Improvement Networks Intrusion Detection System Using Artificial Neural Networks (ANN)

Artificial intelligence is one of the most important fields that exist today because of its many advantages in enhancing and increasing the efficiency of the system through which it is trained. In this paper, we will present artificial intelligence (AI) to improve the performance of the intrusion detection system in ad hoc networks (MANET), which are crucial in agricultural, industry, and IoT applications. MANETs are unique in their flexibility and mobility. MANETs are vulnerable to DDoS attacks due to node mobility, decentralized administration, and low bandwidth. Encryption and authentication cannot fully safeguard MANETs. To protect MANETs from these attacks, we need a more accurate system. IDSs monitor network traffic and report abnormalities for mobile ad hoc networks (MANET). As a monitoring system, it cannot adequately secure the network. Many scientists are using artificial neural networks (ANN) to identify intrusions. This paper focuses ANNs to improve intrusion detection system (IDS), reaction time, together with the rate of packet transport in these networks. In order to prevent packet failure, this method teaches intrusion detection systems (IDSs) to make judgments based on what they learn about attacks in the environment. Convolutional neural network algorithm: Following data collection using NS 2.4 and training in MATLAB using the CNN method, the CNN algorithm enhances the effects of IDSs in networks. Its end-to-end (E2E) performance as well as its average receiving packet speed of our model is superior to those of competing models. With CNN, 84% of our tests yielded positive results by 17 s.