The Network Intrusion Detection System (NIDS) is a crucial aspect of safeguarding against cyber threats in the domain of cybersecurity. The major drawback of conservative approaches like Machine Learning (ML) that involve manual feature selection is their dependency on human involvement, which can hinder their efficacy. Deep learning (DL) is one of the technologies widely used in intrusion detection systems, it increases the performance of the model and securing the networks and classify the attacks. The primary concern regarding both convergence and speed along with the uneven values of the input-hidden layer were addressed as a gap in NIDS. This research compares the Convolutional Neural Network (CNN), Deep Neural Network (DNN), and Artificial Neural Network (ANN) are all related to the field of neural networks. The Performance is evaluated using the following metrics like Accuracy, Precision, Recall and true positive rate. For evaluating the effectiveness of the proposed model in both binary and multiclass classifications, the benchmark dataset CSE-CIC-IDS2018 is utilized. As per the experimental findings, the CNN model demonstrated exceptional performance, achieving an impressive accuracy rate of 99.72%.