Multi-Route Plan for Reliable Services in Fog-Based Healthcare Monitoring Systems

Introduction

The e-healthcare systems both web-based and mobile-based versions use wireless personal area networks (WPAN) or/and wireless body sensor networks (WBSN) for delivering high-quality real-time medical services and efficient medical treatments to patients (Sanna et al., 2014). In such systems, sensors are fixed around the patient’s body to collect vital information on the patient such as oxygen level, sugar level, heart rate, and pulse rate. This collected data is reported immediately to the remote designated physician and/or to a healthcare service provider attempting to ensure taking suitable action when detecting abnormalities, which is an application of Cloud-based e-healthcare systems (Xu et al., 2018; Park et al., 2019). Healthcare data needs a cloud platform to manage the large volume of generated data instead of relying on limited computing resources. However, this causes a high delay that affects healthcare services negatively especially those requiring an immediate response which is one of cloud computing drawbacks (Ahmad et al., 2016; Zhan et al., 2019).

In this regard, in 2014, Cisco announced a new computing concept which is fog computing, a new infrastructure paradigm to go beyond the confines of cloud computing (Bonomi et al., 2014). Fog computing infrastructure, on one hand, consists of many fog nodes, virtualized data centers, and IoT devices, which have an established connection with the cloud for more implementation of permanent storage and powerful computational capabilities (Dutta & Roy, 2017; Naranjo, Shojafar, Mostafaei, Pooranian, & Baccarelli, 2017). Fog computing is held to be a suitable paradigm when it comes to designing real-time or latency-sensitive healthcare applications. It is significantly contributing to healthcare applications such as Ambient Assisting Living (AAL) applications, remote home nursing that serves elderly people, and real-time tracking of chronic diseases (Al-Khafajiy, Webster, Baker, & Waraich, 2018; Awad, Khanapi, Ghani, & Arunkumar, 2019). However, due to its infancy; several other issues relevant to fog computing can be highlighted like sensor node failure or removal, external attack on BAN/WBAN, environmental coincidences, loss or limited power, loss of connectivity, and failure of the network, network congestion. Thus, system failure should be carefully addressed while designing emergency or real-time e-healthcare systems (Kher, 2016).

Node failure has a direct impact on many aspects of the system including deteriorate of the performance of the system, service disruption, increase the cost of operation, delay in service delivery time and unpleasant user experience (Satria, Park, & Jo, 2017; Ullah, Sehr, Akbar, & Ning, 2018). Most importantly, the issue of node failure has a significant negative effect on medical services in a healthcare monitoring system. This is because any medical service disruption due to the node failure might results in a potential loss of human life if the most appropriate action is not taken within the required time. Hence, a significant design feature of these strategies and recovery techniques must ensure service protection with low latency and less expensive execution costs (Khan, Parkinson, & Qin, 2017). Most of the previous approaches designed for healthcare monitoring systems assumed that the main fog node is always available and can entertain the user request, particularly for time-critical cases. However, this assumption is not always true, and the designated fog node might fail to entertain the user request and the latency to respond to the user becomes extremely long. Therefore, preventing the patient from getting the necessary immediate aid. Hence, longer latency might result in negative implications (death of the person) if the fog node fails to send the generated alert within a reasonable time. Thus, an efficient approach is needed to consider such cases. The approach has to identify and establish an alternative path through more than one fog node to be involved in such critical situations to ensure system reliability requirements. This paper proposes an approach that aims at establishing an alternative path to entertain the medical services for healthcare monitoring systems. The approach tries to make sure that the system is reliable in which the critical alerts can be reached to the fog node and the cloud instead of relying on a single path only that might leads to signal loss in any failure case.