Optimization technique based on cluster head selection algorithm for 5G-enabled IoMT smart healthcare framework for industry

1 Introduction

A network of linked medical equipment and software known as the Internet of medical things (IoMT) allows for the sharing of medical records. For instance, fewer people will need to be hospitalized than would be otherwise due to the use of secure networks to communicate medical data and improve communication among doctors and patients. The IoMT has high requirements for storage capacity, processing speed, battery life, and network reliability. Because of 5G’s quick capabilities, IoMT can be employed in human healthcare applications such as diagnostics and therapy [1]. The term the IoMT is used to describe a system of networked electronic devices used in healthcare. Healthcare providers can improve clinical operations and workflow management, as well as remote patient health monitoring, thanks to the interoperability of equipment and sensors. The IoMT bridges the gap between the virtual and real worlds to improve patients’ health through faster, more accurate diagnosis and treatment, as well as through real-time behavioral and physiological changes. The impact on both patients and doctors of the increasing interconnectivity of medical equipment is substantial [2]. Emerging as a global infrastructure to link medical software and hardware applications, the IoMT is changing the way healthcare is delivered throughout the world. Healthcare Internet of things (IoT) is another name for IoMT. The IoMT uses many different types of sensors to monitor a patient’s vitals – such as accelerometers, visual sensors, temperature sensor, carbon dioxide sensors, echocardiographic sensors, pressure sensors, and gyroscope detectors – blood oxygen levels sensors, humidity sensors, respiratory sensor, and blood pressure sensor. Through the use of remote cloud data centers, the IoMT may detect a patient’s health status and relay that information to medical professionals and those responsible for patient’s care [3]. The capacity to remotely monitor healthcare systems is a major factor in IoMT’s rising popularity. This has knock-on effects that boost healthcare quality in areas including response time, disease diagnosis accuracy, and, most crucially, treatment delivery in the event of an emergency. A healthcare monitoring system that makes it easier to keep track of patients, doctors, prescriptions, medicines, and medical gadgets can be of great use to pervasive computing. With the use of machine learning, IoMT may be developed in a methodical way that may soon surpass human capabilities [4]. IoMT’s depiction in the healthcare system is shown in Figure 1.

Figure 1

IoMT in the healthcare system.

A collection of interconnected medical hardware and software is known as the IoMT. Patients, healthcare providers, researchers, and payers can all benefit greatly from the IoMT. Some examples include remote medical help, data insights, prescriptions management, operation augmentation, patient and staff tracking, or insurance claims processing [5]. The following are some of the problems that can arise when using IoMT. Keeping these in mind is essential when developing an IoMT infrastructure. The use of wireless medical equipment raises security concerns. The implementation of an IoMT system faces the difficulty of ensuring the interoperability of connections between devices manufactured by different companies. The IoMT system is a software application of medical analytic techniques, and clinicians need to be trained to use the system’s apps and software.

The most significant challenge of internetwork data transfer is ensuring the safety of transmitted information. IoMT platforms and applications that rely on a centralized cloud pose security risks. Existing methods suggested a secure technique or a mechanism that uses blockchain technology to guarantee the privacy, veracity, and integrity of data transmitted in IoMT. Due to resource requirements, blockchain-based solutions are challenging to implement on the IoMT end devices. Hence, in this article, the whale optimized weighted fuzzy-based cluster head (CH) selection algorithm for 5G-enabled IoMT is proposed.

2 Our contributions

• The Hospital Frankfurt Germany Diabetes Dataset (HFGCC) and Pima Indians Diabetes Dataset (PIDD) have 2,000 patients with diabetes and nondiabetes datasets that were gathered from the Kaggle dataset.

• Whale optimized weighted fuzzy-based cluster head selection algorithm (WOWF-CHSA) is used to improve effective communication for IoMT-based systems on the 5G network was developed.

• Low-overhead multi-hop routing protocol designed to be used for device-to-device (D2D) communication in 5G networks.

• Spider monkey optimization (SMO) is used to further increase the communication efficiency of the IoMT process for the 5G network.

The remaining of this article is categorized as follows: Section 3 presents the related works, Section 4 explains the experiment procedures, Section 5 presents the results of this study, and Section 6 concludes this article.

3 Related work

This article reviews several research papers and technical reports authored by diverse writers. The difficulties of dispersed and integrated systems are discussed, and other writers proposed solutions. The previous article [6] provided a novel clustering optimization method that considers communication distance, node density, speed, directions, and grid size. An intelligent and capable clusters head was chosen using a whale optimization method developed for clusters in vehicular ad hoc networks. The past study [7] introduced the infrastructure, design, and use cases for IoMT devices enabled by 5G. This research not only discusses the benefits of 5G-enabled IoT to the smart healthcare system but also compares and contrasts various wireless communication technologies available for this purpose. The need for and impact of 5G-enabled IoMT system technologies and requirements are discussed. The authors of ref. [8] explored deeper into the use of whale optimization algorithm (WOA), as well as its adaptations and hybridizations, in a wide range of engineering disciplines. The prospects, threats, and limitations to the further study are also discussed. The previous article [9] created a clustering model of medical applications (CMMA) that can be utilized to choose cluster leaders and facilitate efficient communication in IoMT-based software. The analysis of experimental results shows that the suggested CMMA outperforms competing methods in terms of both longevity and efficiency. As a result, the proposed CMMA not only evenly distributes CHs over the network to increase its lifetime but also minimizes the energy consumption of edge-based IoMT systems. The architecture of the IoMT is shown in Figure 2.

Figure 2

Framework of IoMT.

The previous article [10] examined the development of routing protocols and the quality of services in IoMT, including both traditional approaches and cutting-edge innovations. These protocols were created in a way that addresses the present and future needs of the IoMT network, which include the ability to deal with a great variety of data kinds, high data volumes, and power consumption concerns. The previous article [11] classified security risks to an edge network environment according to the primary security goals they threaten to undermine. In addition, on the basis of the existing research, they suggest a taxonomy for security countermeasures against assaults on IoMT edge networks. The purpose of this study is to lay the groundwork for future investigations into the development of efficient security countermeasures to protect the IoMT edge network both from internal and external attacks. The previous study [12] introduced the fuzzy logic clustering technique for the internet of medical things (FC-IoMT) technique, a powerful clustering method for IoMT applications that is based on fuzzy logic. CHs are selected by the provided FC-IoMT method using the following five inputs: energy, distances, delay, capacity, and queue. By using the FC-IoMT method, one can significantly cut down on the amount of energy an IoMT system uses. Using the IoT to improve the quality and efficiency of wearable devices, this article presents a comprehensive overview of how 5G has disrupted the medical healthcare industry. In addition, we also detail a state-of-the-art 5G sensor nodes architecture for noninvasive health monitoring of patients. The previous study [22] presented a new and highly effective method for early breast cancer diagnosis and prognosis. They also compared the predicted algorithm’s acquired accuracy with that of a standard CNN and other current classifiers, indicating that it excels in every respect. The previous research [23] discussed how the IoT and other technologies are being used in healthcare today. The safety of IoT-based healthcare systems was our primary focus because of the current preoccupation with the transmission of sensitive data. They have identified major security and privacy issues in the healthcare profession and offered a comparative analysis of them. A healthcare security case study has been probed. The previous study [24] presented various machine learning (ML) trends in the healthcare sector and explores their implications for the development and use of artificial intelligence (AI) in this industry. AI can be applied in the medical field, the hospital, and the insurance industry. Smart health care, where patients are prioritized and innovation is used to improve outcomes even outside hospitals, is a concept shared by many in the national healthcare industry around the world. The previous study [25] presented a novel resource-efficient flow-enabled distributed mobility anchoring paradigm. Internet protocol wireless has been thought to be a solution to these problems, and the network mobility basic support protocol is intended to support it. However, issues with the radio link were discovered during the handoff phase as a result of the patients’ extensive needs. New developments in IoMT are discussed in the study [26], with special emphasis on the part played by AI. The necessary hardware is discussed, as well as the most recent studies that propose answers to IoMT problems utilizing AI. The primary advantages and disadvantages are laid forth as well. In addition, wearable medical devices (WMDs) are analyzed. Technologies are also used in the categorization of WMDs. Reference [27] presented a structure for healthcare IoMT systems that are based on edge computing and discussed the research challenges connected with their implementation. Edge computing powered by AI has made important contributions to the ultra-reliable communication network of the smart healthcare system, particularly in the areas of reduced delay, increased device connectivity, and faster data transmission. The research offered FairHealth, a healthcare system at the cutting edge of 5G technologies that is driven by long-term proportional fairness. Next, they realize a trade-off between service stability and fairness by designing a Lyapunov-based proportional-fairness resource scheduling method. This approach separates the long-term fairness problem from single-slot subproblems. The previous study [28] presented a search certification framework for IoMT data on blockchain networks that makes use of 5G technology and edge computing to protect user privacy. The framework uses a highly effective multisignature approach to create a foolproof search result certification mechanism for IoMT data stored on the 5G edge blockchain. The previous research [29] investigated the edge-enabled IoMT-based system’s applicability and numerous extraordinary prospects in healthcare. Patients’ health conditions, such as movements, voice signals, temperatures, hypertension, glucose levels, and pulse rate, among many others, can be captured, diagnosed, and monitored in real time using the increased organizational and an assortment of devices and sensors. The technique can also be used to treat patients in critical situations, such as those involving heart attacks, hysteria, anxiety, and epilepsy. The previous study [30] proposed a 5G-based design for smart healthcare information infrastructure and defined a new network element called smart healthcare dedicated cloud platform, all of which collaborate together to meet the needs of individual hospitals while also allowing for growth. In addition, the methods used in implementation and the results of the associated field tests are given, demonstrating the substantial improvement in network performance made possible by the new system architecture developed [31].

4 Experimental procedures

This section outlines the proposed procedure’s overall flow (Figure 3). The schematic representations of a suggested technique include the processes like data collection of IoMT, WOWF-CHSA, low-overhead multi-hop routing protocol, and SMO.

Figure 3

The flow of the proposed work.

4.4 SMO

The cooperative behavior of spider monkeys inspired SMO, a population-based algorithm. It is modeled after the way spider monkeys, who have a fission-fusion social structure and forage intelligently. Members of a bigger or more permanent community establish fleeting little groupings among themselves in a fission–fusion social structure. Depending on the availability and abundance of food, monkeys form smaller and bigger groups.

• Fission–fusion social organization explains why these creatures live in stable groups of 40–50 members.

• A senior female general leads the group as they search for new food sources. It is held up as an example of worldwide supremacy. Whenever she is unable to locate enough food for the group, she divides the group into smaller subgroups (often between three and eight individuals) that go out to find food on their own.

• The daily foraging route taken by each subgroup is planned and decided upon by a single female leader. For this reason, we refer to this head of state as the “local leader.”

• Group members use a special call to talk to one another across great distances. Each spider monkey makes a unique call, and its identity can be quickly determined by others in the group. This aids the spider monkeys in preserving their communities and establishing safe zones.

The following data provide light on the inner workings of a spider monkey society and its peculiar ways of behaving.

• About 40–50 spider monkeys call one another home in the wild.

• During the day, the members of this society forage in tiny groups, venturing in various directions, and then return to their habitat at night to share what they have found.

• While planning a foraging trip, the dominant female spider monkey takes charge.

• If the group leader cannot find enough food, she divides a group into smaller groups and sends each group out foraging on its own.

• Due to their tolerance for one another, members of a society might not be seen closely at one location. When they interact, it is clear from their body language that they are a large group.

The spider monkey communicates its thoughts and observations through body language. They communicate across great distances with distinctive calls like whooping and chattering. Each monkey in a community can be recognized by its distinctive call. Figure 4 depicts the foraging behavior of spider monkeys, which was mentioned previously.

Figure 4

Foraging behavior of spider monkeys.

During the startup process, SMO creates a swarm of N spider monkeys with a uniform distribution, where SM i denotes the ith SM in the swarm. Initiation parameters for each SM i are as follows:

(2) SM i j = SM min j + U ( 0 , 1 ) × ( SM max j − SM min j ,

while U(0, 1) is a random number among 0 and 1, and SM min j and SM max j are the minimum and maximum values for the jth dimensional search space, respectively (0, 1).

A spider monkey’s social position may change depending on the leader’s and group’s previous experiences in the area. Each spider monkey’s new location is evaluated to see whether its fitness value is greater than its previous location. This is an example of the position update equation.

(3) SMnew i j = SM i j + U ( 0 , 1 ) ×   ( LL k j − SM i j ) + U ( − 1 , 1 ) × SM r j − SM i j ,

where LL k j stands for a jth dimension of the kth group’s local leader, U(1, 1) is the uniformly dispersed random variable in the range (1, 1) and SM r j is a jth dimension for a chosen SM from the kth group for that r ≠ i .

All symbols that have been used in the equations should be defined in the following text.

5 Results

This section covers the recommended framework’s general behavior. Parameters such as throughput, end-to-end delay, packet loss, or packet delivery ratio are compared in Figures 5–8. Comparison of existing methods includes vehicle clustering protocol (VCP), adaptive weighted clustering protocol (AWCP), enhanced whale optimization algorithm (EWOA), and adaptive weighted clustering protocol with whale algorithm (AWCP-WA). (Table 3) depicts the comparison of the proposed and existing methods.

Figure 5

Throughput results of proposed and existing methodology.

Figure 6

Results of proposed and existing methodologies’ end-to-end delays.

Figure 7

Results of proposed and existing methodologies for packet loss.

Figure 8

Results of proposed and existing methodologies for packet delivery ratio.

A system’s throughput is the rate at which it processes data, expressed in the number of data units processed per unit of time. Figure 5 shows that the proposed method of WOWF-CHSA has high throughput when compared to the existing methods.

Figure 6 represents the end-to-end delay results with proposed and existing approaches. Figure 6 shows that the existing methods have a high end-to-end delay when compared to the proposed method of whale optimized weighted fuzzy-based CH selection algorithm.

The results of the suggested and existing approaches’ packet loss calculations are shown in Figure 7. According to the aforementioned graph, the proposed approach of whale optimized weighted fuzzy-based cluster selection has a 58% lower packet loss than the existing methods.

A packet delivery ratio can be calculated by dividing the total amount of data packets collected by destinations by the total data packages sent by a source. This metric monitors how much information was properly delivered after being sent. Figure 8 demonstrates that when compared to existing approaches, the proposed method has a high packet delivery ratio of 96%.

The parameters show that the proposed method outperforms the existing method, which has several deep issues. These are a few of the issues with the existing strategy. This issue is remedied by VSP [14], which eliminates potentially dangerous nearby vehicles by filtering them out before the cluster-head selection process. In addition, one solution to this issue is to set a maximum and a minimum for the total number of vehicles allowed in a cluster. The maximum number of cars that can fit within a cluster is determined by a parameter called cluster size. Due to the competing goals and a large number of possible AWCP designs, a multi-objective optimization problem was developed [15,16]. The detection rate was reduced dramatically due to the limited capabilities of intrusion detection and prevention systems. The most effective method of resolving this security issue is to maximize the competing demands [17]. Inputs for AWCP are fed into the optimization problem, which attempts to maximize data delivery rate, minimize clustering overhead, and produce stable cluster topologies. We determined that the proposed method is better to the existing used techniques because it overcomes limitations in the existing used methods.

6 Conclusion

As the number of IoMT devices continues to grow, IoMT communication has become an increasingly vital part of 5G wireless communication networks in the healthcare industry. The proposed approach efficiently deals with the most important aspect of IoMT devices based on the CH selection. Some of the nodes in a cluster serve as the CHs, collecting information from the other nodes in a cluster. Moreover, the results are compared to the current methods in terms of throughput, packet losses, packet delivery ratios, and end-to-end time. Our suggested method performs better than the other methods that are presently in use. Through shedding light on the tremendous potential of the IoT in healthcare and highlighting the most important difficulties in IoMT’s future, we hope this study would be of service to academics and professionals. We proposed the whale optimized weighted fuzzy-based CH selection algorithm for 5G-enabled IoMT. The HFGCC and PIDD have 2,000 patients with diabetes and nondiabetes datasets that were gathered from the Kaggle dataset. The research is made to use to improve the efficiency of communication for IoMT-based systems on the 5G network, which was developed using the whale optimized weighted fuzzy-based CH selection algorithm, low-overhead multi-hop routing protocol, and SMO. The experimental results are provided as throughput, end-to-end delay, pack loss, and pack delivery ratio. A WOWF-CHSA was proposed, and its results were 96% of throughput, 50% of end-to-end delay, 58% of pack loss, and 96% of pack delivery ratio. The suggested method performs better than the existing methods. We emphasized the importance of 5G and the technologies that make it possible to use it to address the problems and constraints with current networks. Finally, we focused mainly on how 5G networks could be used to identify disease risks in the future and create a society that is ready to deal with them by embracing extensive automation and rising digitalization.