (GKO), Indexed Divide and Conquer (IDC), Quantum Fuzzy Inference System (QFIS), Ethernet, Distributed Packet Switching, and Local Computer Networks.


Paper Id: 231577

Published On: 2024-08-07

Published In: Volume 12, Issue 4, July-August 2024

Cite This: A GAPM-DLNN-based Attack Detection for Distributed Packet Switching in Local Computer Networks - Amaresan Venkatesan - IJIRMPS Volume 12, Issue 4, July-August 2024. DOI 10.5281/zenodo.14183832

Share this