Low-latency Patient Monitoring Service for Cloud Computing Based Healthcare System by Applying Reinforcement Learning

Traditional healthcare systems are difficult to provide low-latency patient monitoring services for multiple patients. However, with the widespread application of novel technologies such as cloud computing, Internet of things(loT), 5G wireless communication, health industry is gradually walking into the era of smart and precise medical services with powerful healthcare cloud. Specifically, this kind of promising healthcare system which is based on cloud computing and loT, makes low-latency and multiple patients monitoring service possible, which signif- icantly improves patient's medical experience and doctor's fast access to patient's real-time data of vital signs for later precise diagnosis and analysis. Nonetheless, how to satisfy the low latency requirement while notably reducing the use of communication resource and computation resource of healthcare private cloud that has expensive cost is still a challenging problem. In this paper, we apply reinforcement learning(RL) algorithm which will learn a policy to automatically adjust transmission rate of monitoring services and computation resources of cloud servers, in conjunction with effective two-stage tandem queue system to tackle the above problem. Finally, we conduct extensive experiments to verify the effectiveness and efficiency of our proposed method in cloud computing based healthcare system.