Zobrazit v EDS

Chained Hash Table-Based Filtering Model for Detection of Flooding and Dos Attack in Manet

Uloženo v:
Podrobná bibliografie
Název: Chained Hash Table-Based Filtering Model for Detection of Flooding and Dos Attack in Manet
Autoři: P. V., Shyamily, B. K., Anoop
Zdroj: International Journal of Intelligent Systems and Applications in Engineering; Vol. 12 No. 20s (2024); 301-321
Informace o vydavateli: International Journal of Intelligent Systems and Applications in Engineering, 2024.
Rok vydání: 2024
Témata: Chinese Remainder Theory based Digital Signature Algorithm (CRT-DSA), Correlation Coefficient based Galactic Swarm Optimization (CC-GSO), Particle Swarm Optimization (PSO), Soft kernel Swish based Recurrent Neural Network (SS-RNN), Boltzmann Gibbs Trapezoidal Entropy based Fuzzy (BGTE-Fuzzy) Algorithm, Denial-of-Service (DoS), Mobile Ad Hoc Network (MANET), Chinese Remainder Theory based Digital Signature Algorithm (CRT-DSA), Boltzmann Gibbs Trapezoidal Entropy based Fuzzy (BGTE-Fuzzy) Algorithm, Correlation Coefficient based Galactic Swarm Optimization (CC-GSO), Soft kernel Swish based Recurrent Neural Network (SS-RNN), Particle Swarm Optimization (PSO), Denial-of-Service (DoS), Mobile Ad Hoc Network (MANET)
Popis: Mobile Ad Hoc Network (MANET), which connects mobile devices to all devices in the network where security is an essential task caused by Malicious Nodes (MNs), is a wireless communication technology. One of the crucial attacks that aim to exhaust network resources by flooding it with numerous fake packets as well as messages is the flooding attack. Hence, to detect flooding attacks, many models were developed. However, the models did not detect the sequences of flooding attacks. Therefore, this work proposes a soft kernel Swish-based Recurrent Neural Network (SS-RNN)-based sequential flooding attack detection model in MANET. Primarily, the nodes are initialized. Thereafter, the signature is created for the packet information utilizing the Chinese Remainder Theory-based Digital Signature Algorithm (CRT-DSA). Afterward, the packets are transmitted to nodes for connection establishment. The signed packet is verified, and to eliminate the attacked packets, the legitimate user’s packets are filtered utilizing Boltzmann Gibbs Trapezoidal Entropy based Fuzzy (BGTE-Fuzzy). By utilizing XNOR-HAVAL, the packet information is hashed and added into the chaining grounded on time series to identify flooding attacks, namely HELLO, Internet Control Message Protocol(ICMP), Synchronize an Acknowledgement (SYN-ACK), Acknowledgement (ACK), and Data flooding. After that, using SS-RNN, Denial-of-Service (DoS) attacks are detected. Lastly, for data transmission, the optimal path is created utilizing Correlation Coefficient-based Galactic Swarm Optimization (CC-GSO). To prevent security in MANET, the developed model classifies attacks with an accuracy of 98.41%.
Druh dokumentu: Article
Popis souboru: application/pdf
Jazyk: English
ISSN: 2147-6799
Přístupová URL adresa: https://www.ijisae.org/index.php/IJISAE/article/view/5142
Rights: CC BY SA
Přístupové číslo: edsair.issn21476799..2e876013af9bc72c1fd96d19ebeb8c37
Databáze: OpenAIRE
Popis
Abstrakt:Mobile Ad Hoc Network (MANET), which connects mobile devices to all devices in the network where security is an essential task caused by Malicious Nodes (MNs), is a wireless communication technology. One of the crucial attacks that aim to exhaust network resources by flooding it with numerous fake packets as well as messages is the flooding attack. Hence, to detect flooding attacks, many models were developed. However, the models did not detect the sequences of flooding attacks. Therefore, this work proposes a soft kernel Swish-based Recurrent Neural Network (SS-RNN)-based sequential flooding attack detection model in MANET. Primarily, the nodes are initialized. Thereafter, the signature is created for the packet information utilizing the Chinese Remainder Theory-based Digital Signature Algorithm (CRT-DSA). Afterward, the packets are transmitted to nodes for connection establishment. The signed packet is verified, and to eliminate the attacked packets, the legitimate user’s packets are filtered utilizing Boltzmann Gibbs Trapezoidal Entropy based Fuzzy (BGTE-Fuzzy). By utilizing XNOR-HAVAL, the packet information is hashed and added into the chaining grounded on time series to identify flooding attacks, namely HELLO, Internet Control Message Protocol(ICMP), Synchronize an Acknowledgement (SYN-ACK), Acknowledgement (ACK), and Data flooding. After that, using SS-RNN, Denial-of-Service (DoS) attacks are detected. Lastly, for data transmission, the optimal path is created utilizing Correlation Coefficient-based Galactic Swarm Optimization (CC-GSO). To prevent security in MANET, the developed model classifies attacks with an accuracy of 98.41%.
ISSN:21476799