A Beetle Antennae Search-Based Sparrow Search Algorithm for Network Intrusion Detection

Nowadays, the network Intrusion Detection System (IDS) is a crucial cybersecurity technique that predicts unauthorized access, and exploits malfunction of computer networks. Traditional IDS struggle with classifying multiple, varied attack patterns which often lead to poor performance in detecting critical threats. In order to improve detection accuracy, identify numerous attacks, developing an efficient system becomes crucial which handles imbalanced data. Thus, this paper presents an Adaptive network IDS that employs a hybrid approach by combining the Improved Beetle Antenna Search (BAS) Algorithm and the Sparrow Search Algorithm (SSA) to enhance detection accuracy and adaptability in identifying cyber threats on NSL-KDD and CSE-CIC-IDS2018 datasets. Additionally, a Bi-directional Long Short-Term Memory (Bi-LSTM) network is utilized to reduce the False Positive Rate (FPR) and increase the entire efficiency of the system by effectively overcoming the challenges such as imbalanced data and low detection rate. Experimental evaluations on benchmark intrusion datasets such as NSL-KDD and CSE-CIC-IDS2018 using proposed IBAS-SSA + Bi-LSTM demonstrate higher detection results with accuracy of 98.12% and 98.64% respectively, when compared to other traditional IDS methods.