Implementation and performance evaluation of a real-time tele-monitoring system based on a private cloud platform

To monitor an out-of-hospital cardiac patient's vital signs better, the emergency medical service system, which responds to cardiac arrest outside the hospital setting, has been proposed by previous studies. However, the high data transmission rate requirement for information such as electrocardiogram (ECG) needs to be addressed to ensure complete data transmission and storage in remote database servers. The challenge is how to provide high service availability while many users are accessing the database. This study implemented a private cloud platform using the Microsoft Azure platform and Citrix XenServer to attain high service availability for multi-users while accessing the servers. A Web-based system was designed to provide the following functionalities: (1) administrative management functions, (2) tele-monitoring for single-lead ECG, and (3) emergency alert and response. This system was tested using JMeter for automatic testing. JMeter is a popular desktop application designed to test and measure the performance and functional behavior of client/server applications. Each time, the client user transmits one row of data containing 6 s of ECG data with 512 sampling rate per second and 3750 bytes in one row of data. There were 10, 20, 50, 500, and 1000 client users transmitting data to the server simultaneously. The results show that the latency of XenServer was kept at the same level. The latency of the Azure platform was initially smaller than that of the XenServer but increased gradually with the increase in client users. The latency of the Azure platform was greater than that of XenServer when the client users were greater than 100. The experiment results indicate that, from a performance perspective, a private cloud platform can fulfill the requirements if there are a small number of users. Virtualization is the better solution to satisfy a large number of users accessing the server.