Article Preview
TopIntroduction
The healthcare sector is perceiving a huge growth fueled by a growing population and results in the focused wellness to the consumers. This is due to the information technologies-enabled patient care systems. To serve the same, the organizations are migrating to the cloud-based healthcare services to strengthen their expectation to meet with the demand. The volume of data the healthcare business gathers is mind-boggling. A 2014 report from consulting corporation EMC and research firm IDC place the volume of global healthcare figures at 153 exabytes in 2013 (an exabyte equals one billion gigabytes; five exabytes is equal to the all the words ever vocalized by humans). If the information stored all that data on tablet computers, the authors noted, the stack would reach nearly 5,500 miles. The report expected a 48 percent annual progress rate, meaning that the figure would reach 2,314 exabytes by 2020 and the same has been depicted in Figure 1. Figure 2 shows healthcare cloud systems.
Figure 1. The size of worldwide healthcare data collected
Cloud computing (CC) is an advance terminology; in terms of paid resources (Abdel-Basset et al., 2018). CC permits IT companies to consume the computational resources, just like storage, memory, CPU usage, etc., similar to any of the utility services such as electricity and water. Based on the usages of associated resources and cloud services the client has to pay the amount as per the billings (Cao et al., 2017). There are numerous characteristics of CC and one of the important characteristics of intelligence cloud resource management (Cao et al., 2017); is automatic provisioning and de-provisioning of the resources. In other words, the end-users can automatically ask for more resources and CSP (Cloud service provider) will release the resources when they do not require the said resources then the resources will be unprovisioned (released) (Auto-scaling feature with scale in and scale out). The auto-scaling technique (Alhamazani et al., 2015) can be resolved out by applying the monitoring approach. Monitoring of the cloud infrastructure results in providing application guarantees such as security, availability, and performance. This is also crucial from the perspective of CSP to maintain the demand of the clients without any interruptions. There are various tools (cloud monitoring products and vendors) for monitoring the cloud environment. Figure 3 show the various monitoring tools and their ratings as per the survey done by the Gartner 2018 (Gartner, 2018).
Figure 3. Survey of ratings: cloud monitoring tools
Figure 2 indicates the CC usages for the healthcare systems, such as virtual care and telehealth through internet access to the systems. Medical reminder and refill ordering (automatic), monitoring of real-time supply chain and event-based alerts and logging data (counterfeit and drug theft),artificial Intelligence based decision-making process mechanism, research based on social network, maintaining the individual data privacy, such as for HIPPA(Health Insurance Portability and Accountability Act of 1996) acquiescent and offsite servers with cutting-edge encryption and undeviating medical data in terms of standard EHR (Electronic Health Record) and portability across the healthcare providers.
In the proposed approach, we have adopted the mechanism of reinforcement learning (RL) to train the agent for the cloud environment. RL is a field of machine learning as shown in Figure 4 and its working has been represented in Figure 5.
Figure 4. Reinforcement learning: a part of machine learning