11.1 Introduction

When the internet was in its infancy the word ‘cloud’ was used as a metaphor to describe how the complex telephone networks connected. Now, many people and organizations refer to it as ‘THE cloud’ but it’s not a single entity, and it doesn’t exist in just the one place. So, what exactly is it? Cloud is a model of computing where servers, networks, storage, development tools, and even applications (apps) are enabled through the internet. Instead of organizations having to make major investments to buy equipment, train staff, and provide ongoing maintenance, some or all of these needs are handled by a cloud service provider.

There are five key characteristics of a cloud computing environment, as defined by the National Institute of Standards and Technology (NIST):

Internet Access

With a public cloud environment, users “plug into” the data and applications via an internet connection giving anytime, anywhere access.

Measured Service

Cloud is often pay-as-you-go, where you only pay for what you use. Think about how a utility company meters how much water, electricity, or gas is used and charges based on consumption. The cloud is the same.

On-Demand Self-Service

Services can be requested and provisioned quickly, without the need for manual setup and configuration.

Shared Resource Pooling

Cloud often uses the multi-tenancy model. This means a single application is shared among several users. So, rather than creating a copy of the application for each user, several users, or “tenants” can configure the application to their specific needs.

Rapid Elasticity

Cloud platforms are elastic. An organization can scale its resource usage levels up or down quickly and easily as needs change.

Wireless Body Area Networks (WBANs) are WSNs which are designed to interconnect the bio-sensor or actuators and the human body. An important example is aid for the visually impaired. An artificial retina, consisting of a matrix of micro sensors, can be implanted into the eye beneath the surface of the retina. The artificial retina translates the electrical impulses into neurological signals. WBAN can also be found in the domain of public safety where the data is used by firefighters, policemen, or in a military environment. The WBAN monitors, for example, the level of toxics in the air, and warns the firefighters or soldiers if a life-threatening level is detected. The introduction of a WBAN further enables to tune more effectively the training schedules of professional athletes. All these fields are able to manoeuvre WBAN effectively as it requires a low-power consumption because of limited capacity of the battery of each node, and also it requires low latency and a high reliability of communication.

How can a patient able to know its disease and he can take the help of a physician if he stays in another country. This mechanism also helps his time and space. Figure 11.1 introduces about the architecture and message transmission of diseases and describes about the dual sink used in the human body. By considering this, after all, current times as well people adjust it at everyday life occupations of them. Wireless sensors nodes those are embed or wearable makes WBAN inside human body on the basis of QoS (Quality of Service). The independent work of sensors to sense different human structural information is communicated through wirelessly by outside server for medical purpose [1]. Physical elements of body are observed utilizing sensors that are rate of respiratory, rate of heartbeat, blood pressure, movement of body, levels of glucose, temperature of body, and so on. These assembled body elements, one of two like short level post refined or fresh representatives, are wirelessly sent toward base/sink station for other inspection as well as refining [2, 3]. States of human body are continuously observed through the sensor nodes as well as sensing information is examined for optimal measure. Whether some element(s) are over standard (threshold) span, there is capability of sensors for sending an alert signal [4]. Thus, here, we study dual sink technique utilizing clustering inside body area network (DSCB). This is very important at improving duration of network with effectively using nodes battery time duration. Additionally, connection of nodes toward forward nodes or sink is made certain as utilizing both sink nodes. Inside this protocol, we analyzed the use of that dual sink accompanied by clustering technique. Utilizing clustering technique stables the load of network at sink nodes. At most, nodes opposed to equal cluster transmit sensed information of them as well as another data toward inherent committed sink node that is named by cluster head (CH). It ignores the traffic at one sink by comparing toward sole sink WBANs. By utilizing dual sink aids for keeping relatedness of nodes connected toward arms as well as legs of human too. By forwarding information, every sensor node chooses that finest relay nodes between individual adjacent. It depends on the distance measurement among the data transmission and sink node; then, this transmission rate is very essential for routing. Regarding to motive, calculation of cost function occurs that utilizes route, residual energy, as well as power of transmission. Calculation of SNR link occurs too as well as utilization occurs inside searching power of transmission required as sensor node. Those elements make sure adeptly use of nodes assets for improving capacity of network.

For evolution of WBAN occurs due to immense investigation. By taking an example, examination of individual’s (outdoor) corporal fitness, blood pressure, and heart attack like disease occurs through utilizing sensor, cloud, and machine learning technique [5]. Sensors (retina artificial arm chips) [6–8] inside retina can be embed for helping one blind human being for seeing once more. By using WBAN, the sufferers accompanied by heart disease [9], asthma, diabetes, Alzheimer’s and Parkinson problem, and so on can be performed [10]. Inside conventional wellness programs, there is a necessity for sufferers for staying inside hospital, yet WBAN confesses these sufferers for continuing through usual everyday schedule of them. It mitigates pharmaceutical work cost and also foundation cost. By using such technique, remote making of diseases detection occurs within sooner. Health observing structures are used for humans sanction for performing everyday ventures continuously that eventually improve life standard of them [11–13]. An important element for that common demand of WBAN is to monitor old humans’ health. The growing inhabitants are increasing within whole universe, as well as each further day protection of health cost is enhancing. Above after 50 years, proportion of aged humans is suitably going to be acquiring doubly from 10% to 20% [8, 10, 14]. As well as ratio of this retired person to labors are decreasing inside westernmost earth. Most of the human being number is increasing also. Entirely, these elements are motivating to introduce the WBAN that helps the enhancement of living wellness program. M-health as well as telemedicine [6] get many benefits for this concept.

From Figure 11.1, we know about the clustering process used in the body for data transmission. For clustering purpose we can use C-means, K-means, and optimized clustering (e.g., K-PSO) algorithm. After all, in 2001, the WBANs have concentrated at various techniques according to studies. Those are hardware as well as devices technique, network layer technique, MAC layer technique, as well as security inside WBAN [15]. By putting another way, those areas stated over are main technology areas establishing WBAN technique as well as specify which is become an interesting topic for investigators. As a key parameter of WBANs at time of investigation, there is topic called routing technique. Such routing technologies inside WBANs must take various domains as such kind of networks configuration occurs inside human body. Firstly, whether sensor’s power transmission is excessive which is harmful to human body as well as this absorbs excess battery power [16]. Secondly, as corporal classifications that sensor networks which can comfortably give rise to available space attenuation (path-loss/fading), noise, as well as interference, so bandwidth is changeable as well as control of transmission as per protocol’s limitation. Thirdly, sensor nodes are connected with anatomy which is transferable, because of body’s motion that effects within disjoint accompanied by further nodes. Fourthly, because of duration of battery limitation of sensor nodes’ methodical use called more censorious problem [17]. Possibility may occur or not for sending signals whether sensor node dries up of battery power [18].

Because of resources limitation (processing power, memory, as well as battery power), low transmission reach, as well as path-loss, one well-organized energy routing protocol is required for maintenance of WBAN [19]. Various sensor nodes are located in multiple places of a human body. The movable human structure leads to comparative variation of place can happen inside sensor nodes. By taking an example, also movable human body leads to disjoint between nodes as well as sink can happen. It gives rise to large fall rate of packet. Stable Increased-throughput Multi-hop Protocol for Link Efficiency inside WBANs (SIMPLE) is one of multi-hop routing protocol [20]. By achieving efficient energy as well as high duration of network, this utilizes route as well as residual energy elements to choose the afterward hop. Even so, by using one sink, that disjoints problem abides equal. Inside Distance Aware Relaying Energy (DARE)–efficient routing protocol, sufferers inside the hospital unit are observed for various anatomical elements [21]. It utilizes mobile sink node that located in various place of a unit. By mitigating consumption of energy, one on-body relay node located at chest of sufferer is utilized for receiving information against further nodes as well as passes on this toward sink. The on-body relay node has more energy by comparing with further sensor nodes. While locating of sink does not inside interior of human anatomy, remote nodes absorb higher energy with comparing toward neighbor nodes toward sink. Earlier classification statements give one terrible requirement to routing protocols of further traditional WSNs. For performing systematic routing inside WBANs, this should touch censorious operative needs. Those are duration of network addition, energy efficiency, classifications of anatomy, as well as management of position adjustment. Generally, duration of network addition should be regarded incorporating those classifications of anatomy for balanced performance as well as warranty of productive WBAN control. Additionally, WBANs’ integration is accompanied by cloud computing shows within starting of recent cost effectual as well as data operated structure. It aids to boost this implicit hospital topic within future. WBAN’s policies found at cloud technique possesses more edges like improved efficiency, larger performances, as well as utilities and greater reliability, and even so, this will be quite inside its advance phase as well as might possess various oppositions as well as practical problems [22]. Now, duplicate medicines are available in the market which is dangerous for health of a people. So, IoT-based apps are available in mobile phone. By using mobile phone app, we can predict which medicine is dangerous for our health [28].

11.2 Survey on Architectural WBAN

The WBAN architectural model is classified into four surfaces [7] as in Figure 11.2. The first surface (Surface 1) called BAN surface combines various wireless sensor nodes employing within one restricted physiographic region, so making one Wireless Personal Area Network (WPAN). It depends upon owned style by positioning sensor nodes at human anatomy within mode of wearable sensors stitched inside fabrics, tiny marks (on-anatomy sensor), else placed inside anatomy of human (inside anatomy sensor). Human anatomy is constantly sensed by those sensors by wanted elements as well as sends this toward one outer server directed toward another study. Nodes possess ability of local processing previously transference that is on demand basis. An upper surface one of two locally sends information gathered with the help of sensor nodes else this relays toward median coordinator named sink [23].

Signal-to-noise ratio (SNR), receiver noise figure (RNF), as well as body path-loss (BPL) are three important components that put on sensor node’s power of transmission. SNR based upon standard of transmission link. RNF component is based on gadget. Different devices give different results by it. This receiver affects BPL within utilization as well as radioactivity system [24, 25]. Various customer interactivities with gadgets on Surface 2 are there (interactivity surface customer) that importantly perform like access point (AP). The sensor nodes sense information which is sent toward treated server like pharmaceutical server located on hospital by such surface. With the basis of utilized wireless transmission protocol, Surface 2 holds various gadgets like smart phones or PDAs which are based on Bluetooth. Those gadgets gain as well as send information toward Surface 3. Due to observing of populous, by one AP equips various quarters inside home that is also attached with one wired else wireless network such as Wi-Fi [12].

Surface 3 carries decision measuring unit (DMU). This attaches with back end pharmaceutical server located inside hospital by the World Wide Web. This consequently acts every main tasks of computation. The major task of DMU is for collecting information, filtering as well as analyzing this to make decision. Surface 4 is the end surface of such architecture is called Surface 4. This supplies medical management utilities toward staffs with monitoring. The processed information with the help of DMU is communicated with remote pharmaceutical server. Inside hospital, it locates server, at which physician treated build correct conclusions at gained data. Such surface provides importantly dual various tasks called medical management services as well as urgent services.

11.3 Suggested Strategies

11.3.3 DSCB Protocol

The suggested routing strategy is presented by us in such part known as Dual Sink Software Defined Networking proposal utilizing clustering inside BAN (DSCB) that improves WBAN’s execution as applying clustering strategy between both sinks.

11.3.3.1 Network Topology

DSCB place both the sinks nodes which are called S1 and S2 between 10 sensor nodes at human anatomy. Sink nodes S1 and S2 are dissimilar with further distributed sensors at anatomy. The sink nodes possess superior assets by comparing with further sensor nodes like cell power, transmission power, and memory. Distribution of four sensor nodes occurs at anterior part of human anatomy accompanied by S1 like CH of them as well as four sensor nodes at behind of human anatomy accompanied by S2 like CH of them. A node is located at right hand’s anterior part when a node is located at behind of left hand. The waist locates S1 as well as Lumbar locates S2. With the movable part, these nodes which connected toward hands are attached to one of two of sink nodes utilizing LOS transmission. Nodes 1 to 4 as well as, perhaps, 5 united by S1 build cluster 1 at which nodes 6 to 9 as well as feasibly node 10 united by S2 build cluster 2. Whether sensing of censorious information occurs, it is forwarded straightly toward associated CH, nevertheless whether information is not censorious and this is going to chase multi-hop transmission afterward choosing one sender node between nearer nodes. Figure 11.3 shows the suggested DSCB protocol’s topology.

11.3.3.2 Starting Stage

Inside such stage [26], two sink nodes (S1 as well as S2) telecasts packets, i.e., “Hello” that carry IDs as well as places of them. On acceptance, every sensor node transmits “Reply” information that carries IDs, place, as well as residual energy. With such procedure, every node plus two sink nodes gain data regarding every further node. To find nearer nodes, this aids.

11.3.3.3 Cluster Evolution

Afterward, recognition of nearer is occurred then upcoming procedure is cluster evolution that occurs by data given before within “Hello” as well as “Reply” texts interchanged. Two sinks are known by CHs as individual clusters of them. CHs give to receive sensed information from sensors, combine, as well as forward this toward neighbor AP. Time periods location follows cluster evolution toward every cluster participants, occurred with the help of CHs utilizing TDMA protocol.

11.3.3.4 Sensed Information Stage

Activation of sensor nodes occurs just within assigned time period of them or sensor nodes possess snooze way. As sensor node turns agile, this initializes to sense information. That information sensing is inspected for cruciality firstly. If it is critical, it is sent directly to sink node; else, it is sent to sink node via multi-hop.

11.3.3.5 Choice of Forwarder Stage [26, 27]

DSCB calculates sender node with the help of link’s SNR, Distance (di) in distinction to sink, Residual energy (E_Rn) as well as Transmission power (T_p). The estimation of threshold considering SNR possesses default inside DSCB protocol that is same with “1”. Whether one sensor node does sense little information as well as which is not censorious, this chooses one node in distinction to nearby table of that. After that, numerous hops to sink are added up for selecting nearby direct toward sink. Whether numerous hops possess zero, such node is regarded by straightly attached toward sink as well as so this transmission occurs straightly with no sending node with among. Or SNR of link is computed whether less than 1, for becoming forwarder, it rejects such node as well as it records data of itself. Whether the SNR of link is greater than 1 after that node’s function of cost (FC) be about is computed. For calculating sensor node’s residual energy (E_Rn), this equation is used by us:

(11.1)

At which n indicate numerous nodes utilized, E_Rn is called as residual energy, E_In is called as initial energy as well as E_Cn is called as energy consumption node that computes the equation as below.

(11.2)

At which E_Tr as well as E_Re are accordingly, the energy consumption’s quantity with the help of node transceiver radio at the time of information’s transmission as well as reception as well as E_Cty is called as consumption of energy occurred with the help of electronic circuitry of node. The node energy absorbs at the time of setup period P_su, beginning on time 0 as well as consumption of energy occurred in every round Eⁱ_Rndj (ti) is

(11.3)

(11.4)

At which needed with the help of node at time of period P_su also in every cycle.

For finding energy value needed considering the transmission as well as reception, we use these equations:

(11.5)

(11.6)

(11.7)

At which “di” = entire distance among receiver Re as well as transmitter Tr. E_Re as well as E_Tr = consumption of energy fares for every packet with help of receiver as well as transmitter, accordingly. Likewise, E_Re-Cty as well as E_Tr-Cty = consumption of energy estimates for every bit considering receiver as well as transmitter circuitries of electronics accordingly. k = length of packet at which ∈_am = radio amplifier kind. Loss of co-efficient = l_o that distinct inside human anatomy by comparing with earthly networks; thus, Equation (11.6) is written as l_o as below:

(11.8)

(11.9)

At which P_n.T = power of transmission as wireless signal as well as β = Path-loss parameter. For finding entire distance di among some sensor node as well as nearby sensor node or sink node of itself, below equation is used:

(11.10)

(11.11)

At which F_n.C is function of cost of some node. Afterward computation of nearby F.C, it possesses noted as well as succeeding nearest F.C do computed. One nearer node within minimal F.C does choose like forwarder (Figure 11.4).

11.3.3.6 Energy Consumption as Well as Routing Stage

An important edge with utilizing both sink nodes does which sink’s more nodes undergo straight transmission span. Packet data goes sprightly bring of short delay at utilizing single-hop transmission. Inside DSCB, straight transmission happens within various instances. For few instances, this protocol chooses sender node considering routing motive. Consumption of energy with the help of sensor nodes inside multi-hop transmission indicates [18]:

(11.12)

(11.13)

(11.14)

At which, E_Tr−Mu as well as E_{Re −Mu} represent energy needed considering transmission as well as reception with help of transmitter as well as receiver accordingly inside multi-hop transmission. k = bits dimensions, di = distance among sink node as well as sensor nodes, E_Cty = energy needed considering the transmitters as well as receiver’s electronic circuit, E_am = for amplifying k numerous bits toward distance di for this energy needed, n = numerous nodes, and L_EN = energy loss at a time of communication by medium of transmission. Straight communication’s consumption of energy is [5]:

(11.15)

(11.16)

At which E_{Tr−di (k,di)} = transmission energy of straight transmission.

11.4 CNN-Based Image Segmentation (UNet Model)

Medical image segmentation is the process of automatic or semi-automatic detection of boundaries within a 2D or 3D image in internet-of-medical-things (IoTM) domain. The main difficulty of medical image segmentation is the high variability in medical images. For example, CT images contain a large amount of noise and complex boundaries. Here, we discuss an adaptive fully dense (AFD) neural network for CT image segmentation. By adding the horizontal connections in UNet structure, it can extract various features from all layers adaptively. It uses ensemble training for the output to extract more edge information in the multiple rounds training. To solve complex and large medical image data, it needs many times of up and down sampling process to extract the semantic information features of different regions. Because it is difficult to get the global information from small-scale network, it is often underfitting. Shared encoder structure to combine multiple layers of the UNet is a common way; however, even multiple layers are integrated in a same network structure, the different layers of decoder structure keep independent that cannot improve the usage of the shallow layers of features for network. From Figure 11.2, we know about the architecture of UNet briefly.

At starting, for each register node, BS records each detecting points creates cipher key. In add-on, BS records every confirm utilizers and then generates cipher keys. By keeping identification of detector node with transmitting time-stamp TS, when a sensor node registers with the base station (BS), it stores the record of sensor nodes. To give the additional safety and reliability toward several threads, BS transmits register data. All sensor nodes exist in the network will reply after getting the transmitted message from the BS, by terminating their acknowledgements. In return, this will not forward acknowledgement to BS if a sensor node will not get any message. The BS instantly retransmits the information again to the all the silent nodes. For suggested method accepted that BS is not going to keep any track. To observe patient’s physiological data and to trace and detect patients to manage drug administration and doctor’s sensor networks are associated in modern health care centers. Different uses are glucose level detectors, organ scanning, general health monitoring, and cancer detection. Inside a human body implanting wireless biomedical sensors is promising though great challenges such as ultra-safety, security, and minimal maintainability of the system are associated.

In this wireless networking for carrying out saturated detecting, a WSN is having many detecting points joined between them, which is utilized in varieties of uses like health and environmental monitoring, surveillance, and security. Permitting, detect, reply to situation in the natural surroundings, detector networking is detecting, computing, and communicating infrastructure and in our daily life. The detectors also range in less passive micro detectors such as “smart dust” as higher range like weather sensing. Their computing resources are totally separate like internet system, which appears the device and application possessed nature of these systems. A very essential way of wireless sensor network enters as for the similar set of events and it has various sensors causing sensing data. For conventional ad hoc networks as various algorithm, protocol has been given, which are not matched in special characteristic. Detector networking is a recent type of wireless networks and such as cellular networks, and MANETs are entirely separated from conventional networks. Like this conventional network, to enhance the high bandwidth efficiency and QoS, the tasks management, routing, and mobility management is executed. Under a high-level mobility situation, these networks perform superior throughput or delay characteristics.

As the battery packs can be changed as required, energy consumption is different concerned matter. Therefore, detector networks are created several points that are drawn for unaccompanied work. In MANETs and cellular networks, congestion is of statistical as compared to the multimedia biased message. The data rate is such that it is so less to 1–100 kb/sec. Aims are extending lifetime of networking system unlike conventional networks that ignore connectivity degradation through aggressive energy management as batteries cannot generally changed for operations in unfriendly or interior area. Distribution of message is primarily one directional in sensor networks from sensor points to sink point.

Enabling differences in between WSN and conventional wireless networks given below are some points of WSNs:

• Detecting point is compactly placed.
• Detecting point is liable to failures.
• Detecting point is restricted with memory, energy, and computing power.
• The topology of a sensor network can be changed frequently.
• Detecting point in WSN is of various orders such as magnitude is greater as compared to points in another conventional Wi-Fi network.
• Where maximum ad hoc networks communicate, data mainly utilize a broadcast communication prototype.
• Node is distributed in a two-dimensional space and cannot be recharged after deployment.
• Nodes are quasi-stationary.
• These are fully depending on communication interval where nodes broadcast in same ranges.
• Nodes based decisions on local information.
• It can be described utilizing GPS, signal strength or direction where nodes are location-aware.
• The energy consumption among nodes is unbalanced. Sensor network do not make any assumptions about:
  • The network length and thickness.
  • The distribution of the nodes.
  • Separation of energy consumption with in nodes.
  • Synchronization of the network.

This model and said assumptions are correct for several real networks. Sensor nodes gather their local information and transmit them to the data center in a sensor network. Regularly, the message is location-dependent, so the nodes have knowledge of their own position through GPS or by different means. Whereas, density is not known. Forwarding the image wavelet coefficients by priority is the fundamental knowledge of the suggested technique. Preserve efficiency using progressive image forwarding attempts by this technique. To organize data packages of many priorities, the wavelet image compression gives at the source. Considering picture dimension is M × N pixels and picture is degraded into r resolution level, then the 2D-DWT is constantly put to r-1 levels. Likewise, data packet priority can be carried out. To obtain each and every priority level at the descend it is optional, only for the primary level 0 which is an important portion of the image energy. We avail an image request-based scenario in this approach. Having a multiple hop transmission, the plea is set up by the main station and then transferred through the in-between nodes. In the request communication procedure, every sensor node complicated having in its memory each request parameters. The very much essential request parameters are QoI (quality of image) which hold priority levels PL and Pc, compression ratio, PSNR, and rate. When achieving a particular processing level, the requested parameters may be gained, based on the wireless implementations.

11.5 Emerging Trends in IoT Healthcare

The emerging trends in cloud computing, mobile applications, and wearable devices facilitate IoT’s role for making healthcare a smart and personalized system. Various medical devices, sensors, and diagnostic and imaging devices can be viewed as smart devices or objects constituting a core part of the IoT.

Challenges

• Constrained data processing methods
• Unique identity
• Dynamic and self-adapting self-configuring interoperability
• Integrated into information network

Constrained Data Processing Method

The process of converting raw data using medium-like manual or automatic tools into meaningful output information is as follows:

• Conversion is converting data to another format.
• Validation is ensuring that supplied data is clean, correct and useful.
• Sorting is arranging items in some sequence and/or in different sets.
• Aggregation is combining multiple pieces of data.
• Analysis is the collection, organization, analysis, interpretation, and presentation of data.
• Reporting is list detail or summary data or computed information.
• Presentation data is helpful in taking decisions.
• Characteristics of Unique Identity

Figure 11.3 tells about the fog server. The aggregated data is processed and evaluated by a decision framework and the abnormal measurements trigger an alarm system. IoT schemes have intelligent crossing points which familiarize grounded on the framework. The devices of IoT interfaces permit operators to inquiry about the strategies, display their position, and control them remotely, in connotation with configuration, control, and data management setup. IoT has increased tremendous fame in medical because of its capacity to have applications for which the administrations can be conveyed to shoppers quickly at insignificant expense. An imperative application is the utilization of IoT and cloud innovations to help specialists in giving increasingly successful demonstrative procedures. Specifically, here, we examine electrocardiogram (ECG) information investigation utilizing IoT and the cloud. The slender improvement of Internet availability and its openness from any gadget whenever has made Internet of Things an appealing alternative for creating well-being observing frameworks. ECG information examination and observing comprise a situation that normally appropriates into such situation. ECG represents the important appearance of the contractile action of myocardium of the heart. Such action conveys an exact wave that is reiterated after some phase and that addresses the heart rate. The examination of the condition of the ECG signal is a crucial problem and is the most broadly perceived procedure to deal with distinguish coronary ailment. IoT advancements permit the remote checking of a patient’s heart rate, information examination in immaterial time, the notice of therapeutic guide workforce, and experts should these data reveal possibly unsafe situations [1]. Thus, a patient in threat can be checked deprived of embarking to a crisis center for ECG signal examination. Meanwhile, experts and crisis treatment can immediately be informed with respect to cases that require their thought.

• Dynamic and Self-Adapting

  IoT schemes have intelligent crossing points which familiarize grounded on the framework. By taking the help of surveillance camera which can adjust their approaches depends on whether it is night or day. Camera could switch from lower resolution to higher resolution approaches.

IoT devices are to manage themselves, both in terms of their software/ hardware configuration and their resource utilization:

1. Energy
2. Communication
3. Bandwidth
4. Medium access

• Self-Configuring

  IoT-based instruments have self-configuring proficiency which permitting a huge number of plans to work collectively to give convinced functionality. These types of devices have capability to construct themselves, to arrange the networking, and then to fetch newest software renovations with minimal or customer intervention. Self-configuration mainly contains of the activities of neighbor and the discovery of services, organization of network and provisioning of resources.

• Interoperability

  IoT devices provision a large number of communication protocols which are interoperable and can interconnect with the other smart IoT instruments and also takes the help of the infrastructure.

• Integrated Into Information Network

  In human services space, WBAN has happened as a noticeable innovation which is equipped for giving better strategies for ongoing patient well-being checking at medical clinics, refuges and even at their homes. As of late, WBAN has increased incredible intrigue and demonstrated a standout among the most investigated innovations by human services offices on account of its essential job and wide scope of utilization in clinical sciences. WBAN includes correspondence between little sensor hubs with every now and again evolving condition, consequently bunches of issues still should be tended to. A portion of the serious issues are physical layer issues, interoperability and versatility issue, dependability, asset the executives, ease of use, energy utilization, and QoS issues. This exploration paper incorporates an extensive overview of late patterns in WBAN look into, gives forthcoming answers for some serious issues utilizing intellectual methodology, and a proposed idea of cognitive radio (CR)–based WBAN engineering. Hence, a traditional WBAN engineering can be ad-libbed to a versatile, increasingly dependable and proficient WBAN framework utilizing cognitive-based methodology. WBAN is a remote systems administration innovation, in view of radio frequency (RF) that interconnects various little hubs with sensor or actuator capacities. These hubs work in close region to, on or couple of cm inside a human body, to help different restorative region and non-medicinal territory applications [1]. WBAN innovation is profoundly refreshing in the field of medicinal science and human social insurance [2–5]. Additionally, huge commitment is conveyed in the field of Biomedical and other logical regions [6]. Also, its applications are broad in non-restorative territories like purchaser gadgets and individual diversion. A great deal of research work is experiencing on WBANs. The primary issues concentrated upon are size of system, result precision, hub thickness, control supply, versatility, information rate, vitality utilization, QoS, and real-time correspondence. WBAN hubs use scaled down batteries because of their little size. Thus, the system must work and perform in a power productive way with the goal that the existence term of intensity sources can be augmented. A large portion of the work in this specific space has been on advancement of better MAC conventions for vitality effective preparing. By and by, there are two distinct methodologies of MAC convention planning for sensor systems. Initial one is contention-based MAC convention plan. Case of this sort of MAC convention is Carrier Sense Multiple AccessCollision Avoidance (CSMA/CA). This structure has their hubs needs for channel access before transmitting information. The advantages of CSMA/CA-based conventions incorporate no time synchronization limitations, simple flexibility to arrange varieties, and versatility. The other methodology is schedule-based MAC convention. Case of this sort of convention is a TDMA based, in which time opened access to the channel is given. Henceforth, various clients get isolated availabilities for information transmission. These openings can be of fixed or variable length. Schedule vacancy controller (TSC) is utilized for giving availabilities. The advantages of this methodology are diminished inactive tuning in, over heading, and impact. TDMA-based methodology is very utilized in vitality effective MAC convention [21]. A tale approach of heartbeat fueled the MAC convention is given by [26]. This convention is TDMA-based and utilized for body sensor systems (BSNs). The work incorporates use of heart beat mood to perform time synchronization and consequently gives a vitality proficient MAC layer by evading power utilization related with time synchronization reference point transmission. Utilizing a robot to revive batteries and exchange information can drastically build the life expectancy of a remote sensor arrange. In this, the way of the robot is constrained by waypoints, and the districts where every sensor can be adjusted are featured. We utilize a blend of angle drop and a “numerous voyaging sales rep issue” look calculation to move the waypoints toward districts where sensor hubs can be revived while guaranteeing waypoints remain near one another. An auxiliary well-being remote sensor arrange (WSN) should keep going for quite a long time, yet conventional dispensable batteries cannot continue such a system. Energy is the significant obstruction to supportability of WSNs. Most vitality is devoured by (i) remote transmissions of saw information, and (ii) long-remove multi-jump transmissions from the source sensors to the sink. This paper investigates how to misuse developing remote power exchange innovation by utilizing automated unmanned vehicles (UVs) to support the WSNs. These UVs slice information transmissions from long to short-separations, gather detected data, and recharge WSN’s vitality.

11.6 Tier Health IoT Model

The protocol is intended to use as various types of communication smart protocol standards which required to proliferation the flexibility as well as interoperability of the smart system; then, it provides various types of services such as fusion, local storage data aggregation, filtering, actuation, compression, and analysis, and the model of smart e-health protocol has cast-off for fog computing prototype which deals with a hierarchical structural design and an additional reactive scheme. This acts as an intermediate module between the end-users and cloud system which accomplish the advantages by giving priority-based facilities. Figure 11.4 describes about the various types of medical sensor, connectivity, data storage, device management, data analytics, and user interface. Figure 11.5 tells about the modern e-health protocols.

11.7 Role of IoT in Big Data Analytics

Big data platform permits the incorporation and keeping of large volume and variety of information related to healthcare. Figure 11.6 describes about role of IoT in big data analytics.

This can ultimately provide the following:

• Highly configurable information incorporation alerts for realtime disease suffering engagement information customization by taking the help of parsers; in addition, this system delivers automated message analytics and propels data to patients.

11.8 Tier Wireless Body Area Network Architecture

By using cloud computing, a patient can communicate with physician and the WBAN architecture is basically a three-tier system as shown in Figure 11.7 [21].

This is composed of several biosensors which are deployed on the body. The first tier consists of body sensor nodes, the wireless communication system (devices) becomes the second tier, and the medical center or the application specific center becomes the third tier. Tier 1 consists of an intelligent node which is capable of sensing, processing, and communicating. Some sensors used are ECG sensor for monitoring heart activity, EMB (electromyography) sensor for monitoring muscle activity, consists of a blood pressure sensor and a tilt sensor for monitoring, and many other sensor. Once we collect the data of required parameters from the biosensors, it is transmitted to personal digital assistants (PDAs). Usually, the PDAs are within the transmission range of the biosensors. The PDA then transmits the data to Tier 2 which can be IEEE 802.15.6 (for implantable nodes), IEEE 802.15.4 (ZigBee), IEEE 802.11 (Wi-Fi), or IEEE 802.15.1 (Bluetooth), etc. In Tier 2, there is an interface in the WBAN sensor nodes through Zigbee or Bluetooth, ZigBee, IEEE 802, and others like Ultra-Wideband (UWB) technology, Zarlink technology, and ANT protocol (Adaptive Network Topology). Wi-Fi cannot provide timing guarantees on packet delivery, while beacon-enabled ZigBee can provide real-time communication by supporting GTS. Zigbee slow rate can be considered as a shortcoming. It is connected with the medical server through mobile telephone networks (2G, GPRS, and 3G) or WLANs—Internet. Its functions are as follows:

• Registers type and number sensor node.
• Manages the network channel sharing, time synchronization, and processing data and send data to the BS.

Next, in Tier 3, we have a set of end points (PDAs) which are linked to a mega database system where the application specific data is analyzed by the specialists in the domain. The communication in WBAN takes place via the sensor nodes. The energy consumed by these nodes tends to finish over a certain period of time; it then becomes of prime importance to restore or replace the batteries of these sensor nodes. In Tier 3, the major functions include the following:

• To authenticate users;
• To save patient data into medical records;
• To analyze the data;
• To recognize serious health cases in order to contact emergency care givers;
• To forward new instruction to user.

Therefore, we need a system which provides us with low energy consumption and maximum network lifetime. Clustering of the nodes is one such solution where the number of direct transmission from source to sink is more. Usually, the clustering approach is good for monitoring applications which require continuous sensor data stream. WBANs are emerging as a technology of great importance in the field of health-care, sports, military, and position tracking. It has a broader area of application because of its characteristics such as portability, real-time monitoring, low cost, and real-time feedback. Efficient data communication and limited energy resources are some of the major issues of WBANs. Despite the recent developments in communication technologies for WBANs, the reliability of packet transmission, especially for emergency and critical data transfer, remains a significant challenge. This may be that most of the existing techniques in WBAN use single-channel for data transmission with no intelligence. The cognitive bonded channel rovides high data rate for emergency and the demanding situation. WBAN consists of low-power sensor nodes where nodes are deployed on or inside the human body to the monitoring of various physiological parameters. It provides the daily activity of the patient and its health condition. WBAN is significantly used in medical applications and health care. In the medical online monitoring environment, WBAN provides low cost and flexibility to monitoring, patient, and medical professionals. Both healthcare and surveillance have been more modified with recent technological advances. Advances in electronics especially in communications technology and microelectronics are leading to more and more personal health monitoring and advanced healthcare products with a wide range of products that already available in our society. Various sensor applications and systems are developed with a wide range of features for heartbeats or temperature, proper insulin level, ECG, and for even wireless pacemakers. The introduction of advanced telecommunications technologies into the healthcare environment and the use of wireless communication solutions for healthcare products have led to increased user-friendliness and accessibility for users and health service providers. In the medical field, WBAN plays an important role to monitor patient health situations for early diagnoses. These sensors sense human body activity and send it to the cluster head or coordinator node. The cluster head is a high power node that collects information from neighboring nodes and sends it to the BS or doctor. Numbers of sensor nodes that can be implanted in the human body, each sensor performs its own functions such as fear detection, heartbeat, blood pressure, etc. In WBAN, some events need high data rate transmission, like in an emergency situation high data rate is required to send the patient health information to the monitoring unit. In wireless communication, CR is a transceiver which senses a frequency spectrum and is capable to concatenate free adjacent channel for high data rate. In CR, the PU, i.e., the primary user to transmit data, has the highest priority to use the channel. If PU is not using the channel and channel is free then it is allocated to the SU (secondary user). Many WBANs coexist, in which multiple WBANs communicate with medical staff for regular health monitoring. These WBANs consist of low-power sensor nodes which have a low data rate. They always rely on a single channel. Due to multiple adjacent WBANs and nearby IoT devices, the co-existence interference affects reliability and overall performance of the system. However, numerous challenges are present in WBANs and their reliability is affected by wireless sensor nodes with limited resources. Also, because of advent and advancement in sensor technology, low-power electronics, and low-power RF design have enabled the development of small, relatively inexpensive, and low-power sensors, called micro-sensors, which can be connected via a wireless network. These wireless micro-sensor networks represent a new paradigm for extracting data from the environment and enable the reliable monitoring of a variety of environments for applications that include surveillance, machine failure diagnosis, and chemical/ biological detection. There are two main challenges while designing this kind of networks, namely, communication bandwidth and energy, which are significantly more limited in this kind of WBAN network as compared to any tethered network environment of the same above maintained constraint. These constraints require innovative design techniques to use the available bandwidth and energy efficiently. In order to design good protocols for wireless micro-sensor networks, it is important to understand the parameters that are relevant to the sensor applications. While there are many ways in which the properties of a sensor network protocol can be evaluated, we use the following metrics. They are as follows:

1. Ease of Deployment

   Sensor networks may contain hundreds or thousands of nodes, and they may need to be deployed in remote or dangerous environments, allowing users to extract information in ways that would not have been possible otherwise. This requires that nodes be able to communicate with each other even in the absence of an established network infrastructure and predefined node locations.

2. System Lifetime

   These networks should function for as long as possible. It may be inconvenient or impossible to recharge node batteries. Therefore, all aspects of the node, from the hardware to the protocols, must be designed to be extremely energy efficient.

3. Latency

   Data from sensor networks are typically time sensitive, so it is important to receive the data in a timely manner.

4. Quality

   The notion of “quality” in a micro-sensor network is very different than in traditional wireless data networks. For sensor networks, the end user does not require all the data in the network because 1) the data from neighboring nodes are highly correlated, making the data redundant and 2) the end user cares about a higher-level description of events occurring in the environment being monitored. The quality of the network is, therefore, based on the quality of the aggregate data set, so protocols should be designed to optimize for the unique, application-specific quality of a sensor network.

   It is well known that cloud computing has many potential advantages, and many enterprise applications and data are migrating to public or hybrid cloud. But regarding some business critical applications, the organizations, especially large enterprises, still would not move them to cloud. The market size of cloud computing shared is still far behind the one expected. From the consumer’s perspective, cloud computing security concerns, especially data references and privacy protection issues, remain the primary inhibiter for adoption of cloud computing services.

   • ➢ Cloud computing is the delivery of hosting services that are provided to a client over the network. It is a compilation of existing techniques and technologies packaged with a new infrastructure paradigm that offers scalability, elasticity, business agility, faster startup time, reduced management costs, and just-in-time availability of techniques.

Cloud Service Models

Deployment of cloud services are based on the services such as infrastructure as a service, software as a service, and platform as a service.

Cloud Delivery Models

• ➢ Hybrid cloud: This is a mixture of two or more clouds that have unique entities.
• ➢ Public cloud: It contains all resources generally inside a company and keeps a lot of sensitive information.
• ➢ Community cloud: Cloud infrastructure is shared by several clouds, i.e., collection of several clouds. It supports a specific community that has shared concerns.

Application of WBAN can categorized depending on the domain of application. In what follows, we present major WBAN domains of application [21] in Table 11.1.

1. WBAN Application for Medical Treatment and Diagnosis There are myriad of possibilities where WBANs are useful for diagnosis or treatment of diseases. Many researchers have conducted research in this regard.
   • Remote Patient Monitoring

     Telemedicine and remote patient monitoring are the main applications of WBAN. Telemedicine means diagnosis and treatment of patients located at a remote location using information technology. WBAN has made it possible for delivery of certain healthcare services for patients at a distant location. Using telemedicine, more and more patients can be served. Body sensors collect signals from the body and transfer it to the distant physicians and doctors for processing. Doctors can use this information for health estimation for medical diagnosis and prescription. This will create a smart health care system. Daily-life activities of patients can be monitored to collect vital parameters from the human body.

   • Rehabilitation

     Through rehabilitative treatment methods, patients can restore their normal functional capabilities. Proper rehabilitation measures and therapy can enable a person, who has experienced a stroke, to function independently. These patients are constantly monitored to maintain a correct motion pattern. The main application of WBAN in this area includes sensor diversification, data fusion, real-time feedback, and home-based rehabilitation health through devices that constantly monitor bodily activities. This will create awareness regarding certain physiological activities.

   • Biofeedback

     Through WBAN, remote monitoring of human body can be done. The data collected by sensors can be accessed to gather valuable parameters from the body. Patients can look after and maintain their health through the mechanism of biofeedback like temperature analysis, blood pressure detection, ECG, etc. Biofeedback means maintaining and improving health through devices that constantly monitors bodily activities. This will create awareness regarding certain physiological activities.

   • Assisted Living

     This helps in improving the quality of life. Assisted living technologies enable elderly and disabled people to be monitored at their individual homes. This will lower the healthcare costs. Through these devices and technologies, the condition of the health of the people can be estimated appropriately.

2. WBAN Application for Training Schedules of Professional Athletes

   WBAN further enables to tune more effectively the training schedules of professional athletes.

3. WBAN Application if Public Safety and Preventing Medical Accidents

   Approximately, 98,000 people die every year due to medical accidents caused by human error. Sensor network can maintain a log of previous medical accidents and can notify the occurrence of the same accident and thus can reduce many medical accidents.

4. WBAN Application for Safeguarding of Uniformed Personnel

   WBAN can be used by firefighters, policemen or in a military environment. The WBAN monitors the level of toxics in the air and warns the firefighters or soldiers if a life-threatening level is detected.

5. Application of WBAN in Consumer Electronics

   Next to purely medical applications, a WBAN can include appliances such as an MP3 player, head-mounted (computer) displays, microphone, camera, advanced human-computer interfaces such as a neural interface, gaming purposes, and virtual reality.

11.9 Conclusion

We discussed about WBAN; here, sensor nodes are connected with our body. If any trouble happens in our body, the messages are transferred to the physician and then the physician transfers the health related issue to the specialist through cloud system. Here, we already discussed how cloud computing, machine learning, and wireless sensor network play vital role in smart health care. We discussed a densely connected encoder-decoder structure that shares the coder and integrates the decoders of different depths. The output of multiple decoders is correlated by the densely connected structure. Furthermore, we discussed about an adaptive segmentation algorithm for shallow and deep features using the UNet structure. The horizontal and vertical comparison between the two data sets verified the advantages of the model in the segmentation of complex boundaries. Experiments demonstrated that the depth feature adaptive segmentation algorithm can effectively use the information of different depths and the segmentation results generated by different depth decoders and can learn the final segmentation results from them, thus improving the accuracy of image segmentation.