In this paper, a high-performance network intrusion detection system based on deep learning is proposed for situations in which there are significant imbalances between normal and abnormal traffic. Based on the unsupervised learning models autoencoder (AE) and the generative adversarial networks (GAN) model during deep learning, the study aim is to solve the imbalance of data and intrusion detection of high performance. The AE-CGAN (autoencoder-conditional GAN) model is proposed to improve the performance of intrusion detection. This model oversamples rare classes based on the GAN model in order to solve the performance degradation caused by data imbalance after processing the characteristics of the data to a lower level using the autoencoder model. To measure the performance of the AE-CGAN model, data is classified using random forest (RF), a typical machine learning classification algorithm. In this experiment, we used the canadian institute for cybersecurity intrusion detection system (CICIDS)2017 dataset, the latest public dataset of network intrusion detection system (NIDS), and compared the three models to confirm efficacy of the proposed model. We compared the performance of three types of models. These included single-RF, a classification model using only a classification algorithm, AE-RF which is processed by classifying data features, and the AE-CGAN model which is classified after solving the data feature processing and data imbalance. Experimental results showed that the performance of the AE-CGAN model proposed in this paper was the highest. In particular, when the data were unbalanced, the performances of recall and F1 score, which are more accurate performance indicators, were 93.29% and 95.38%, respectively. The AE-CGAN model showed much better performance.

Application areas that utilize the concept of IoT can be broadened to healthcare or remote monitoring areas. In this paper, a remote monitoring system for patients at home in IoT environments is proposed, constructed, and evaluated through several experiments. To make it operable in IoT environments, a protocol conversion scheme between ISO/IEEE 11073 protocol and oneM2M protocol, and a Multiclass Q-learning scheduling algorithm based on the urgency of biomedical data delivery to medical staff are proposed. In addition, for the sake of patients’ privacy, two security schemes are proposed—the separate storage scheme of data in parts and the Buddy-ACK authorization scheme. The experiment on the constructed system showed that the system worked well and the Multiclass Q-learning scheduling algorithm performs better than the Multiclass Based Dynamic Priority scheduling algorithm. We also found that the throughputs of the Multiclass Q-learning scheduling algorithm increase almost linearly as the measurement time increases, whereas the throughputs of the Multiclass Based Dynamic Priority algorithm increase with decreases in the increasing ratio.

Pulse oximetry data such as saturation of peripheral oxygen (SpO₂) and pulse rate are vital signals for early diagnosis of heart disease. Therefore, various pulse oximeters have been developed continuously. However, some of the existing pulse oximeters are not equipped with communication capabilities, and consequently, the continuous monitoring of patient health is restricted. Moreover, even though certain oximeters have been built as network models, they focus on exchanging only pulse oximetry data, and they do not provide sufficient device management functions. In this paper, we propose an advanced pulse oximetry system for remote monitoring and management. The system consists of a networked pulse oximeter and a personal monitoring server. The proposed pulse oximeter measures a patient’s pulse oximetry data and transmits the data to the personal monitoring server. The personal monitoring server then analyzes the received data and displays the results to the patient. Furthermore, for device management purposes, operational errors that occur in the pulse oximeter are reported to the personal monitoring server, and the system configurations of the pulse oximeter, such as thresholds and measurement targets, are modified by the server. We verify that the proposed pulse oximetry system operates efficiently and that it is appropriate for monitoring and managing a pulse oximeter in real time.

We propose an integrated gateway for various personal health devices (PHDs). This gateway receives measurements from various PHDs and conveys them to a remote monitoring server (MS). It provides two kinds of transmission modes: immediate transmission and integrated transmission. The former mode operates if a measurement exceeds a predetermined threshold or in the case of an emergency. In the latter mode, the gateway retains the measurements instead of forwarding them. When the reporting time comes, the gateway extracts all the stored measurements, integrates them into one message, and transmits the integrated message to the MS. Through this mechanism, the transmission overhead can be reduced. On the basis of the proposed gateway, we construct a u-healthcare system comprising an activity monitor, a medication dispenser, and a pulse oximeter. The evaluation results show that the size of separate messages from various PHDs is reduced through the integration process, and the process does not require much time; the integration time is negligible.

We propose a smart medication dispenser having a high degree of scalability and remote manageability. We construct the dispenser to have extensible hardware architecture for achieving scalability, and we install an agent program in it for achieving remote manageability. The dispenser operates as follows: when the real-time clock reaches the predetermined medication time and the user presses the dispense button at that time, the predetermined medication is dispensed from the medication dispensing tray (MDT). In the proposed dispenser, the medication for each patient is stored in an MDT. One smart medication dispenser contains mainly one MDT; however, the dispenser can be extended to include more MDTs in order to support multiple users using one dispenser. For remote management, the proposed dispenser transmits the medication status and the system configurations to the monitoring server. In the case of a specific event such as a shortage of medication, memory overload, software error, or non-adherence, the event is transmitted immediately. All these operations are performed automatically without the intervention of patients, through the agent program installed in the dispenser. Results of implementation and verification show that the proposed dispenser operates normally and performs the management operations from the medication monitoring server suitably.