Medical informatics, healthcare, and Internet health technologies are all combined in the cutting-edge system known as e-health. This combination encourages technological advancements to address persistent issues like lower costs and higher-quality healthcare. Likewise, the Internet of Things (IoT) enables a wide range of innovative initiatives to be implemented with shared and programmable technologies such as the Internet, cloud computing, storage, network-connected gear, software, and databases. When combined with adaptable, easily accessible, and enhanced patient health services, IoT and cloud computing has developed into revolutionary inventions that strengthen one another’s capabilities. This connection is easier to adopt in comparison to traditional networks, makes information more secure during engagements, increases data access, and increases productivity. Healthcare services can be considerably improved by IoT and cloud-based e-Health systems, which also encourage continuous, methodical development. IoT and cloud-based e-Health platforms allow users, providers, and servers to interact with each other while maintaining medical data in the cloud [1].
On the other side, there are numerous important problems with cloud computing, including traffic overload, huge data processing volumes, and data transmission delays. The main cause of these difficulties is that cloud servers are physically far away from IoT devices [2]. Healthcare apps cannot tolerate delays due to the critical nature of the industry. Employing specific cloud computing services to collect and analyze medical data from patients dispersed across a large geographic area is not practicable due to the significant transmission latency and high network utilization [3]. Fog computing has become a paradigm change to address the fundamental issues with conventional cloud computing as mentioned above [4]. As a component of the fog computing architecture, a wide variety of devices are linked to the network to supply computational and storage resources. Fog computing provides an infrastructure that is more flexible, and secure, and requires lesser bandwidth. The idea of healthcare systems reinforces that, in most countries, healthcare has challenges that only worsen with the aging population [5].
Health monitoring systems and Internet of Things advancements in wireless and cellular networks greatly enhance performance and save medical costs. Providing inexpensive home monitoring equipment which can spot early signs of deterioration, prompt service and treatment can be provided. IoT encourages observation of critically ill patients. Moreover, in developing countries, the population is mostly residing in rural areas where medical assistance is not easily available. Also in disaster-stricken areas, a framework is needed which can assist critical patients remotely using existing infrastructure. As a result, a framework for IoT-assisted health monitoring based on fog computing is suggested. The design of the suggested health monitoring system provides patients with ongoing real-time medical response, and fog computing has demonstrated its effectiveness in time-sensitive applications. Fog computing provides a way to handle the massive amounts of data created by end-user devices by deploying resources close to ending users. Due to the requirement for quick and effective data processing, fog computing should thus be employed in the scenario that has been presented.
The presented research is a bidirectional approach to improve real-time data transmission for health monitors by reducing network latency and usage. To that end, a simplified approach for large-scale IoT health monitoring systems is devised, which provides a solution for IoT device selection of optimal fog nodes to reduce both communication and processing delays. Additionally, an improved dynamic approach for load balancing and task scheduling is also suggested. Embedding the best practices from the IoT, Fog, and Cloud planes, our aim in this work is to offer software architecture for IoT-based healthcare systems to fulfill the aforementioned non-functional needs.
However, the computing and storage capacities of fog servers are constrained. The load on the fog server increases as the number of available queries rises in a vast system [6]. Patient requirements to install the proposed fog-based health monitoring system globally will put pressure on a single fog node. The fog node in this case gets overloaded while the other fog nodes are most likely dormant, increasing response time and causing the delay. Moreover, the selection of an inappropriate fog node for the IoT device gateway to transmit data increases network latency.
The three architectural layers that are suggested include sensors that are attached to the patients and can detect and transmit vital signs like body temperature, heart rate, and pulse rate to the IoT device gateways in the first tier. The fog servers, which are located within the same area as the base stations, make up the intermediate fog layer (BS). The fog nodes receive the real-time monitoring data from the first tier’s IoT devices, process it to determine whether the patient is in a critical condition, and then transmit their findings to the cloud server through a proxy server set up in the third layer. The fog nodes also notify the patient’s PDA devices of the results of their health status monitoring. Fog computing provides a way to handle a sizable volume of data produced by end-user devices by bringing resources close to end-users. Fog computing is therefore appropriate for the recommended technique because real-time efficient data processing is required. However, due to the pressure placed on fog nodes and the scanning of the IoT gateway for a suitable fog node to relay monitoring data, transmission is prone to delays. The IoT sensor gateway experiences a communication delay (Dc) as a result of the time needed to connect with an appropriate fog node, which renders it inappropriate for real-time monitoring. Additionally, the burden on the fog nodes increases the processing time (Dp), slowing transmission. To increase the quality of service (QoS), both network delays should be addressed. The overall delay can be written as,
$$\mathrm{Total}\;\mathrm{delay},\;{\mathrm D}_{\mathrm t}={\mathrm D}_{\mathrm c}+{\mathrm D}_{\mathrm p}$$
To manage the total network delay, this research suggests an innovative two-way method in the suggested architecture, which manages the incoming monitoring traffic at fog nodes. Firstly, by connecting to the optimum fog node, an efficient scanning mechanism (ESM) for the IoT gateway suggests a method to minimize communication lag. Secondly, we present a network-assisted Load Balancing scheme for real-time monitoring data (LBRT). The strategy effectively distributes the load to nearby fog nodes to reduce latency and network utilization because the health monitoring system’s time-sensitive nature makes it necessary. As we know, IoT data flows experience traffic and processing latency, as described in [7]. The frequent data transmission to the cloud server is reduced by using fog-based computing in the system design, thus lowering the system’s latency. The following are the contributions of this paper:
-
1.
A three-tier architecture is proposed for the fog-IoT health monitoring system, with fog nodes located in the middle layer. Smartphones connected to body sensors relay the patient’s physiological data streams to the fog node. To determine whether a patient’s health status is critical or not, fog nodes process the incoming data streams. The patient’s health information is thereafter sent to their smartphone and sent to the cloud for storage.
-
2.
In widely deployed health monitoring systems, the load-balancing of the real-time data scheme (LBRT) balances the load among fog nodes.
-
3.
The IoT gateway is suggested to use the efficient scanning mechanism (ESM) to rapidly and effectively choose a suitable fog node for transmission.
-
4.
Two criteria, namely latency and network utilization, are used to assess the performance of LBRT for health monitoring.
-
5.
Using the iFogSim toolkit, extensive simulations are performed to compare the performance of the LBRT to the conventional fog node placement method (FNPA) [8], and the load balancing scheme (LBS) [9].
The following is how the paper is set up: The context and inspiration of this work are presented in Section II. The most recent research on load balancing in fog-based systems and the design of health monitoring systems is presented in Section III. The proposed architecture for health monitoring systems is detailed in Section IV, and the real-time (LBRT) load balancing algorithm is covered in Section V. The discussions in Section VI and Section VII provide the conclusion and suggestions for future development.