Sistem Deteksi Intrusi pada Server secara Realtime Menggunakan Seleksi Fitur dan Firebase Cloud Messaging

Abstract

Intrusion detection is one of the fundamental parts of a security tool, such as adaptive security tools, intrusion detection systems, intrusion prevention systems and firewalls. There are various kinds of intrusion detection techniques used, the main problem of this intrusion technique is the performance problem. The accuracy of the intrusion detection technique greatly affects its performance, which needs to be improved to reduce the false alarm rate and increase the detection rate. In solving performance problems, multilayer perceptron, support vector machine (SVM), and other techniques have been used recently. This technique shows limitations and is inefficient for use in large data sets, such as system and network data. Intrusion detection systems are used in analyzing a very large data traffic; thus, an efficient classification technique is needed to overcome these problems. This issue is considered in this paper. The well-known machine learning techniques, namely, SVM, random forest, and extreme machine learning will be applied. These techniques are well known for their ability to classify. NSL knowledge discovery and data mining datasets were used, which were considered as benchmarks in the evaluation of intrusion detection mechanisms. The results show that ELM outperforms other approaches. Utilization of Firebase Cloud Messaging because it can work with multiple platforms in addition to the availability of a file store that can store all logs created by the JALA application.