PERFORMANCE EVALUATION OF A MACHINE LEARNING-BASED ANOMALY DETECTION SYSTEM FOR WIRELESS COMMUNICATION NETWORK
Aditya Shukla, Sawmya Thakur, S. V. Sankpal, Atharv Pingale, Pallavi H. Chitte · International Journal Of Trendy Research In Engineering And Technology · 2023
Wireless communication networks have become a critical part of our daily lives, and the Internet of Things (IoT) has enabled the connection of various devices to these networks.However, this has also made these networks vulnerable to various types of attacks, leading to anomalies in the data generated by these networks.This research paper presents an investigation of machine learning algorithms for anomaly detection in wireless communication networks using IoT intrusion dataset.We present a comparative analysis of popular machine learning algorithms: Decision tree, Random Forest, CNN, and XGBoost, for anomaly detection in wireless communication networks.The research paper aims to compare and improve the accuracy of anomaly detection algorithms for wireless communication networks by utilizing feature engineering techniques.These techniques involve extracting relevant features from the dataset to enhance the performance of the algorithm.To achieve this objective, the study conducts an in-depth exploration of the impact of different hyperparameters on the performance of the algorithms.The study utilises feature engineering techniques to extract relevant features from the dataset and explores the impact of different hyperparameters on the performance of the algorithms To evaluate the performance of the proposed algorithms, the study uses cross-validation techniques and measures several performance metrics, including accuracy, precision and recall.These help to determine the effectiveness of the proposed algorithms in identifying anomalous events in wireless communication networks.Moreover, the research assesses the efficacy of the aforementioned algorithms by comparing their performance with other contemporary state-of-the-art algorithms.This comparative analysis helps to determine which algorithm is best suited for detecting anomalies in wireless communication networks, depending on the characteristics of the available dataset.