Rapid technological advances and network progress has occurred in recent decades, as has the global growth of services via the Internet. Consequently, piracy has become more prevalent, and many modern systems have been infiltrated, making it vital to build information security tools to identify new threats. An intrusion detection system (IDS) is a critical information security technology that detects network fluctuations with the help of machine learning (ML) and deep learning (DL) approaches. However, conventional techniques could be more effective in dealing with advanced attacks. So, this paper proposes an efficient DL approach for network intrusion detection (NID) using an optimal weight-based deep neural network (OWDNN). The network traffic data was initially collected from three openly available datasets: NSL-KDD, CSE-CIC-IDS2018 and UNSW-NB15. Then preprocessing was carried out on the collected data based on missing values imputation, one-hot encoding, and normalization. After that, the data under-sampling process is performed using the butterfly-optimized k-means clustering (BOKMC) algorithm to balance the unbalanced dataset. The relevant features from the balanced dataset are selected using inception version 3 with multi-head attention (IV3MHA) mechanism to reduce the computation burden of the classifier. After that, the dimensionality of the selected feature is reduced based on principal component analysis (PCA). Finally, the classification is done using OWDNN, which classifies the network traffic as normal and anomalous. Experiments on NSL-KDD, CSE-CIC-IDS2018 and UNSW-NB15 datasets show that the OWDNN performs better than the other ID methods.

No abstract

The primary memory of the computer gets corrupted due to microparticles that possess a great deal of energy. This causes transient kind of errors and it is called subtle errors. High corruption of data in computer systems has been occurred because of these errors. Major sources of subtle errors include ionic particles, huge radiations say example alpha, beta and gamma rays, protons, neutrons etc., When TCAM configuration changes from 120 nm to 20 nm technology, there was a drastic increase in subtle errors more than multiples of eight. Devices operated in low power regions are more prone to subtle errors. Memory storage devices like secondary memory element called cache, registers are more feasible to subtle errors. 1 Behavior of subtle errors never damage circuit configuration but changes the information content leading to erroneous data. The proposed architecture is called N-SEARCH KEY TCAM, where in this system, it has been predicted that occurrence of subtle errors are very small. This innovative method of detecting and correcting both single and burst type of subtle errors has many modules like TCAM, Multiplexer, control circuit and Error correction code frame. Moreover, the proposed architecture N-SEARCH KEY ternary memory uses a special LUT for matching process. A key named Search key was generated which contains don’t care bits for matching subtle errors. Unlike regular TCAM, this design uses sequential LUT which enhances matching time of order [Formula: see text](wnlog2[Formula: see text]. Initially, proposed scheme have been used to find single bit subtle errors by changing forbearance factor ([Formula: see text] and chances of occurrence of error among TCAM information. The design was enhanced to detect multiple subtle errors and rectified the same. If match not occurs with predetermined information, then IP filter forbids the unauthorized users from using global network resources, thus providing better information security. By the proposed method, number of improper segregation ratio decreases with increase in Forbearance ratio parameter and time required to detect and correct the same is also considerably reduced. By controlling the forbearance factor [Formula: see text], the best duration schemes can be chosen. Through the given architecture, average time consumption for error checking of subtle errors would range from 0.6 to 0.7 [Formula: see text]s.