Optimal transfer-interval frequency to minimize data loss in BLE network for healthcare service

The use of internet of things technology has substantially increased in healthcare. Accordingly, we have developed a digital therapeutics service, DfuKare, which contains a smart insole with multiple sensors named SmarTinsole for diabetes and diabetic foot ulcer patients. However, there are still challenges associated with using multiple sensors in digital therapeutics due to the characteristics of wireless communication, such as network connectivity, real-time capability, and data reliability & integrity. To address these problems, we developed a mobile application that communicates with SmarTinsole called Dfu-App. This paper describes the detailed implementation process from the communication perspective in particular. Moreover, we divide the experiment into three situations that are applicable to a real-world environment and then measure the data loss rate by changing the data transfer-interval frequency of SmarTinsole (10 Hz, 20 Hz, and 40 Hz). With 10 Hz, the data loss rate is 0 % in all cases, but in terms of the collected data, the performance is lower than that with either 20 Hz or 40 Hz. Although the results obtained with 20 Hz and 40 Hz show similar performance, we confirm that 20 Hz is the most appropriate data transfer-interval frequency while considering the hardware overhead of 40 Hz to be troublesome. Based on these results, we verify that communication with SmarTinsole and Dfu-App is reliable and secure, and it is expected that the data that can be collected using DfuKare will be helpful for building a dataset and designing an artificial intelligence model in healthcare service.