The paper presents a network intrusion detection system (NIDSE) designed for smart home devices using an ensemble model, specifically XGBoosting. The system aims to identify various network attacks, including host discovery, SYN, ACK, and HTTP flooding, to enhance the security of smart home systems. The NIDSE is evaluated on the IoT Network Intrusion (IoT-NI) dataset, which contains nine types of network attacks. The performance of the NIDSE is assessed using cross-validation, and it achieves a micro average precision of 94% and a macro average precision of 85%. Compared to other machine learning algorithms like random forest (RF), balanced RF, and support vector machine (SVM), the XGBoosting classifier outperforms them in terms of accuracy, precision, and recall. The study highlights the effectiveness of the proposed NIDSE in detecting and classifying multiple network attacks, providing robust endpoint security for smart home devices.The paper presents a network intrusion detection system (NIDSE) designed for smart home devices using an ensemble model, specifically XGBoosting. The system aims to identify various network attacks, including host discovery, SYN, ACK, and HTTP flooding, to enhance the security of smart home systems. The NIDSE is evaluated on the IoT Network Intrusion (IoT-NI) dataset, which contains nine types of network attacks. The performance of the NIDSE is assessed using cross-validation, and it achieves a micro average precision of 94% and a macro average precision of 85%. Compared to other machine learning algorithms like random forest (RF), balanced RF, and support vector machine (SVM), the XGBoosting classifier outperforms them in terms of accuracy, precision, and recall. The study highlights the effectiveness of the proposed NIDSE in detecting and classifying multiple network attacks, providing robust endpoint security for smart home devices.