Embedded Dual Band Rfid Based Blood Glucose Monitoring System For Internet Of Medical Things

Manually recorded health information could lead to errors such as inaccurate patient identification and mismatch patient data that could seriously affect patient safety. In order to reduce the risks of error for patients with diabetes, a new design of wireless blood glucose monitoring system with the embedment of dual band RFID for Internet of Medical Things is being developed. Using this method, passive RFID allows short-range communication to read automatically the patient identification number and active RFID extends long-range communication for recording and monitoring blood glucose data through multi-hop WSN. The work presented in this thesis contributes mainly to the embedded system and its application in healthcare to reduce the burden of recording, tracing and monitoring the patient‘s data by embedding blood glucose sensor, passive RFID, active RFID, WSN, M2M and IoMT into a single platform. A new design concept is established for the patient identification mechanism, where the mechanism is embedded in the source device to enhance the ability of the system to automatically assign the identification number to each blood glucose measurement (mmol/L) during multiple patients monitoring. Additionally, the results from the experiments conducted showed that the developed system produced better overall performance compared to the Bluetooth BGM and conventional BGM system in terms of the shortest recording time and the ability to retransmit data. In the reliability analysis using ANOVA and DOE statistical methods, the result validates that the number of hop and number of end node significantly affects the PDR performance of conventional CSMA/CA. These two parameters are then taken into account in experimental setup for performance evaluation of the enhanced CSMA/CA (EN-CSMA/CA) algorithm that uses an external interrupt mechanism and a cross layer approach. The PDR increased from 94% (conventional CSMA/CA) to 99.33% (EN-CSMA/CA), an improvement of 5.33%. The PDR model estimates that for the best and worst scenario, the percentage of PDR is 100.0% and 51.67%, respectively. To optimize the arrangement of routers for real implementation of the developed system in health facilities, the developed path loss model estimates that the router should be positioned at a distance of 30 m from each other, which agrees with the test results which indicate that the router should be positioned ≤ 40 m in order to achieve the best PDR performance