Analysis of Conventional Techniques on Detecting DDoS Attack in Wireless Sensor Network with Varied Feature Combination

Numerous security threats, such as sinkhole attacks, tampering attacks, and denial-of-service (DoS) assaults, are present in wireless sensor networks (WSNs). Thus, it is essential to design an effective system in order to identify and detect these assaults in WSNs in a timely manner. This article performs a thorough examination of several standard models used to identify intrusions in network environments. Features like statistical, higher order statistical features and technical indicators are considered in this analysis for the effective extraction of relevant features. Then, the conventional classifiers like Long Short-Term Memory (LSTM), Neural Network (NN), Support Vector Machine (SVM), Recurrent Neural Network (RNN), and Deep Belief Network (DBN) is deployed and analysed with respect to the extracted features. At last, the conventional techniques are analyzed with varied feature combinations. By investigating these models, we hope to gain a better grasp of their strengths and limitations, opening the door for the creation of WSN intrusion detection systems (IDS’s) that are more successful. This research also adds significant insights to the continuing efforts intended at guaranteeing the security and dependability of WSNs across diverse applications and domains by illuminating the subtleties of intrusion detection models.