Evolution and Adoption of Next Generation IoT-Driven Health Care 4.0 Systems

Abstract

The uprising of Internet of Things (IoT) has dramatically influenced the world’s technology in terms of interoperability, connectivity and interconnectivity with help of smart sensors, devices, objects, data and applications. General population aging, dearth of healthcare resources and upsurge in healthcare costs makes IoT advancements in healthcare all the more essential in order to confront these challenges. The revolution in IoT healthcare is redeveloping the healthcare sector in every aspect – social, technical and economical. This article presents a comprehensive survey on upcoming technologies helpful in healthcare 4.0 systems where the major focus is on emerging technologies like fog computing, cloud computing, machine learning and Bigdata analytics and that are all based on IoT based healthcare applications. In addition, the authors also provided an exhaustive survey on Wide Body Area Network (WBAN)-based IoT health care systems and discussed their network topology, architecture, platform, services and their applications. In addition, this study analyses IoT healthcare security challenges, possible threats, attack taxonomies and how blockchain technology can be helpful in countering these challenges. Lastly, exhaustive state-of-the-art technologies, challenges identified so far and possible future scope of this domain is also discussed in this survey.

1 Introduction

With the advent of Internet technologies, the healthcare sector has witnessed a lot of changes at a rapid pace. E-health is among the rapidly advancing services in which several automation services such as equipment for heterogeneous ubiquitous healthcare and advance clinical treatments, are being provided to the doctors [1]. From past many years, it has been observed that many patients have been found affected with several chronic diseases, which has overwhelmed our healthcare systems [2]. This has mounted a lot of pressure on our healthcare organizations as to how can healthcare solutions be designed such that they are economic, without maintaining the healthcare (HC) quality. Among the diseases that exist, some diseases such as Vertigo and Parkinson’s disease are some of the critical neurological diseases or NDs that need real time and constant monitoring. IoT is a major advancement in the era of Bigdata, which has been supporting several real-time engineering applications via enhanced services. Data collected from healthcare systems can be transmitted via these IoT devices to be analyzed, in order to discover detailed health information, early diagnosis and decision making over critical health situations in a timely manner. Eventually over the years, Bigdata’s prevalence in healthcare has paved a path for development of health 4.0 applications that are based on machine learning. Moreover, IoT in WBANs (Wireless body Area Networks) has been gaining a lot of focus from researchers in recent years. It has been observed from a study by World Health Organization (WHO) that these NDs contribute to more than 40% of the all the diseases that have affected the world, as can be seen from Fig. 1.

Fig. 1

Categories of diseases affecting the world and their percentage shares as per World Health Organization

Till now, numerous NDs and associated diseases have been uniquely identified and measured based on a number of different parameters. One such parameter is DALY or disability adjusted life years which is defined as the total healthy life years lost due to disability. Figure 2 indicates that NDs contribute to more than 6% of total DALYs caused by different diseases. This percentage is surprisingly higher than HIV AIDS which contributes to around 5.5% of total DALYs. This highlights the fact that NDs are a major cause of concern in healthcare and need to be discussed better.

Fig. 2

Percentage contribution of different disease categories to total DALYs

To provide some relief in the health sector, IoT’s (Internet of Things) capabilities can be explored. As a matter of fact, healthcare and medical sector together constitute to be a core attractive application area for IoT [3]. IoT has the capability to be of great advantage when it comes to health sector applications such as care for old-aged people, fitness and health programs, management of.

long-term diseases and remote health monitoring (RHM). Another important HC application where IoT can be utilized is, the patient being compliant with the medicinal dosage and other treatment suggestions at home, in accordance with the recommendation provided by the healthcare workers. Hence different sensors in medical devices such as imaging devices and other diagnostic devices are all referred to, as smart devices having IoT as their core principle. Healthcare costs reduction and enhancement of the life quality to enrich user’s experience are the potential assets of these IoT HC services. IoT is also beneficial, according to healthcare workers’ point of view, because it can reduce the downtime of other healthcare devices by using downtime monitoring of those devices, by utilizing the potential capabilities of IoT. IoT can be applied in different ways in the medical sector to be of benefit. For example: it can correctly determine the optimal timeline for inventory replenishment of medical supplies of healthcare devices in hospitals, in order to make ensure their uninterrupted and continuous operation. In another scenario, IoT can be used for optimal scheduling of healthcare resources that are limited in number, thus ensuring a better service to a greater number of patients [4].

A recent important trend has been the ease of cost-effective interactions between the individual patients and the HC clinics and agencies through secure and seamless connectivity. Latest HC networks that are powered via wireless-technologies, are being expected to help deal with the long-term diseases, their timely diagnosis, RHM and other health emergencies. HC servers along with gateways and HC databases are playing a key role in creation and maintenance of HC records and the delivery of health services on-demand, for all legit stakeholders. In recent years, the domain of IoT successively and massively gathered wide attention from research scholars around the world to address its potential in the healthcare sector, by taking into consideration the several practical challenges, that the field poses. As a result of it, there now exist multiple service solutions, applications and prototypes in IoT healthcare. New applications and services, network architectures and platforms, interoperability and security are some of the major research trends in IoT healthcare. Many countries and health organizations around the globe have already formulated some policies for the deployment of IoT in the HC field. Today, an in-depth understanding of existing research works in IoT healthcare is expected and required by various stakeholders to pursue further research in this area. Through this paper, we examine the trends in IoT HC research and inspect several issues that need to be addressed to revolutionize the HC industry using IoT. In upcoming sections, we see how IoT based on fog computing along with cloud solutions is helpful in catering to these critical requirements of the healthcare sector.

1.1 Fog Computing

Development of Cyber Physical Systems (CPS), mobile internet and IoT has led to various objects including people, things and machines to get connected in the information space, as per will [5, 6]. This has led to an unparalleled generation of huge amount of heterogeneous data. Cisco’s estimations claims that by 2020, there will exist around fifty billion devices connected to the internet. Huge volumes of data being continually generated is estimated to hit five-hundred Zettabytes, of which approximately 45% of the IoT data will be stored and processed at the network’s edge by the year 2019[7, 8]. As expected, with the rising growth in volumes of data generation, the pace at which it is being generated has also been rising quickly. An analysis of recent HC data from the IoT apps suggests that only thirty million users generate as high as 25,000 data tuples per second [9]. Our current storage and processing capabilities are not sufficient to satisfy this rapidly rising data volumes [10]. Moreover, apart from that our traditional computing models namely cloud computing and distributing computing also proved to be insufficient, when it comes to satisfying the need of handling this huge data. This led to the development of new technology that is termed as Fog Computing by Cisco.

1.2 Why Fog Computing?

Cloud computing which is a conventional computing paradigm has been utilized ever since, because of its high efficiency in processing data with massive storage and high computability. [12, 13]. However, most of the computation required are computed in the cloud itself because cloud computing (CC) is essentially centralized in nature. So, it requires that the data as well as other requests too, have to be sent over to the CC servers. Though we have seen an appreciable rise in processor speeds over the years, the same cannot be said about our network bandwidths. Network bandwidth has become the bottleneck for CC, thus deterring it from handling vast volumes of data which eventually results in long delays or latency.

In some application scenarios, we strictly require quick response time and support for mobility. Example of such applications are emergency response, smart traffic signals system, smart power-grids, smart HC and other apps sensitive to latency [14,15,16,17,18]. The delay occurring due to data transmissions to and from cloud servers, is unacceptable in these cases. Moreover, depending on.

the situation and kind of data, some decisions can be made locally instead of being transmitted to the cloud. If some decision making is required to be done in the cloud only, then also it’s inefficient transmit complete data to CC servers for processing/storage, as the entire data is rarely needed for analysis or decision making. In other words, the challenges posed in terms of network bandwidth, latencies, reliability challenges and other security issues caused by explosive growth rate of data, cannot be solved by complete dependence on the CC model.

Fog computing or FC, which possesses the capability of seamlessly integrating edge network devices and our CC center, is deemed to be a much effective solution to address the problems left unaddressed by CC model, as highlighted before. FC is a computing architecture, geographically distributed in nature, in which various heterogeneous devices are connected at network’s edge, together providing elastic storage, communication and computation services [19]. Extension of CC services to network’s is the most prominent feature of FC as represented in Fig. 3. FC works by pooling the local resources together, thus bringing the storage, computation resources, communication and control, nearer to the users. Most of the communications and data transfers are now between devices and fog instead of devices and cloud and thus, the data transmission time and volume of network transmissions become reduced [20]. Thus, FC paradigm resolves the network bandwidth bottleneck issue in cloud and is suitable for latency sensitive applications i.e. all applications that need to respond in real time. Table 1 summarizes these differences between FC and CC.

Fig. 3

Fog architecture as an intermediate between IoT devices and Cloud Layer

Table 1 Comparison between CC and FC

1.3 Advantages/Requirements of Fog Computing in IoT

FC performs storage, communication and computational tasks on network edge devices which are in end user’s vicinity. Its service facilities are in close proximation of its users. This is termed as FC’s basic feature and its most significant advantage compared to the traditional CC model. Apart from this, FC has several other features and merits illustrated in Fig. 4 and listed as follows:

• (a) Minimal Latency: The sensory information is gathered from the edge nodes located at the network’s edge then process that data and finally store that data in the LAN. This leads to significant decrease in data transmission over the Internet and premium quality localized services are provided. Thus, latency is reduced and demands of real time applications is fulfilled [21]. Sarkar et al. [22] performed theoretical modelling of FC system architecture and observed that the delays in FC systems were comparatively much higher than that of CC systems. Hu et al. [23] conducted experiments in the domain of face recognition, by deploying both FC systems and CC servers. He observed that the response time was very less in FC as compared to CC.

• (b) Interoperability: FC nodes as well as the end devices are usually manufactured by multiple different providers and need to be deployed in several diverse environments. Hence it is required that FC should be capable of inter-operating with different device and node manufacturers, to seamlessly provide support to several devices seeking its service [24]. In the streaming service supported by FC, coordination among several providers is required as the service is federated among several different domains [21]. In FC based smart transport systems, Realtime analysis of data is needed along with dynamic transmission of data between smart vehicles, fog nodes, fog application and traffic lights. To achieve data sharing and collaboration, resource management guided by a set of policies is proposed in order to enable interoperability combined with secure collaboration between the resources requested by user in FC [25, 26]. All policy requirements such as network requirements, operational requirement and security requirements are supported by well-defined policy specifications. It is due to these policy specifications that uniform and secure collaboration is established whenever the communication occurs in the distributed or dynamic environment. Hence in this manner, interworking, interconnection and interoperation of different heterogeneous resources and devices is realized by fog computing.

• (c) Mobility: In the particular case of fog computing, we have quite a lot of diverse end devices, traffic cameras for instance, that remain static and many other devices such as smartwatches, vehicles and smartphones that cause frequent spatial mobility at the terminal layer. In the same way FC nodes in our fog environment can either be either static or mobile computing resource platform and accordingly these can be deployed in coffee shops and airport terminals or on mobile trains and buses [27,28,29]. Apart from that, there is direct communication between the various devices as well. The data now need be sent over to the CC servers. Realization of mobile data analysis is achieved by the processing of voluminous data that these IoT devices generate, in the end devices itself or in the intermediate devices. So, service to more extensive nodes can be provided by it. For direct communication with the mobile devices and users, FC apps can use routing, communication and addressing protocols. For example, ID or location separation protocols for mobile nodes (LISP-MN), involves decoupling of location identity and host entity and to support mobility techniques, it requires a distributed directory system [30, 31]. Moreover, the mobile devices’ access depends on physical closeness in some communicational technologies. For instance: NFC (Near Field Communication), Bluetooth and Millimeter Wave Communication. In this manner, any irregularities in the network connection possibly caused due to mobility, can be avoided [32]. Also, by employing the concept of ‘Data Sherpa’, it is possible to transmit data from static sensors on the edge to smartphones (on edge) via Bluetooth, whenever the they are in physical proximity and then the smartphones can send the data to FC or CC servers [27, 33]. FC supports the location-based mobility which empowers the admin to control where the mobile devices and the users can come in and the protocol via which they can access that information [34]. This leads to improvement in both the quality of service and the performance of the system.

• (d) Security: The services hosted by fog computing are closer to the end users in terms of physical distance. This in turn is advantageous in terms of both privacy protection and data security. Firstly, data is protected in FC by encryption techniques and measures for isolation. FC nodes also present encryption schemes, access control measures, isolation measures and data integrity checks to ensure the security of private and confidential information. Moreover, FC also eludes the potential risks from an upgrade in the system. Remote upgrading of traditional services has been found to be generally inefficient and there are associated risks such as middleware upgrade may lose contact while upgrading. FC doesn’t require an Over the Air (OTA) middleware upgrade in the system, it just updates the micro applications and algorithms at the fog end.

• (e) Geographical Distribution: The applications and services of fog computing involve a geographically distributed deployment as opposed to cloud computing which follows a centralized model. It comprises of many geographically decentralized nodes, which can track and locate the end devices for the mobility support. The FC architecture of distributed nodes ensures the accessibility of the data analytics and statistics with low-latency to the user, rather than storing and processing of data in centralized CC environment, which is far away from the user. This feature of fog supports better location-based services, quicker analysis of Bigdata and powerful capability of decision making in real-time. Interworking and interconnection among ubiquitous things are one of the goals to be achieved in IoT. The things referred to over here, are both voluminous in number plus widely distributed. Decentralized data analytics and geographical distribution are two characteristics which can help satisfy the abovementioned requirements efficiently. For instance, in Internet of vehicles (IoV), FC can present multiple IoV services such as traffic security, infotainment, road and lanes conditions etc., based on interactions between vehicles and between vehicles and access points [24].

  Fig. 4

  

  IoT Requirements from Fog Computing

The main contributions of this survey are as follows:

• 1. Highlighting the need of IoT and FC in healthcare and how FC is an utmost requirement in healthcare in contrast to CC.

• 2. Study of IoT healthcare networks under three different components: Network topology, network architecture and network platform.

• 3. Study of classification of healthcare IoT into two broad categories: applications and services. Surveys of existing IoT solutions have been conducted in this study. Application of machine learning algorithms such as SVM by different studies in the bigdata analytics environment have been compared in our paper.

• 4. Provides a detailed insight into the requirements and challenges to be dealt with while developing wearable and products and prototypes designed by new age wearable tech-companies.

• 5. Consolidated study on security challenges possible in IoT healthcare systems and how blockchain when integrated with IoT can serve as a possible solution to avert such security issues in healthcare

The rest of the paper is organized as follows: Sect. 2 introduces integration of IoT in healthcare and detailed study of IoT networks in healthcare and its composition. In Sect. 3 we discuss about IoT services and applications and survey of various IT solutions working on edge cutting tech-algorithms of machine learning and AI in bigdata IoT environment. In Sect. 4, we talk about wearable technology, its worldwide statistics and various studies conducted to counter challenges faced in building wearables. Section 5 lists out various security issues and blockchain as a viable option to counter those challenges. Lastly, we conclude our study with summarization and future works in Sect. 6.

2 IoT in Healthcare 4.0

As mentioned before, in the medical care sector, IoT performs vital role in several healthcare 4.0 related applications. Automated medical data collection method is used to rule out the possibility of human errors during the process of data collection. This helps in improvement of the quality of diagnosis and eliminates all risks due to human errors which can be a reason for false diagnosis which can eventually be harmful for the patient’s health. The role of IoT in healthcare is widespread, but is most vividly visualized under the following three healthcare paradigms:

• 1. RHM: It is one the most vital paradigm for real world health applications. Numerous people around the world including the children and elderly, suffer from chronic illnesses that need to be monitored daily [35]. Remote monitoring will reduce patients’ trips to the hospital for daily checkups. Remote access sensors help in early and timely diagnosis of diseases [36]. Thus, patients suffering from critical diseases like cardiopulmonary diseases, heart diseases and asthma, can benefit a lot from RHM. Our real-time monitoring system includes an RHM center and an RHM unit. It performs a real-time analysis on the sensor data and alerts with a warning sign in case of emergencies. The signals collected through the sensors are accumulated and transmitted to the respective health centers via WLANs. Thus, remote monitoring provides precise and real-time monitoring of patients’ data, so that emergency healthcare units are informed timely, in case of patient being in a health emergency and they can provide necessary help.

• 2. Clinical Care: Every hospital has numerous patients with critical illness that need to be admitted to the ICUs and require constant monitoring can be helped via IoT monitoring systems. These monitoring systems consists of sensors that are deployed to accumulate physiological data about the patient. This data thus gathered, is examined by the IoT system and sent to patient’s family, doctors and nurses for further analysis, if required [37].

• 3. Context-Awareness: It assesses the patient’s health status in the environment in which he/she resides, that can help doctors and other healthcare staff to comprehend the relationship between the surroundings and the patient’s health. This is done because changes observed in residing environment of the patient can impact his vulnerabilities to diseases and can even lead to deterioration of his health [38]. The use of specialized sensors that capture patient’s physiological and physical information including exercises like weightlifting, cycling and sleep pattern and the environmental set-up in which he is, to give a deeper insight into his/her health status. Further, it will be helpful in locating the patients in case of any emergencies and being better prepared for dealing with such situations. We have covered the brief classification of IoT application areas in healthcare so far. For successfully setting up an IoT healthcare network in any healthcare application, we need to look at how the different IoT Healthcare networks are structured, its architecture and other technical aspects. This is discussed under upcoming headings.

Healthcare Networks in IoT: Multiple sensors along with diverse heterogeneous devices need to work in collaboration in IoT ecosystem, in order to lead to successful functioning and advancement of our healthcare infrastructure [39]. Reception and transmission of healthcare data of healthcare data is achieved by optimization of communication protocols and algorithms which is further based on an efficient WBAN [40].

2.1 IoT Network Topology

It is characterized by an aggregation of different components playing diverse roles in IoT healthcare 4.0 systems. It also describes numerous features of persistent e-healthcare environments [41]. IoT in WBANs is used for remote monitoring of patients. The handy medical IoT instruments and on body sensors connected to patient’s body gather all information related to patient’s vitals and this data is then sent over to the fog nodes for analysis. This helps in building patient’s health profile using IoT and then the health workers can respond accordingly [42]. In Fig. 5, we have represented the healthcare data’s streaming by using interconnected network with WiMAX, the access service networks and the Internet Protocol (IP) network. The topology represents the principle role of gateways. Figure 5 also represents an iMedpack, an IoT devices that manages problems in the pharmaceutical sector by preventing the medicinal misuse [43]. Moreover, various interfaces and sensors of wireless topologies, present gateways that are critical from the healthcare apps’ perspective [44]. The e-healthcare system ensures the detection of health issues by connecting different IoT devices to the e-health gateway where the data collected by sensors is stored, processed, analyzed and the analysis is then presented. Hence every IoT topology and associated clinical equipment are clubbed with good amount of e-healthcare infrastructure. The IoT network topology thus provides the activities related to healthcare system and is there for the perception of the medical staff [45]. The topology thus includes, the health system in semantic e-health monitoring system.

Fig. 5

Network topology in RHM based on IoT

2.2 IoT Network Architecture

The basic IoT healthcare 4.0 architecture is composed of three layers namely: the perception layer followed by fog network layer and finally application layer as show in Fig. 6 [46]. The prime role of perception layer is identification of devices and gather the information. This layer is the first layer and also more frequently contacted layer by other entities in our healthcare system namely patients, devices and nurses. Medical devices and sensors such as tracking devices and heart rate monitors are some of the entities that are responsible for collecting the information. The second layer is the fog network layer that transmits the information collected in the perception layer to application layer. The data collected from the perception layer is transmitted to the network layer via different protocols. The network layer is responsible for creating a permanent connection for the transfer of data from perception layer to application layer and resolving interruptions if any. Third layer is our application layer. This layer consists of fog local network example: localized servers linked to localized Hospital Information System [47,48,49,50]. As we know that IoT requires various heterogeneous devices to be connected, it is important that we develop a configurable and flexible architecture [46, 48,49,50]. Globally, there exists no mutual agreement for the standardization of IoT healthcare architecture and as a result there are different architectures proposed in different research papers [51]. Hence it is inferred that based on the requirements and other factors like infrastructure, various different architectures with different number of layers in them are probable. In Table 2, we have summarized various studies based on the total layers that their model employs along with the communication model used in each of them respectively.

Fig. 6

A Fog Network Architecture in IoT healthcare

Table 2 Summary of e-healthcare Studies based on their IoT architecture and Communication Protocols

2.3 IoT Network Platform

Fog computing platform along with IoT medical/healthcare network platform are two essential platform systems in the IoT system. As represented in Fig. 7, the service platform model majorly concentrates on resident health information. The network platform depicts the cataloguing of e-healthcare structure and demonstrates the procedure with the help of which the healthcare staff can use foremost application layer to gain access to different databases, as per the ‘support layer’ [69]. It ensures that an automated design methodology platform and the associated interoperability is basically for the rehabilitation method, presented in it. Access layer, business implementation layer along with data persistence layer and the support layers are the four layers included in the framework [38]. One enabling function’s required by the gateway to handle multiple sensors having multiple users. This method makes it possible for the IoT ecosystem to manage several users simultaneously by using numerous sensors while collecting the data for healthcare. The database platform in IoT is exploited in a ‘multi-tenant’ fashion and it is the ‘resource layer’ that is accountable for the interpolation of healthcare data, applying resource control mechanism and data sharing in healthcare systems. Hence, the displayed meticulously organized framework has the capability to provide user environment and semantic potential of IoT in the medical sector.

Fig. 7

IoT platform framework for e-healthcare

So far, we have discussed the core underlying IoT principles required for IoT health networks development. In the next section, we would provide a broad classification of IoT healthcare into services and applications.

3 Applications & Services in IoT e-healthcare

IoT e-health solutions are applicable to several diversified fields including but not limited to elderly patients’ care, continuous supervision of long-lasting diseases and also management of people’s personal health and their fitness etc. This paper does a broad categorization of this extensive topic into IoT applications and IoT services. Single conditioned and grouped condition applications are further sub-divisions of the IoT applications. Single condition applications deal with one specific disease whereas a grouped condition application refers to a bunch of medical conditions or diseases. This categorization is illustrated in Fig. 8.

Fig. 8

Categorization of e-healthcare 4.0 into services and applications

3.1 IoT e-Healthcare 4.0 Services

IoT is expected to support the diversified set of HC services, with each and every service providing a specific bunch of e-health solutions. A service is generally generic and functions as an elementary block to form diverse applications and e-healthcare solutions. Already existing general services needed by IoT frameworks require trivial amendments, to achieve accurate functioning in e-healthcare models. These include services for resource utilization in a shared network environment, notification services, protocols for establishing cross platform connectivity and internet services for devices, which might be heterogeneous. Linking protocols are also included to establish connectivity. Some of the IoT services in healthcare are:

• 1. Ambient Assisted Living (AAL): It is an artificial intelligence (AI) powered IoT platform that is developed to address the health of the elderly, aged or incapacitated individuals. AAL’s aim is to expand the life span of the elderly in a manner convenient and safe for them, within their residence of living. Various studies have been discussing AAL with IoT for years now. An architecture with security and communication control has been introduced for implementing IoT’s AAL in [70]. The proposed architecture can be considered to be the framework to provide health service for the aged and incapacitated patients. Implementation of this architecture includes underlying technologies that include NFC and RFID for passive communication and 6LoWPAN for active communication.

• 2. Adverse Drug Reaction (ADR): It refers to a medication induced injury [71]. This can happen owing to many reasons such as only one drug dose, or its consumption over long period of time or as a result of mixing of more than one non compatible drugs. ADR is generally not specific to particular medicines or diseases, but is inherently generic, so there’s a requirement to individually develop some frequently occurring technical faults with their solutions, known as the ADR services. An ADR that’s based on IoT has also been proposed where terminal of the patient detects and determines the drug by using barcode or even NFC enabled devices in [72]. This information is coordinated to an intelligent pharmaceutical information system, that evaluates the drug to be consistent with its e-health informational records and allergy history profile. As a part of iMedBox, iMedPack has been developed, to address ADR by using CDM (controlled delamination material) and EFID technologies [73].

• 3. Embedded Context Prediction (ECP): These are frameworks that are generally generic and are required by the developers to develop a context-aware healthcare application on the IoT network. A similar framework in ubiquitous healthcare’s context has been developed in [74]. Context-aware ubiquitous healthcare systems in nature have given rise to several sets of challenges and disputes as expressed in [75]. Similar research issues and challenges are required to be considered and resolved for building context-aware e-health apps over the IoT network and the prediction of context is then utilized for IoT based RHM in [76].

• 4. Internet of Medical Things: m-IoT consists of HC sensors, technology for communication and also mobile computing, all coordinating to provide HC services as shown in [77]. M-IoT is like a novel model for healthcare connectivity that can connect the 6LoWPAN to the evolving 4G networks to provide internet-based health services in the future. Though m-IoT in general is representative of IoT health services, there also exist certain specific features that are intrinsic to the participating entities’ global mobility. This led to the formulation of m-IoT services. The utilization of m-IoT services was analyzed on the basis of its glucose level sensing potential, in a noninvasive manner and its architecture and all the implementation challenges all have been covered in [78]. m-IoT ecosystems and context awareness issues are two challenges that exist in m-IoT [79]. Another system with mobility on the basis of message exchange has been proposed in [80].

3.2 IoT e-Healthcare 4.0 Applications

Apart from e-health services, we also have IoT applications that need to be discussed. IoT services provide help to design and build IoT applications which are then used directly by the users. Hence to differentiate them in a better manner, we say that services are developer centric, but applications are on the other hand user centric. Apart from the IoT applications considered in this section, we have different healthcare devices, gadgets and wearables available in the market by various brands are discussed in the upcoming sections. Some of the IoT healthcare applications are as mentioned:

• 1. ECG Monitoring: Electrocardiogram Monitoring or ECG Monitoring records the electrical activity of the heart by electrocardiography. It involves measuring the heart rate and determination of its rhythm along with diagnosis of long-lasting intervals of QT, myocardial anemia and multi facet arrhythmias [71]. Using IoT in ECG monitoring can provide maximum possible information and be used to full extent as in [81]. Several studies have discussed ECG monitoring using IoT in detail in [74, 82,83,84,85,86,87]. An ECG system based on IoT consisting of a wireless reception processor and a wireless portable acquisition transmitter has been proposed in [88]. The system detects abnormal data to identify cardiac activity in real-time by integrating a search automated functionality. There also exists a detailed algorithm for detection of these signals of ECG in IoT network’s app layer [89].

• 2. Monitoring Oxygen Saturation: We use pulse oximetry to continuously and noninvasively supervise the saturation levels of oxygen in blood. IoT’s pulse oximetry’s potential is covered in CoAP based HC service survey in [90]. Wrist OX2 is a wearable that functions as a pulse oximeter whose functionality is discussed in [82]. The device features Bluetooth based health profile along with sensor connectivity to a platform called Monere. Also, a low cost and low power consuming pulse oximeter has been proposed for RHM in [91]. The device is used for constant monitoring of patients’ health status over the developed IoT network. For telemedicine application, a pulse oximetric system is proposed in [92].

• 3. Medication Management: Around the world there exists financial wastage and threat to the general public health because of noncompliance issues in medication. IoT has some promising solutions to offer in order to resolve this issue. Intelligent packaging methods for medication management driven by IoT has been proposed as in [93]. These methods involve a precursor for iMedBox and performs field trials for verifying the system. Wireless communications-controlled delamination materials are used for controlled sealing deployed in these packaging methods. An e-health architecture which is developed on the basis of radio frequency identification tags for establishing medication control and audit, over an IoT HC network has been proposed in [94].

• 4. Wheelchair Management: Researches around the globe have been working on creating a fully automated smart wheelchair that would be really helpful for the people who are physically disabled. IoT can prove to beneficial in building this solution. IoT based smart wheelchairs for our healthcare system have been proposed in [85]. This design comprises of WBANs that’s laced with multiple connected sensors, which eventually provide all the functionalities suited for IoT requirements. IoT in combination with P2P i.e. peer to peer network has been implemented to develop a health support system in [95]. This approach supports functionalities including the status detection of the wheelchair user and also the functionality to control the chair’s vibrational levels. Intel also has an IoT department that has developed ‘connected wheelchair’ which is an IoT driven wheelchair [96]. Developments like this only go on to strengthen and emphasize the fact that normal things around us have the budding potential to be evolved into data driven machines that are constantly connected. This device is loaded with useful features like constant measuring and monitoring of the wheelchair patient’s body vitals. It also has the capability to gather data about the surroundings of the user in order to determine the accessibility of the patient’s location. In the forthcoming sub-sections, we are going to be looking at the IoT healthcare services and applications that we have seen so far from different perspectives.

3.3 Sensors and Devices in IoT Healthcare 4.0 Applications

We have already discussed IoT applications and services. In this sub-section, various IoT applications domain along with the sensors involved are discussed. Table 3 can be referred for a brief summarization of these devices and sensors. Some of these application domains along with the sensors used, sensing mechanism, communication protocols used, and challenges faced are described as follows:

• 1. Health Monitoring (Blood Pressure Sensor): BP measurement is the measurement of the pressure exerted against the walls of the arteries, by the blood that is flowing through it. If the blood is flowing normally through the artery, BP measurement comes out to be normal. However, if there is any kind of restriction in the blood flow, owing to various reasons, the BP measurement goes high. High BP is known to be the cause of various health problems [97]. Sphygmomanometer is the name of the device that helps us to measure the BP. The BP is also of two kinds. One is the systolic blood pressure and the other is known as the diastolic blood pressure. The BP at the time of the heartbeat is systolic BP, while the one that’s measured between the two consecutive heartbeats is known as diastolic BP. There exist various BP measurement techniques.

  Mercurial sphygmomanometer is considered to be the standard sphygmomanometer. It comprises of a glass tube and a mercurial reservoir in conjugation. Another device is Aneroid sphygmomanometer, which is quite similar to a Mercurial sphygmomanometer, the only difference being that it utilizes mechanical dial, in lieu of mercury, to show the BP. Oscillometer principle for calculating the BP forms underlying principle, on which the/ digital manometers work. It utilizes the electrical pressure sensor that measures the BP which is then displayed digitally. The drawback in the different methods that have been discussed so far is that these all involve a cuff being used. This is not feasible in a pervasive constant monitoring environment, where the monitoring would be required even when the patient is deep asleep.

  To circumvent this drawback, there are other feasible methods that exist for constant BP monitoring. One of them is called PTT or Pulse Transit Time [98, 99]. PTT is defined to be the time latency that exists between the 2 arterial sites. PTT’s estimate is usually determined by the time difference that exists between the distal arterial waveform and proximal arterial waveform. BP and PTT are both related in an inverse proportionate manner. PTT measurements that are measured in milliseconds are then finally calibrated to BP measured in mm of Hg where mm stands for millimeter and Hg is the scientific name for Mercury. BP measurement using these methods is thus without cuffs and also can be connected to smartphones.

• 2. Medical Care Units (Light sensor): MCUs have a lot of parameters to be taken care of and maintained. Illumination is one such parameter. IES or Illuminating Engineering Society recommends that the light intensity levels in the hospital premises needs to be maintained to be in range of 10 lumens per square foot to 20 lumens per square foot. Heightened light intensity levels can hinder with the nighttime sleep of patients and also leads to unnecessary power wastage. Light sensor or LS is used in the luminaire along with the occupancy sensor or OS [104]. This luminaire functions as our light intensity control system. The sensors involved here are all connected via networking, to the centralized controller. OS is used as it can identify if there are any people present in its vicinity. LS is used to add the daylight illumination and luminaire’s illumination to calculate total light intensity in its proximity. The information thus obtained by LS and OS is sent to the centralized controller. The controller is driven by a light intensity governance and control algorithm, which takes the input from OS and LS, and determines the increment/decrement amount required in the current artificial luminaire’s intensity. The determined calculations are eventually sent back tour artificial luminaires which then use those calculation for light intensity enhancement/reduction.

• 3. Fitness Devices (Wrist Band): The technical advancement in the domain of healthcare sensors has taken long strides ensuing the production of wristbands that are commercially vended today and worn by masses. Most common ones are the fitness trackers. They have the ability to determine the wearer’s physical activity levels, measure and monitor their health vitals such as heart rate, pulse rate sleep pattern, calories burnt etc. [105]. These devices also offer some health advices to users regarding the physical exercises to be practiced to maintain good health, dietary nutrition etc. [106]. Fitness trackers is perfect example of how healthcare sensors have been used efficiently in noninvasive healthcare monitoring. Moreover, these trackers can be connected to smartphones where data is aggregated over period of usage and data analysis when applied on that data can provide even a more detailed information about the person’s health status and other trends.

Table 3 Summarization of different sensors used in various application domains; communication protocols used & challenges faced

3.4 Relevant IoT Healthcare 4.0 Services from Industry

The growth and rapid advancements witnessed in use of IoT in healthcare has given the IoT industry a well-deserved boost. This has led to the established corporates around the world along with new age, edge cutting tech-startups in the direction of developing healthcare products. Table 4 is a brief summarization of these solutions. The table has the services offered, provider company’s name, product developed along with its description.

Table 4 More Relevant Known IoT Health Solutions from Industry

RHM products are of utmost benefit for the patients who are required to be monitored constantly, but at home and not admitted to some hospital. RHM monitors the patient’s health status constantly, collects that data and then transmits it to the health workers team which can then analyze that data remotely. Pourhomayoun et al. [107] in his approach has suggested that RHM frameworks possess the capability to effectively decrease the patient re-admission rates in the hospitals. EarlySense is an RHM solution that follows a wholesome approach [108]. It continuously measure the heart rate of the patient along with his/her respiration rate. It is well equipped to prevent patient’s fall. It senses deterioration, if any, in patient’s health status and also prevents ulcer formation under pressure. The solution looks promising but comes with a drawback of not being easily available globally, with its offices restricted to the territories of Israel and United States of America.

‘Falling’ has been found to be a major reason for lethal injuries in old-aged people [109]. Fade App has been developed aiming to solve this problem [110]. It’s simplistic in nature, user friendly and addresses this AAL issue in a rather sophisticated manner. A study conducted in 2010 in the USA, it was found that of the total 21,649 fatalities that occurred that year due to people ‘falling’, 5,402 was the figure related to population above the age of 65 [109]. The mushrooming population of the aging people has only stressed the importance of health apps like Fade in AAL. As per Global Health Observatory, in the year 2015, the life expectancy that’s estimated at birth averaged to be 63 years. Some devices have been developed specifically as solutions for AAL. ‘BeClose’ is one such solution that makes use of various sensors in combination with the power of AI, eventually leading to the elderly people getting more freedom [111]. Wearables is another category of healthcare solution that has been trending recently. Wearable fitness trackers and watches are being used by around fifteen percent of its total healthcare consumers. Sale of approximately 110 million such wearables is estimated for the year 2018 [112].

The aim is to provide the patients with quality healthcare service and maintaining their comfort levels simultaneously. Hence, the major factors considered while analyzing and comparing the various wearables are ease of usability, security and patient’s comfort. So, even though ‘MC10’ is good as it puts to use modern sensors, that are capable of collecting user’s physiological data, the smartwatch by Bittium is seen as a better solution [113]113. This smartwatch has the ability to measure user’s health vitals, track his location and monitor body ergonomics. It can also monitor the dosage and timing of medication. Moreover, it also provides security as a key feature as all data communication between the cloud and device sensors are secured using reliable connectivity solutions. Apple watch is another popular wearable, that has seen tremendous commercial growth and popularity [115]. The device has a lot of health-apps associated with it, making it feasible to measure the user’s BMI, estimated glomerular filtration rate (eGFR), body surface area and similar other measures required as a parameter to detect cardiovascular diseases presence. This wearable is so prevalent that in 2016 only, around 6 million units were traded successfully, underlining the importance and popularity of smartphones-based health solutions in recent times.

As per Saúde Business, an estimated 100 million medical health apps are known to exist worldwide [116]. The diabetic population in 2014 has increased drastically up to 422 million, a compared to 1980 when it was 108 million. As per WHO, this trend will probably continue and grow considerably. For diabetics, ‘On Track Diabetes’ solution is considered to be a blessing [117]. It’s an application, developed specifically for diabetics with ‘type-2 diabetes’. This app helps these users to manage the disease in a well-equipped manner, by regularly being able to perform glucose level checks, BP checks, pulse rate check and body weight measurement checks among others.

3.5 Research Work Classification Based on Condition & Services/Application

Islam et al. in their research has stated that IoT based patient monitoring systems can be classified as either services or applications [118]. Services are basically generic in nature and serve in application creation. Applications on the other hand are used directly by the patients and doctors. Table 5 is a summarization of various works in IoT healthcare by different authors, classified for different disease and application/service based.

Table 5 Different studies surveyed and categorized by Disease and Application/Service Type

3.6 Classification Based on IoT implementation with Data Analytics & Machine Learning

In Table 6, we have the classification of various research work by authors how have tried to implement IoT based healthcare models, which can eventually be deployed to be of critical use in the healthcare sector. Research works have been categorized on multiple parameters including if they can be used for emergency aid or not, what all technologies are involved, IEEE or other standards followed, multi-device support available or not and implementation of other critical and useful technologies such as ML and DM.

Table 6 Comparison and Contrast Analysis of different IoT healthcare models

• A. Emergency Aid: IoT in healthcare focusses on data and also on extending support in case of crisis and emergency situations. In such cases, the healthcare system should be able to trigger signals such as alarms thus informing the healthcare workers and emergency services.

• B. Technology: RFID, 4G, 3G etc. are some latest technologies supported by IoT. With the help of these technologies only, can data about the patient’s health status be transmitted to the fog nodes/cloud for storage or further processing [172]. Different health care models developed using IoT can be compared on this basis.

• C. Standards: IEEE 802.11/b/g/n, IEEE 802.15.4, IEEE 802.15.6, ZigBee, WBAN, and NL 7 etc. are some of the protocols or different standards that can be used for building our IoT health system.

• D. Multi device support: Efficient systems are those that have the ability to support multiple devices including RFID sensors, smartphone sensors, wearables etc.

3.7 HealthCare 4.0 with Bigdata

Big data has proven to be of great benefit in the health domain by predicting diseases, curing pandemics, saving people’s lives and improving the overall health quality of the population, in general [173]. Bigdata’s role in health sector has been shown in Fig. 9. The heterogenous data or the data that consists of structured and unstructured data, is collected by the system. This kind of data may be generated by IoT devices deployed for healthcare. The data thus collected is further consolidated and visualized in FC. Information is also preprocessed by applying techniques of data cleansing, missing values removal, data extraction, which is followed by analysis, the results of which are then transmitted to the concerned health care workers.

Fig. 9

Bigdata Architecture for IoT Healthcare

The data thus collected is voluminous in nature which makes it challenging to store it and analyze it in Bigdata environment. Different studies have been conducted in this area to find solutions to this problem. Table 7 lists the proposed work and their outcomes with their use cases. Moreover, the advantages and drawbacks of bigdata with healthcare have been summarized in Table 8. Manogaran et. al. [174] in their research have come up with a novel architecture to process as well as store the data coming in from sensors in healthcare apps. Meta Fog-Redirection with Grouping and Choosing architecture is the name of the developed system. The architecture functions as an ecosystem for IoT bigdata and governs the data against any trespassing activities. IoT devices. The data collected by IoT devices if first sent to FC and analyzed. If there is an emergency detected, doctors are alerted. Data after necessary filtering is sent to CC where Apache HBase and Apache Pig are utilized for storing bigdata.

Table 7 Big data summary in H-IOT

Table 8 Related work in Bigdata in IoT Healthcare 4.0

In RHM, managing bigdata is also a challenge. Another challenge of big data analysis is patient prioritization that exists in the field of telemedicine. Kalid et. al. [175] co-relates the prioritization of patients to RHM by the use of big data. Telemedicine applied in treatment of long-term diseases such as cardiovascular diseases, BP problems etc. requires the patients to be monitored uninterruptedly. Sensors are used for gaining Realtime status updates about patient’s health information, following which telemedicine provides care service with help of healthcare staff. One issue that one might face with telemedicine is its scalability as we witness a surge in patients count. As the count multiplies, the prioritization overhead may become too large to be of practical use. In order to counter with this possibility, the study has taken the age factor of the population as well as disasters into consideration. Prioritization has been designed to happen under three different umbrellas; transplant, operations and operation room. Triage is the technique that determines chronological ordering among patients, depending on complexity of health issue.

Firouzi et. al. [176] in his study reviewed the IoT based healthcare system that consisted of sensors and integration of bigdata ecosystem. The system included (a) BSN that records the patient’s physical and physiological information. (b) FC layer for information pre-processing, transformation of protocols from one to another, data mining and other such tasks. (c) CC and bigdata ecosystem to store the incoming data and perform predictive analysis that might be able to find some trends in patient’s health information status in the long-term.

Soaring number of outpatients in hospitals, requires more man power of health and wellness workers to diagnose and treat them. Hue et. al [177] came with a new system called Simultaneously Aided Diagnostic Model or SADM that can help in reducing the workload on the doctors. The process begins with data collection. In data collection, outpatients’ medical history, treatments taken in the past, related results and costs are all accumulated. This is followed by preprocessing of collected data, feature extraction techniques application, machine learning algorithm execution. Eventually a diagnosis is formulated and shared with the doctors. For instance, with the help of neural networks and support vector machines, SADM can perform classification of hyperlipidemia. Bigdata is generally stored in multi-dimensional arrays known as tensors based on attributes’ hierarchy at the abstraction level.

Sandhu et. al. [178] in their research have come up with a data-mining solution which is based on tensor and granularity computing approach. This solution addresses the problem of abstraction, different data formatting and data privacy. Granularity computing is a technique to solve problems at diverse granularity levels and then extracting important data from them by eliminating unimportant data in the dataset. The approach works in three different phases. Data matrix being the initial phase is responsible for storing raw data including text, audio and video into the health data tensor. In the subsequent phase, the data matrix is spread out on diverse granular levels using attribute hierarchies. In the final phase, the granules are utilized for processing the queries and deriving the results. The results thus gathered are evaluated and were found to provide faster processing and economic when compared with CANDLEINC or PARAFAC2. The dataset can be transformed to tensor format with the help of toolbox in MATLAB [179].

So far, we have seen references to various sensors and wearables in IoT. The next section delves deeper into wearable technology and provides an in-depth insight into it. Wearables are actually a quintessential example to realize how the masses in general are already becoming a part of and contributing to the IoT healthcare ecosystem. Furthermore, the Table 9 presents the outcomes of the proposed work from the references of recent state of-the-art.

Table 9 ML application summary in H-IoT

4 Wearable Body Area Networks

Wearables are major pervasive devices in IoT that are common to be observed in our daily lives. Smart clothes for an instance, along with smart watches and other wearables are endowed with optimal data processing power, which when integrated with the customer centric services of IoT can be miraculous as far as our healthcare systems are concerned. Health and fitness, tracking and safety are some domains that have been monopolizing the markets of wearables. The confluence between the digital world and our actual real world is being boosted by our wearable devices. This boost naturally brings more population into the IoT domain. The quality of life for human beings cannot be improved by smartphones alone but, is feasible when smartphones are connected with wearables as wearables have the capability to sense and assemble corporeal data. Wearables execute many minuscule jobs including verification of the incoming messages and also urgent computation of important information in a natural & acceptable manner. Wearables provide multiple value-added services. Some examples are psychological and physical health monitoring financial payments indoor localization and sports analytics [180,181,182,183,184,185,186].

As per market statistical analysis, the annual shipment of wearables is expected to hit 0.2 billion by the year 2019 [187, 188]. Wearable technology having a market value of approximately 19.6 billion $ in 2016 is expected to rise rapidly and is estimated to hit 57.6 billion $ by the year 2022, i.e. a 3X growth [189]. The main reason behind this surge in wearables’ popularity is that it has completely revolutionized the manner of interaction between the consumers and their surroundings. As much as 74% of the people accept that the wearables have been helpful in supporting their communication with the physical things surrounding them. Hereafter, it’s expected that 33% of the population will wear at least five wearables from the year 2020. 60% of the people also expect that in the coming five years, wearables will be utilized only for tracking people’s health information.

In the near future, people will be equipped with multiple internet- connected devices, and would facilitate the users to communicate with their surroundings. To enable the users to accomplish this along while receiving information in continuous & secure way would require these wearables to evolve rapidly. Smartwatch segment is estimated to grow exponentially by the year 2021, with approximately 81,000,000 units expected to be sold, constituting 16% of the total sales of wearables. As per Gartner report, wearables’ worldwide shipment is expected to increase by a margin of 25.8% annually. At this rate it would be reaching 0.225 billion $ by the year 2019. Figure 10 shows a graphical representation of the worldwide wearable sales statistics in various regions from the year 2015 to 2017, along with estimates of year 2020–2022.

Fig. 10

Wearable Devices sales statistics across globe for years 2016 and 2017, with estimates from 2020–2022

It’s concluded from the graph that the wearable device count having 4G network connectivity is approximately 217 million in North America. By 2022, it’s being expected that we might see these numbers getting doubled, thus depicting the high demand of these devices across the globe. WBAN stands for wireless body area network that is defined as wireless communication network that’s ideally suited for healthcare and aids the constant monitoring of people’s health [190]. Sensors endowed with wireless connectivity are either embedded to the internals of patients’ body or attached to the externals itself. Data thus collected by the sensor, is sent to the suitable servers, over a wireless network that’s low powered, via a gateway. WBAN popularity continues to soar as the variety and count of these wearables continues to expand and becomes economic. WBANs in healthcare can be used for a variety of purposes such as distribution calculation of a patient’s body temperature or detect his/her fall. Thus, AAL as discusses earlier, is made possible via WBANs [191]. WBANs assist in patient’s rehabilitation by presenting data that indicates an improvement/deterioration in patient’s health status.

WBANs constantly monitor and analyze the data collected from patient’s vitals regularly and publishing emergency notifications in case an anomaly is found. This leads to an early diagnosis of any underlying health issues that the patient might be suffering from. Remote telemedicine is another healthcare field that’s supported by WBANs. All the data that’s congregated via WBANs can be utilized in upcoming researches to design new healthcare solutions and cater to effective information delivery to medical sector stakeholders [192, 193]. The enabling technologies for BAN include: Wearable devices, energy harvesting, ultra-low power wireless communication network protocols [194], Wireless Sensor Networks (WSN), 6LoWPAN.

4.1 Technologies Consideration in WBANs

There are multiple prevalent requirements, design and technology considerations in wireless communications, that can also be applied in WBAN healthcare applications. The requirements are categorized under the following sub-headings and represented in Fig. 11:

Fig. 11

Classification of IoT based WBANs requirements

A. Security: These days, the smart systems cater to many diverse applications. The speedy reconditioning of the information technology is facilitating the developers to achieve authentication as a common measure for all applications. As the patients’ healthcare data is by default, sensitive and confidential in nature, our smart systems must have access control protocols with stringent measures for data quality and its security in place [195]. Doctors can monitor the health status of their patients, remotely via WBANs. Nonetheless, the major characteristic of any IoT environment is still system’s privacy and security. BSNC or Body Sensor Network Care is an example of healthcare network proposed by Gope and Hwang in their research [196]. The BSNC was proposed to secure patient’s confidential health data. The researchers classified security of the framework overall into 2 components: one being the network security (NS) and other is data security (DS). NS includes authentication protocols and anonymity. DS on the other hand comprises data privacy, integrity and freshness. Hence, all these measures assist the proposed system in accomplishing both goals i.e. data and network security.

This approach eventually covers mutual authentication, anonymity, secure patient information management, countering forgery attacks and decrease computational overhead. Another approach targeted at providing security & uninterrupted connectivity to medical systems for patient’s data, known as CCN or Content Centric Networking was proposed by Lal and Kumar in [197]. The authors aimed at building a healthcare system driven by economic and low power consuming devices, eventually contributing to an improvement in the quality of lives for patients’ ailing from various diseases. The approach provides uninterrupted connectivity, lowered latency, improved data rates, security and improved speed but faced some challenges including signal collisions in network. Boukerche and Renin [198] had also proposed a model known as the Novel Trust Evaluation Model entrusted to grant security that prevented malicious nodes amidst data transmission. Assets in a multicast arrangement were all secured by this proposed approach, followed by system’s overall performance assessment.

(a) Information security: For any smart healthcare system, it is essential that it provides data that is reliable and secure. Novel authentication and key agreement protocol by Iqbal et al. [199] is another security protocol that ensures communication channel’s security, medical devices and remote servers’ security. This protocol employs a lightweight sensor which is resource curbing and is ideally suited to secure the confidential health data. Hence this proposal has an added advantage of the devices being both small sized and cheap, for providing security to health data. Novel Anonymous Authentication is a novel approach proposed by Wu et al. [200] that provides anonymous authentication mechanism as a security aspect for our healthcare application. In addition, it also successfully lowered the processing time and costs for communication. The costs were proven to be reduced as much as 31.58%. Thus, the system performance so achieved, are better than the prevalent methods.

(b) System privacy: System confidentiality and privacy can be viewed as synonymous aspects. Privacy includes both physical privacy as well as the access control rights. It’s about implementing privacy principles on the user’s personal data. While confidentiality compels the doctors to keep the inappropriate private health data of their patients. A secure and privacy assuring data congregation methodology has been proposed by Ara et al. [201]. Portable digital assistants have been used as the intermediate medium to transmit the data collected from the WBAN’s sensors that constantly sense and collect the health data of patients, to the remote servers. Power along with storage and privacy are some of the system’s complexities. Techniques for the aggregation of the data were introduces for the purpose of decreasing the overhead caused due to these complexities. Hence the system is now empowered with improved optimizations, an increase in its efficiency and security in transmission of data. Shen et al. [202] proposed an approach called as Secure sensor association and key management protocol, aimed at providing integrity and confidentiality for patients in the healthcare system. The sensor system was equipped with a hash function and elliptical cryptography for authentication and key-generation. The advantages offered were that the algorithm proposed to offer confidentiality, security and privacy in healthcare were easy and efficient to execute, thus reducing the computational cost as well as the cost incurred due to communications.

B. Energy: The IoT devices in healthcare have limited power. These devices employ a power saving feature to preserve the power. Reading of sensors is not required at the time of device switching. The power consumption and the lifetime of the network have a deep relation between them. A major challenge in healthcare is the quality of patients’ monitoring in WBAN. To deal with this challenge, it becomes necessary to reduce the power consumption levels of the coordinator node, which would eventually lead to an improvement in the system’s quality. Coordinator node’s major responsibility is information packaging and transferring to the base station and congregate the sensor data to the base station. If the battery is not replenished by the patients regularly, it can prove to be hazardous to their health. It is better to charge the coordinator node’s battery regularly in order to enhance the overall system’s lifetime. Hence an energy efficient security protocol is required that’s built keeping the battery constraint property in mind. Wu et al. [203] proposed an optimal self-governing WBAN execution technique in WBAN healthcare systems.

The work concentrates on limited lifetime of WBAN sensor nodes and the idea of exploiting and harvesting the solar energy which can then be utilized as a power source that drives our sensor nodes. In this manner, the nodes’ lifetime gets extended and patient monitoring can be performed in a continual manner. Another approach known as ANT + protocol was proposed in Mehmood et al. research which to lower down the energy drainage in WBANs [204]. The software framework introduced by the author assimilates ANT + protocols with WBANs to provide healthcare services. Hence it can be said that the proposed approach accomplishes increased power storage, extends the life of devices, supports patient’s uninterrupted health monitoring and decreases the battery replacement requirement’s frequency.

Omeni et al. [205] in his work has laid out a medium access method that’s been optimized to save energy drainage issue and decreases the nearby mode’s collisions. The algorithms are dodged via a channel evaluation algorithm that senses whether the channel is clear. The algorithm’s underlying basis is the listen before transmit standard. Wakeup withdrawal time has been introduced on the approach to avoid those timings overlap. Hence the proposed system achieves reduction in power consumption and collisions and reduced complexity in healthcare applications. Hoang et al. [206] has proposed another approach for keeping tabs on power consumption limits. He has advocated the use of thermal energy harvesting to implement the WBAN using warmth from the human body. The author has described the concept of surplus power in the sensor nodes due to gateways and how the WBAN’s life expectancy can be enhanced by recharging the gateways via thermal energy. This approach thus, reduces the energy drainage and complexity of the system while increasing its reliability.

(a) Power: Power is a factor of utmost importance in e-health 4.0 system. In IoT, the devices are expected to deliver diverse functionalities like transmission mode, sleep mode and receive mode. These power requirements can become challenging to the communication layer. Hybrid data compression scheme or HDCS was proposed by Deepu et al. [207]. On the basis of energy requirements, this scheme helps in the categorization of the power. It is meant to reduce power consumption in IoT wireless sensor networks. Data compression was explored in lossy as well as lossless methods. Eventually it allowed the hybrid transmission mode with features like ‘selection of data rates’ and then conserves power in the wireless transmission mode. Hence the hybrid transmission mode in the proposed approach, paves a way for transmission that’s power-aware, enhanced tolerance towards errors and efficient use of local storage. Ultra-low power and traffic adaptive MAC was proposed by Ullah et al. [208]. It aimed at lowering the power consumption needs and reducing the traffic challenges as idle listening or overhearing challenges. This technique reduces costs and power consumption and improves reliability.

C. Ubiquitous Healthcare: The population is getting more mindful for their health than ever and good health is becoming everyone’s main concern these days. The ubiquitous or pervasive healthcare has had a great impact on mobile application with U-telemedicine that were integrated in IoT WBAN in healthcare monitoring. Medical services, information technology and telecom technology when integrated, is capable of providing medical services such as disease diagnosis and its treatment without any constraints of time and distance. Smart e-healthcare gateway proposed by Rahmani et al. [209] introduced intelligence in the existing IoT pervasive healthcare systems. Through this intelligence, it provided multiple hi-tech services that include localized storage along with real-time health status monitoring and embedded data mining. The system proposed decreases ambiguities in data and increases reliability and tenacity of the system. It also enhances the time and space coverage and enhances data quality. To manage the gateway, command set’s expansion generated a complete data. To make diverse nodes having different priorities interoperable, various protocols sockets associated with the transport layer have been listed in a generic library. Wang et al. proposed SNT i.e. Sensor Network Tasking that’s been developed on the basis of pervasive healthcare systems [210]. It gathers patients’ data and provides data-based probabilities. It also provided a relational model between the key indicators for information collection & antecedence in WBAN. Hence the proposed approach achieves better priority management and gain in utility, reduced power loss and sequenced information consolidation. Chung et al. [211] introduced a new approach for ECG systems in pervasive healthcare. It provides a paradigm for the transmission of all the data collected by the sensor nodes by measuring the physical data to the server nodes without any kind of data loss in between. Though the system was proposed at small-scale, its sensing and communication were precise, and it was compact and lowered the power consumption levels.

D. Resource management: Management of resources is a critical aspect in the healthcare system. Only via a proper resource management can good quality healthcare be provided. This is the reason why the domain of resource management in healthcare requires more R&D. The management approaches of healthcare resources require a better output in order to make healthcare accessible for everyone around the globe. Bhatia and Sood [130] have proposed a temporal information analysis technique in their research for a smart ICU monitoring system. It optimizes the process of real time monitoring. The proposed work goes on to accomplish precision monitoring, decrements in fatality rates and real time health status monitoring. But it also suffered from some disadvantages including issues with uninterrupted data transmissions and inefficiencies in network loads among others. Jeong and Shin [212] in their research work have proposed a mobile vehicle implanted with IoT healthcare devices. The vehicle would then serve as healthcare services providing facility for the public. As a result of the proposal, the vehicle functions as a lifesaving utility, capable of carrying the patients to nearest medical facility in case of emergencies and even could call the emergency services automatically.

Wireless sensor network for monitoring and alarming was proposed by Al-Aubidy et al. [213] which catered to patient health status monitoring. As per the criticality of health status, it kicks off the alarms that work in real time. The monitoring is offered using a rich GUI for the users. The work was proposed keeping in view the patients suffering from critical or life-threatening diseases. The system proposed offers several advantages over others such as economic and well-informed decision-making capabilities, live monitoring over patients’ health and data scanning over a reliable communication network.

E. Quality of Services: QoS is another parameter used in healthcare systems that is real-time critical. Various concerns are widespread that create difficulties in achieving high quality in IoT systems in terms of latency, consolidated data quality, resource consumption and optimizing energy requirements. It is the quality of the sensors attached to the patients’ body that is deterministic of the precision and sensitivity of the measurements sensed by these sensors. The parameters against which the quality of BSNs is measured is often fixed based on the data quality that’s provided as a reply to a query. Some of them are:

• Latency and its variations: It is a reference to latency and possible variation since latency that can occur in the process of data consolidation from the nodes.

• Throughput & Bandwidth Capacity: It demonstrates a sensor’s network capacity to transmit the information on a link, in a given time constraint.

• Precision: Precision may vary from one sensor to another and hence so does each of their capability to determine the actual measurements precisely.

• Trustworthiness: When multiple sensors are providing data for the same area, the data sets thus provided can intersect with each other. Identification of an efficient sensors subset that must be kept continuously active to satisfy the subscription requirements is a decision that needs to be made depending on parameters including sensor precision levels, amount of trustworthiness and current battery capacity. Each of these parameters can be utilized for the quantitative assessment of the healthcare data.

There are many prevalent parameters that help insure our IoT environment, but these measures are applied measures and hence can assist in better results output. Also, it helps in achieving high reliability, security, confidentiality, signal improvement among other advantages. Kisseleffet et al. [214] had proposed ‘Distributed beamforming’ for building a BSN that is based on the concept of magnetic induction. The system offers multiple advantages including an improvement in QoS and data-rates, while simultaneously reducing the susceptibility against signal attenuation. ‘Signal to noise ratio’ was increased by this system. Satija et al. [215] in their research have proposed a quality monitoring framework for signals being generated and transmitted in real-time and have termed the system as ‘ECG telemetry system’. This method was able to classify the gathered ECG signals and is both light-weight and power efficient. Therefore, it can be concluded that this system offers higher reliability and an improved life-expectancy when compared to other existing systems in the same domain of research.

F. Real-time wireless health monitoring: A health monitoring system that’s wireless and works in real-time is a research domain that’s gaining a lot of attention recently. Such a system will assist the health workers to provide timely assistance to the patient, if and when required. Velrani et al. [216] have authored a study which introduces AHS i.e. Automation healthcare System that provides healthcare services to the users comprising health monitoring in real-time and some security protocols as well. It is a model built on WBAN based healthcare system in IoT. The model utilizes wearable tags for monitoring in real time. So, the main advantages of this model can be listed as cost reduction in healthcare, increased security features and small size for easy handling, which makes it a great deal specially when there is a dearth of medical staff. For providing security as well as privacy to the user’s data, P2P protocols have been implemented in this model. Hand hygiene is an important aspect that needs to be monitored, which otherwise can prove to be the cause of various diseases in the patient. Bal et al. [217] in their research have introduced a sensor-based monitoring system for hand hygiene. The system calculates health hygiene of individuals in real-time and check if it is compliant with the health standards being followed at the medical institution where this system has been deployed. This system is not only scalable but also comes with the advantage of easy installation. Compared to other existing studies in this domain, this system looks more promising than others as it offers proper and systematic usage of health facilities.

4.2 IoT Healthcare 4.0 Industry Trends in Wearables

Advancement of IoT in healthcare domain has surged significantly in recent times and his has given a boost to more and more entrepreneurs venturing in this technical space. New age startups are interested in this research area and have developed various wearable solutions that are now available commercially and offer different benefits that can be useful for the users. Some of the popular ones are represented in Fig. 12 and mentioned below:

• 1. Motiv Ring: This product boasts of bringing fitness together with form and function. Among the fitness trackers available in the market, this product is one of the smallest and still is loaded with multiple features for its size. It is capable of measuring person’s vitals like calories burned off, heartrate, ‘resting heart-rate’, total steps walked every day. It monitors the heartrate patterns and movement patterns of individuals and then determines a metric that shows the person the manner in which the individual’s daily activities are contributing to his/her health in general. Compared to generic wrist worn wearables category, this product is worn as a ring on fingers and thus is proven to be of more comfort to the wearers. Along with it you have a mobile app called as ‘Motiv’, that allows to set-up your own individual goals such as step-count, sleep time and sleep duration. The device along with the app detect and provide you with the accuracy with which you are able to achieve those targets set by you. For every physical activity that a person performs, he/she can see the details associated with that particular activity such as calories burnt in that activity alone or max heart rate achieved in that activity. The app is compatible with all iphones featuring iOS 9.0 and above and all android devices featuring android 6.0 or higher.

Fig. 12

Wearable IoT devices categorized by body parts

• 2. BloomLife: This product has been built with the pregnant women as their target consumers. It is a wearable that is worn around the stomach with the help of a strap and helps the pregnant ladies to keep a track of all kind of contraction of muscles in the uterus. The wearable is a small sized pod with the sensors built in. The device is comfortable enough to be worn for as long as the individual wants, without any signs of discomfort. As the device has been designed specially to monitor the contractions, it is of specific use when the woman is in her final trimester of pregnancy. The device is helpful as it maintains historical data of the woman’s contractions along with measuring the intensity and frequency of these contractions, eventually guiding the women to get a fair understanding of their pregnancy progress. It also helps them determine if what they are experiencing is not actual but false labor pain also referred to as, ‘Braxton Hicks contractions’ or BHC. These BHC are common and are expected in most pregnancies and they feel quite similar to labor. The difference is that BHC are not as frequent and intense as the actual labor contractions. Bloomlife is there to help the women identify, when it’s gas or HPC or actual labor contractions, thus reducing their stress related to identify these differences themselves.

• 3. Doppel Wristband: This product is the resultant of application of research work accomplished in the fields of neuroscience and psychology respectively, that present the natural and intuitive manner in which human beings generally react to various diverse rhythms. Research has proven that a positive, relaxed and calm emotional state of mind is associated with a slower rhythm tempo, while other emotions like excitement and joyfulness is associated with a relatively faster rhythm. Doppel produces rhythm similar to our heartbeat rhythm which to us human beings, is the most natural rhythm that we can relate to. Our brain in turn, feels this rhythm and responds to it in the same manner in which it reacts to music. The only difference is that unlike music which can be distracting at times, Doppel’s rhythm does not produce any sound. One is able to just feel the rhythmic vibrations thus generated and feel calm and composed with an improvement in focusing power.

• 4. iThermonitor: This product makes use of ‘ambient compensation technology’, is worn on the person’s axilla and is used to measure the temperature of the core accurately. Monitoring is constant and the data is transmitted with a periodicity of 4 s. Bluetooth is used for the data transmission. The product features wireless adaptors that are compatible to be installed with the existing patient monitors. It is capable to monitor the patient’s body temperature in real-time without causing any disruptions in the ongoing workflow, neither does it require any changes to be made to the existing infrastructure. This product’s precision to measure the temperature is at par with the invasive options available. Hence it can be considered to be used in oust the need for spot-checks and complex probation for positioning of invasive options. And unlike other alternatives, this product has an added advantage that the patient can wear it to his/her home after being discharged from the hospital and sill be monitored for body temperature checks. All these wearables have been summarized in the Table 10.

Table 10 Wearable IoT Devices Classified by Body Parts

So far in this study, we have discussed that how the fog systems have been largely successful to process huge chunks of datasets locally, are portable in nature and are easily installable even if the hardware is heterogeneous. These are the features that are responsible for making fog systems as developer’s choice when it comes to deploying latency sensitive applications. For instance, IoT devices are devices that are generally required to process huge datasets within given time constraints. Such a varying functionality range in applications exposes the security threats to data, virtualization and the network in general. In the next section, we determine the impact of these security issues and possible solutions, providing future security-relevant directions to those responsible for designing, developing, and maintaining Fog systems.

5 System Security and its Importance in Healthcare 4.0

The computational framework and the data storage ability, both are centralized in nature in cloud computing paradigm and hence, vulnerable to be hacked by invaders. Amazon and Google – two of the largest cloud vendors in the market, have reported unfortunate instances of massive data leaks from their respective companies. Hence security is one of the major obstructions, that’s been blocking the growth of cloud computing adaption over the years. Fog computing, which is termed as cloud computing’s significant extension, is lauded as a better architecture than cloud, in terms of security. This is mainly due to two reasons. One is that the data that’s collected is briefly maintained on the fog nodes, that are nearest to the data generation sources. This reduces the reliance on web connectivity. Data’s analysis, transfer and storage when localized, makes it challenging for intruders to be able to access user’s personal data.

Secondly, data transfer between the cloud and the end devices is not occurring in real-time anymore. This makes it challenging for the intruders to detect any kind of confidential data related to some particular user. But fog computing is not certified as fully secure facility, since it is an inheritor from cloud and thus, some security risks are also inherited by fog computing in its hierarchy. There are multiple services deployed on the fog nodes for the benefit of the users, as per contract. The vendors are in a position where it’s a possibility that might be snooping on the personal information and the other data stored by the fog tenant. Fog vendors’ employees might go against the company policy and gain access to the sensitive and confidential user data and data leak can occur. Apart from this, attackers or intruders from outside may also try to target the fog nodes, for their own malicious goals. This might lead to the fog nodes themselves turning malicious, thus making the other nodes in the network, also susceptible to getting compromised. As most healthcare networks make use of FC, it makes our valuable and confidential healthcare data vulnerable to thefts and manipulations. Following is a comprehensive consolidation of the various possible attacks in the fog healthcare network [218]. Some of these security issues are highlighted in Fig. 13.

• Forgery: The attackers might malevolently impersonate the identity of the victims, produce and propagate fake data about those victims in order to misguide other bodies existing in the network. As this fake information adds to the already existing data transfers in the network, the overall data might become bulky, thus affecting the limited network resources including energy, bandwidth and storage. Bottlenecks thus produced in bandwidths can be critical in healthcare as discussed in Sect. 1.

• Tampering: The tampering attack includes data packets drop, latencies and modification of the healthcare data being transmitted in fog networks, thus leading to a disruption in services and deterioration of efficiency. Service disruptions can prove to be fatal in any healthcare system, thus tampering is a major threat to fog healthcare networks. Detection of tampering is a difficult task as some other factors such as wireless transmission channel’s current state and mobility of the users may also lead to similar transmission latencies and failures.

• Spam: It includes generation and broadcasting of undesirable and useless data like redundant information and fake data collected from the users. Spam attacks usually result in wastage of network resources, duping victim’s friends and social circles and even confidential data leakage. Spamming in healthcare networks again constricts bandwidth which is a critical challenge as health applications are time sensitive in nature.

• Sybil: The attackers carrying out a sybil attack, create fake identities or take undue advantage of pseudonyms, so that they can manipulate and compromise with fog’s effectiveness. For instance, they might be generating inaccurate crowdsensing documents such that any results derived from the docs are not trustworthy. Moreover, sybil attacks can also lead to invasion of legit user’s personal healthcare data.

• Jamming: In this type of attack, the attacker intentionally generates vast volumes of fraudulent messages so as to throttle the computational and other resources such as communication channel’s capacity. This attack renders the normal operations and computations infeasible for the legitimate users. If the attacker is successful in jamming the network, it might lead to a lot of causalities and fatalities of the patients in our e-healthcare network.

Fig. 13

Security concerns overview

• Eavesdropping: The attackers tap into the communication channels to seize the data packets under transmission and capture its contents. Leakage of patient’s healthcare data is critical as that is classified information as per the law in many countries. Patients become more susceptible to such attacks if there are no data encryption systems being used in the e-health network.

• DOS attack: The attackers make the fog nodes unavailable to legitimate users. This is done by overwhelming the fog nodes with huge superficial requests. This disrupts the fog nodes e-health services and hence users find them unavailable when required. Compared to CC servers, FC nodes are considered to be more susceptible to this DOS attack, as this attack modulus operandi is to overwhelm the healthcare network resources which are already very limited in a FC e-health network.

• MITM: An attacker with malicious intents plays the middleman by meddling with the data exchange occurring between two parties and tries to relay or even modify that data. While this happens, both the communicating parties are still under misapprehension that they are communicating between themselves only. This is a major hurdle in the private & confidential communication over the e-health network, that is supposed to be between the doctor and his/her patients and the intruder performs an MITM attack.

All the above security challenges are supposed to be dealt with, to make our fog network secure. But of these, MITM attack is of particular concern and needs to be looked into detail. Section 5.1 is dedicated to discussing the MITM attack:

5.1 Man-in-the-Middle Attack:

MITM is one of the most critical potential attack likely against our fog networks. This subsection is devoted to describe the threats posed to our fog networks due to these MITM attacks. The modulus operandi for this attack is that the gateways that have been functioning as our fog nodes are jeopardized or are replaced with malicious nodes [219]. For instance, patients who want to have an online consultation with their doctor might try to connect to some public access points that might have been maligned to function deceitfully, by falsely using the authentic identifiers. Once the gateways have been compromised by the attackers, all the confidential communications made by the patients will be stolen. MITM works in a very subtle manner in fog networks. One more peculiarity related to MITM attack is that it consumes only negligible amount of memory and processor utilization. This in turn leads to the failure of our aberration detection mechanisms, which generally detect attacks based on CPU and memory usage patterns. To highlight the stealth level of MITM in fog, the attack has been demonstrated using an environment depicted in Fig. 14. In this case study, we have patient on 3G network who tries video calling with another WLAN user which is a doctor in our case. MITM attack focusses on compromising the two users’ communication in the middle. This is carried out by compromising the fog nodes which are gateways in this case. The implementation of this attack relies on two major steps.

Fig. 14

Communication being hijacked in Fog computing

In the first step, the gateways are compromised and in the second step, a malicious piece of code is inserted into that compromised gateway. For compromising the gateways, a duplicate access point can be placed in the environment. Another method includes a ROM refresh of the existing normally functioning gateways. These methods are easy to implement in any real-world environment, where the attack is being deliberated to be carried out. The former approach is adapted in our example and for gateways, Broadcom BCM5354 have been utilized [220]. BCM5354 consists of a MIPS32 processor for delivering high performance, a USB controller and IEEE 802.11 b/g Media Access Controller/physical layer. The end points in our case: 3G cellphone and WIFI connected laptop are supposed to engage in a video communication with our gateway BCM5354 in between. We perform a ROM refresh and a system upgrade on our gateway so that it now runs on a Linux Kernel. We make changes to the TCP/IP stack in our gateways, by embedding a hook program onto it. Hook programming refers to an alteration in the system calls, by the insertion of a pieces of code into those system calls [221]. Commonly, hook’s mechanism is to swap a function pointer that points to the system call, with the hook’s pointer. The processing is then carried out and when it’s complete, the real function pointer is called. The architecture of the system can be seen in Fig. 15. Further, the gateway’s controlling requires the deployment of relevant data structures and APIs. Some examples are initialization and bootstrapping code. WLAN’s data packets are going to be sent to the associated 3G modules and processed. A 3G USB modem is plugged on our gateway. H.324 M is also implemented on the gateway for audio–video tunnel with a 3G circuit switched call setup. Adaptive Multi Rate and H.263 protocols have also been executed for the audio, video coding and decoding functionality.

Fig. 15

System design of MITM attack in Fog Computing Health Networks

Workflow of MITM attack: The gateway between the WLAN user and 3G user acts as an intermediate in order to standardize the incoming communication generated by diverse protocols. Hence, the data will be received by the gateway first, before forwarding it to the destination. MITM comprises of four steps as illustrated in Fig. 15. Beginning two steps involves the redirection of the data packets sent by the 3G user, using the hook process that’s embedded in the gateway. The data is redirected to the attacker. The attacker can then modify that data using his computer. The data is then transmitted by the attacker back to where it came from i.e. the gateway. During the last step, the data transmitted by the attacker towards the gateway, is forwarded to a WLAN user. This process goes both ways i.e. data sent by WLAN user is also first sent to the attacker and he process continues.

5.2 Security Solutions for FC e-Health 4.0 Network:

As discussed earlier, the fog network introduced between the cloud servers and users, is particularly vulnerable to various security risks. Moreover, in fog computing, standard security measures or state of the art security mechanisms do not exist, unlike cloud computing. Also:

• Fog platforms have a comparatively smaller set of available computational resources. This puts a constraint on the implementation limits of complete security solutions. Without these full security solutions, it’s difficult to identify and prevent the malicious attacks.

• Fog platforms are the favored targets by the cyber attackers. This is because fog node data throughput is in high volumes. Moreover, the probability of capturing confidential data is more in fog nodes, as they are the intermediaries between the cloud and IoT end devices.

• Based on the configuration of the network as well as the physical location, fog platforms are in general highly accessible when compared to cloud platforms. Due to this, there is a much higher possibility of attacks.

Table 11 is a brief summarization of the security challenges looming over the respective application domains, how these challenges pose a threat and their possible solutions. The table highlights the security issues in fog with respect to availability, integrity and confidentiality. As the healthcare networks use FC necessarily, for the advantages mentioned in Sect. 1, any vulnerabilities in FC, makes the patient’s confidential data and e-health service sin general at risk.

Table 11 Various security challenges in healthcare Fog Systems

So far, we have seen the major security risks associated with fog, its negative repercussions and some solutions to deal with them. Apart from the techniques mentioned so far, Blockchain is an emerging technical field that can be particularly helpful in FC health network as it can protect the confidential data. In the next sub section, we will be focusing on blockchain and how it can be beneficial in solving some of the security threats to fog.

5.3 Blockchain Technology as a Security Technique in e-healthcare 4.0

People these days can connect in many diverse manners through hypermedia. These hypermedia communications are said to be quite multifaceted, as they contain several different data formats, sources and data resolutions all combined [222]. Multimedia communication also has medical sector – a rapidly and significantly rowing field as its key component. Multimedia is responsible for providing a mechanism to enhance the communication quality between the health workers and the patients. This enhancement in interaction quality is proving to be a step forward as far as the advancements in the recovery and remedial processes are concerned [223, 224]. Multimedia usage in healthcare has started the paradigm of storage, transmission and processing of patient’s sensitive health data over to the health workers through online services. This data can be in different forms like pictures, videos, text etc. Healthcare organizations these days are striving to become more and more user-centric and efficient. Technology has been playing a vital role in today’s era, by increasing the patient’s healthcare quality standards. It also decrements its cost as it allocates all the required medical resources in a very efficient manner. IoT brings into picture thousands of devices and storage services networked together, with well-defined cooperation mechanism between them to communicate. In this bigdata can be analyzed for making well informed decisions [225]. In the present day, IoT based health service systems that comprise of multiple devices networked together so they are able to together record and share with each other important information are gaining a lot of attention. These devices communicate via a service layer that is secure and connected to central control server in CC. Moreover, IoT smart devices and sensors are coordinating together to eventually reach the goal of pervasiveness [226, 227]. These devices being interconnected provide us with various kinds of data including energy consumption patterns, current status of the devices and environmental information which can then be aggregated and finally distributed further in a secure and efficient manner. Moreover, as the devices are connected to Internet, it is easy to reach and control them remotely irrespective of the time or location constraint. Data related to healthcare is classified information that can be an attractive attack prospect for the cyber-criminals. Edward Snowden inside information leak incident has raised trust issues for the IoT development companies to trust their employees with the health data of their clientele [228]. Hence security and privacy needs special attention and needs to be put on.

the pedestal, especially when it comes to development decision related to healthcare IoT. For increasing the security for our systems, there is an urgent need of come up with some transparent approaches to develop the upcoming IoT solutions [229]. Security should be at the center or heart of our IoT network and in order to secure our patients’ health data, transparency has a key role to play.

Furthermore, in our IoT health systems as the healthcare data is recorded, produced and shared in an electronic format and also storage of this crucial data online, puts the patients’ confidential health data at high risk of potential security/privacy breach [230, 231]. Multimedia devices provided by the vendors can also be compromised by the vendor in order to take undue advantage of those devices by intentionally introducing security loopholes in them, which can later be exploited to take control of user’s sensitive information. Recently, blockchain is becoming a great topic of interest across organizations belonging to different sectors such as finance, real estate and particularly the healthcare sector [232, 233]. Blockchain technology has been witnessing a steep surge in popularity as it grants complete transparency among all individuals in the organization. Moreover, it is also capable of ensuring security as well as transparency among all the users and this stands true even in the case of IoT devices getting hacked by intruders. It is also able to bear out, organize and track the communications. It is able to do so by storing the information coming in large volumes from different devices and helping the parties to be formulated without a central cloud.

It confines all the activities and tracks the data from different IoT devices as the consignment moves between different places. For instance, if there’s a pharmaceutical firm that produces medicines, it can track its medicines starting the point of shipping to the point of delivery to medicinal stores, all with the help of blockchain. Blockchain when applied in healthcare industry in general, as shown in Fig. 16, will capture and track the intermediate activates if any, patient records and patient’s information accessibility. Moreover, the shipment of medicines in an IoT system where the products keep moving different places, can be tracked by both the receiver party and the sender party. Blockchain technology in combination with healthcare IoT in general is going to be beneficial for the society in numerous ways as it ensures the security of patient’s health data and also ensures transparency by providing the patients the rights to view their heath documents and also the power to decide who else can have access to them. Blockchain is thus an exciting approach for developing the security solutions for IoT but is still in a rather nascent stage [287,288,289,290,291,292,293,294,295,296].

Fig. 16

Blockchain Technology in IoT based Health Systems

One rather interesting prospect to be considered in IoT security solution is its scalability as the devices in our IoT health network are going to increase rapidly and continuously. So, considering this thing into the picture, strategizing a scalable security solution without any compromises in the system’s security is a challenging task in itself. Some of the undertakings in this direction have been summarized in Table 12. The table shows how different approaches use a variety of hardware platforms like Arduino or Raspberry Pi along with the security mechanism implemented.

Table 12 Various Security Mechanisms Deployed in Various Research Studies

6 Concluding Remarks

Researches throughout the globe, have all been exploring diverse technical solutions to improve the healthcare sector. These solutions should also be integration-compatible with the pre-existing services, something that can be achieved by taking advantage of IoT’s potential. This survey discusses the need of FC in e-healthcare 4.0 and its advantages over CC. This survey further explores the niche aspects of healthcare networks based on IoT and how different IoT network topologies, IoT platform and IoT architectures can be helpful in information being transmitted and received. Researches and developers will both benefit from the e-healthcare 4.0 architectures discussed in the paper as it would facilitate the readers as the initial research to venture into novel IoT health architecture development. Detailed R&D is being carried out in healthcare IoT domain, leading to its broad categorization into IoT applications & IoT services which complement each other and can be a boon for humanity. Machine Learning, AI, data mining are the technologies that are being explored to be of use in the e-healthcare 4.0 domain and serve as an interesting research area. The authors further debates how IoT is transforming the e-healthcare as we perceive it, through wearables. This survey, thereafter, takes at another essential aspect of healthcare networks in IoT and how security threats like MITM attacks can be a matter of concern. Blockchain can be helpful in dealing with these security challenges, especially those related to maintaining the confidentiality of patient’s records that are shared over the IoT network and also for medical shipment tracking. Thus, this comprehensive and exhaustive survey captures IoT healthcare in a broad horizon and would thus facilitate for further research in this domain by engineers, scientists, researchers and industry experts. The benefits that the new IoT healthcare paradigm would bring in, are going to far outweigh the issues that it brings along, which can always be dealt with, with more research going into it.