Internet of things-based smart wearable system to monitor sports person health

Abstract

BACKGROUND:

The modern Internet of Things (IoT) makes small devices that can sense, process, interact, connect devices, and other sensors ready to understand the environment. IoT technologies and intelligent health apps have multiplied. The main challenges in the sports environment are playing without injuries and healthily.

OBJECTIVE:

In this paper the Internet of Things-based Smart Wearable System (IoT-SWS) is introduced for monitoring sports person activity to improve sports person health and performance in a healthy way.

METHOD:

Wearable systems are commonly used to capture individual sports details on a real-time basis. Collecting data from wearable devices and IoT technologies can help organizations learn how to optimize in-game strategies, identify opponents’ vulnerabilities, and make smarter draft choices and trading decisions for a sportsperson.

RESULTS:

The experimental result shows that IoT-SWS achieve the highest accuracy of 98.22% and efficient in predicting the sports person’s health to improve sports person performance reliably.

Keywords

IoT wearable device Internet of things-based smart wearable system (IOT-SWS)

1. Introduction to monitoring sports person health condition

IoT innovation with devices is growing increasingly to support individuals with their health information. In sports, adaptive tools could provide actual data to determine athletes’ pulse, which would aid in physical exercise [1]. IoT has possible uses in medical treatment and is commonly distributed in different implementations of healthcare applications. With the tremendous advancement of digital wearable technologies sensors, IoT increase in health care accelerates each day [2]. A more comprehensive range of technology, such as wireless communication, telemonitoring, network access, etc., has been the subject of practical IoT healthcare services [3]. Some other field in which many steps are taken involves recording, tracking, and preserving public health and property using Connected technologies, sensors.

The IoT-enabled equipment has made monitoring equipment feasible in healthcare that enables clinical health and welfare to be maintained, thereby allowing clients to follow practical diagnostic steps. IoT affects medical care with the enormous amount of patient commitment and management in the healthcare industry [4]. Besides, healthcare monitoring networks eliminate hospitalization requirements and avoid reception. Medical treatment costs are relatively lowered by the substantial use of IoT-based e-health tracking to boost therapy performance [5]. IoT’s use in health coverage is the most objective quality for different people, including clients, relatives, doctors, and sports players [6]. IoT-based e-health systems are typically restricted to tracking patient safety and clinical administration. IoT currently provides a wide variety of existing health surveillance programs with accurate tracking and disturbing services [7]. According to a series of methods, the technology has recently been used mainly in microelectronics, communications systems, data assesses, and sensor processing. The wearable devices and techniques are mainly used to track living things and preserve people’s clinical outcomes in the health environment. The standard system includes enormous sensor instruments and electronic front end tasks, making cognitive data challenging to obtain. The wearable devices have many parts, such as glass, shoes, bracelets, caps, sockets, clothing, headphones, and smartphones [8]. A development with tremendous advantages for sportsmen with connected devices healthcare systems is used to track and evaluate physical health in sports implementations [9]. The sportsman’s medical information is gathered via a collection of connected smart devices and/or sensors within the person’s body in IoT-based health service for sports purposes. Smart socks, intelligent jeans, smart-belts, BlueTooth key tracking, smart wrist bands, smart-fingers elegant tops, smart-watches, and many others are included in the portable smart devices.

With such intelligent machines, various data from athletes are collected and analyzed via IoT networks, such as heartbeat, sugar levels, intensity, the density of the blood, burnt energy, the sportsman’s distances, and numerous other physical training information. IoT-based e-health framework is a sports setting used by sportspeople that use portable smart devices to capture health data [10, 11, 12]. Each sportsman’s behavior is fully tracked and recognized by wearable equipment information [13]. The sportsman’s medical factors are thoroughly evaluated during numerous aspects, such as pulse rate, blood pressure, sugar, and oxygen concentration during the run, crash, walk, bend, lied, climb [14]. Furthermore, the sportsman’s health problems are tracked and analyzed for prone signs throughout their sufficient game time [15]. The storage methods such as database centers or computing systems allow the sensor network’s data to be processed for better analysis. The most significant advancement is a health monitoring process in the sports world [16, 17]; careful consideration and more informative data processing systems are required [18]. The Health monitoring system monitors pulse rate, blood sugar, pulse oxygen levels, and breathing can continuously change with various physical activities [19]. If certain crucial factors are not controlled and expected, a person can even die. In the sports atmosphere [20], IoT-based e-health technologies [21, 22] are an evolving area and need enhancement [23]. The wearable system is used throughout the different technologies to track the sports person’s fitness since its success defines their health.

In this paper, IT-SWS has been proposed to track sports person behavior to enhance health and record individual sports details in real-time, wearable devices. The processing of data from wearable devices and IoT technologies can help companies refine their in-game plans, detect competitors’ vulnerabilities, and make better draught decisions and sports people’s trade decisions.

The remaining work is outlined as follows: Section 2 provides insights about background studies. Section 3 discusses an IT-SWS to monitor the sports person’s health condition. Section 4 validates the results. Section 5 concludes the research.

2. Background study on monitoring sports person health condition

This section discusses several works that have been carried out by various researchers; Xiao et al. [24] developed a wearable heart rate monitoring (WHRM) system. WHRM presents a wearable heart rate surveillance device based on the Internet for a sophisticated sports bracelet analysis to track users’ adjustments in their sports heart rate. Data is then transmitted to specific server PC or mobile phones in virtual environments for tracking, management, and processing via ZigBee Wireless Detector, Bluetooth, and other communications technologies.

Zhou et al. [25] introduced Cloud Assisted Recurrent Neural Network (CA-RNN). CA-RNN proposed the wearable electromagnetic monitoring devices that help the physical therapist physicians monitor and evaluate critical rehabilitation parameters. By motion of vehicle and leg motion, the swimming mechanism is easily detected. This analysis concentrates on a multi-target optimization problem, which controls the swimming and balances the nerve cell cortex mass.

Ardalan et al. [26] discussed Smart Wearable Sweat Patch (SWSP). SWSP sensor is developed Sub-mattered dielectric material sensor sondes, and microelectronic channels included cotton threads for collecting sweat from the skin surface and bringing the work to the paper-based detecting samples. A 3D printer creates the pictures module with Ultraviolet light and an optic sensor that enables digital sensor images to be captured on-site via a smartphone. The findings believably demonstrated the SWSP sensor’s capability for massive implementations in health customization, sports quality tracking.

Butkeviciute et al. [27] proposed Mobile EEG Analysis in Sports Exercises (MASE). In connection with sports health, MASE is used for extracting movement objects from electroencephalography (EEG) signals. Clinical human signals EEG, ECG is monitored by a smart wearing Internet of Things (IoT)-based motion capturing system. Then movements are extracted with the ECG referential signal and with the Sparsity filter algorithm for pattern elimination, Baseline estimation.

Yang et al. [28] elaborated Evaluation of physiological workload assessment (EPWA) methods. The evaluation’s objective is to test three models in simulation work to estimate Work metabolism compared to indirect calorimetry. The techniques included the heart rate (HR)-Flex model, the HR-connected model, which combines HR with accelerometers (ACCs), and HR arm-leg ACC, which combines HR with ACCs used in the handles. The HR+arm-leg ACC template must be expected to be estimated for WM if possible in sensor devices.

Huifeng et al [29] developed the wearable sensor based on the IoT (WS-IoT). Thus the WS-IoT wearable sensors are introduced in this paper for sportspeople who track their fitness regularly. The purpose of WS-IoT is to describe the sports science medical centers and the success of the sporting event to encourage the use of innovation for competitors to contribute to various fields of sports. Using wearable smart sensors, health information is collected and monitored. Efficient machine learning technology is introduced for the study and tracking of sports people’s fitness.

Wen et al. [30] introduced the chemical sensing technology-assisted wearable system (CST-WS). Firstly, the production of wearable chemical sensors for sweat identification of indications is examined concerning the transduction function and structural specification in the healthcare framework. The development from a simple device structure to an integrated wearable break and saliva sensors solution is reviewed. The following care is checked through the micro-needle technology officials as an indispensable part of the closer loop sensing-therapy method.

Johnson et al. [31] developed Multidimensional ground reaction forces and moments (MGRFM). The kinematic data from wearable sensor gyroscopes can use newly supervised learning methods to analyze multilayered terrestrial response acceleration almost in realtime. Concurrent neural network (CNN) models have been equipped using data from the laboratory-based stance process and virtual sensor velocities for action and sidestep movement from almost half a million legacy motion studies. Five sensor vibrations reported during individual interlaboratory data collection sessions predicted each model.

Zhang et al. [32] proposed Deep learning-enabled triboelectric smart socks (DL-TSS). The cost-effective intelligent triboelectric socks to capture energy generated from body movements are generated for transmitting wireless sensory data. The methods are proposed to provide an optimized deep learning method with a socket signal edge framework for a gait study, developing a recognition accuracy of 15 respondents of 93.54% and detecting five individual human behaviors with an exactness of 96.67 percent.

As observed from the literature study, IoT-SWS has been implemented for tracking the sports person’s movement to enhance health and performance. On a real-time basis, the sports activities of a person are obtained by the wearable system.

3. Internet of things-based smart wearable system (IoT-SWS)

An IoT-based SWS architecture is used for continuous health monitoring in a sports environment has been proposed. Various health data are collected through different smart wearable devices. Each individual movement activity is monitored sequentially, and the related information is collected and recognized through the wearable IoT sensors, as shown in Fig. 1. The parameters analyzed on a sports person include pulse rate, Blood Pressure, blood glucose level, Presence of oxygen level, and respiratory factors. These Factors are monitored during involvement in running, walking, jumping, climbing, lying, and different sports-related activities. In their busy working times, it is essential to analyze the susceptible symptoms of monitoring. The cloud computing database has been used for storing the information collected from the wearable sensors for further data management and analysis for the classification process.

Figure 1.

Architecture of IoT-SWS.

IoT is composed of three layers: the compilation layer, the system layer and the device layer. The detector layer consists of numerous sensors and portal technology architecture, including a carbon dioxide saturation sensor, a temperature detector, a moisture detector, and the sensing devices as well as other sensing devices. The sensory element’s decision relates to the eyes, ears, nose, throat, and skin’s nerve cells. It is the Network center of items for object recognition and knowledge processing. Its primary role is the identification of entities and data collection. Data storage is a technique that enables development tools like cluster technologies, grid technologies, or spreading file systems to combine a range of storage systems across the infrastructure to provide storage space and company exposure. The core network consists of numerous secure networks, the telephone, telecommunications systems, similar to the middle of the human nerves and mind, for the sharing and retrieval of the gathered knowledge on the awareness layer. The operating system is the link between the Internet of things and the client’s experience. The network is commonly used in inventory management and distribution, smart shipping, smart construction, environmental surveillance, occupational safety, healthcare, elegant house, etc. The need for a heart rate tracking device for training refers to the vital signs tracking portion of personalized medicine, that can be broken up into medical treatment.

As is known, the health factors of a sportsman continuously result in deviation due to the different physical activities. This leads to sudden failure in the process of health monitoring, prediction, and other data analysis process. For resolving such an issue, an efficient method for collecting data from wearable IoT sensors, analyzing, and classification using Probabilistic Radial Basis Function Neural network (PRBFN) in sportsperson health monitoring has been presented. The real-time health monitoring through wearable sensors is giving data during various movements of sports activities. Here a heart rate monitoring sensor and a respiratory sensor have been considered for data monitoring. Besides, an ECG sensor has been used to convert the analog signal to a digital format for further processing. For getting more accurate values, the process of data collection through sensors has been done twice and stored in the specified data cloud through the IoT gateway. It is necessary to perform the task of regression for prediction. The PRBFN performs the regression process of the stored data. The statistical analysis for avoiding the computation defects in differentiating the abnormality and casual variation in health data due to sports activities has been performed by the PRBFN. It uses the radial basis function to approximate the data regression analysis.

The regression analysis is established from the information estimation for forecasting the irregular heart-beat that can be used for real-time tracking to prevent heart-rates differences in sports. The smart wearable system is used to collect and accomplish a real-time heart rate analysis; the pulse data is gathered. User 1 and user 2 represent the sportsperson with the wearable devices, the heart rate and the other parameters are monitored, and the person’s health condition is given to the medical service for further implementation as shown in Fig. 2.

Figure 2.

Way of monitoring sports person health condition.

The cardiac algorithm is used to analyze the information obtained via contact with Bluetooth and smartphones; data filtered to the cell phone would be transmitted. The user-specific significant clinical metrics are stored in the watch data collection terminal, primarily using heart rate measures and the amount of exercise. Heart rate can be measured and interpreted by ECG, blood pressure, pulse-wave indicators and other patterns, while ECG and heartbeat pulses are closely linked to behavior analysis and physiological strength, which can effectively be interfered with by electric impulses, electric power force or movement. Currently, no system can capture ECG and pulse wave information from the body in an extensive, fast, and accurate manner. The efficiency of ECG and pulse impulses was measured on this basis by evaluating the features of their time field. A complex adaptation filter is used to determine the optimum two-way heart rhythm as per the evaluation criterion.

3.1 Data collection

To gather actual data via sensors at various times during athletes’ actions, the IoT-based interface design template had suggested for a heartbeat management system. A wearable smart device that tracks players’ heartbeat and pulse rate is integrated into the machine to translate the pulse from the electrocardiogram (ECG) sensors and scheduled for optical signal converting the analog signal. Data collected by translating the analog signal to the digital signal was transmitted to the device by replying to the data transmission order. The reading by means of sensors is replayed twice such that the value obtained allows even more correct reading in an effort to stop conflicting results, trying to read from the sensor. The data system is known as the distributed storage service, used to gather data from various types of sensors, taking into account a particular connectivity protocol to construct a module containing meta-data.

Variables of physical fitness and data are moved through a wireless data acquisition and processing device, via Bluetooth or ZigBee. Primarily using the mobile cognitive device, the user safety variables gathered by the selection port are summarised and processed. Data are then uploaded to the central server via Cellular networks, Wi-Fi and other networking methods for data analytics and storage. The smart mobile interface uses multiple hardware solutions to satisfy the needs of specific users or groups simultaneously, as per the different device configurations. The pulse and heart rate smart device is used to collect data on the actual time, including the sportsman’s heartbeat and pulse rate. The pulse and cardiac rate measuring module built in the smart devices are efficient and accurate in collecting the data of sportspersons. The interaction pressure of the transmitter is significant, and the signal-to-noise ratio is healthy, which is not ideal for practical uses.

Figure 3.

Data collection by the wearable device.

The data is collected by wearable devices; the information is communicated by Bluetooth via mobile communication. The total heart rate of the sportsperson is collected, as shown in Fig. 3.

The raw data obtained from the wearable sensors through the IoT gateway are applied to dimensionality reduction to avoid the grouping process. The rate of the curve for the processing health information has been gathered at different positions of sportsperson during active playing. The feature extraction is being done to differentiate the abnormality, rhythmic functions, relaxing state, and cyclic functions.

Firstly, the mean value at a relaxing state before involving in sports activity has been calculated. Secondly, at the finishing form of a heavy playing activity, the features generated are extracted using the wearable devices. Finally, after the workout, the components are computed at the recovery state. PRBFN gives a linear output of approximation, control, prediction, and classification based on segmented time series. This PRBF forms an input layer, a hidden layer with activated function followed by an output layer that gives a linear output from the non-linear input of activation function. The data obtained from the wearable sensor are presented for the health information processing center. All the information or data collected is stored in a storage unit, and the communication is made by smartphone via Bluetooth device, as shown in the Fig. 4.

Figure 4.

Health information processing stage.

3.2 PRBF

The PRBF linear neurons provide the input sums. The local output neuron of the input layer is fed into the hidden layer and represented as vectors. The vector is used to classify, and it described as $X\in Y^{n}$ . The scalar part of the function in the network output $Y^{n}\to Y$ is represented, as shown in Eq. (1).

$\displaystyle\vartheta\left(x\right)=\sum_{i=1}^{R}a_{i}\rho\left({\left|{% \left|{X-c_{i}}\right|}\right|}\right)$ (1)

The scalar part of the function in the network is obtained from Eq. (1), where $R$ is the total radial neurons, $c_{i}$ is the central neuron vectors from the hidden layer for the respective neuron $a_{i}$ and $i$ with weight vectors $X\in Y^{n}$ and $\vartheta\left(x\right)$ . The distance of the Center vectors raises the probabilistic radial function using the symmetric centrifugal process. The Gaussian property of the obtained PRBF is expressed in Eq. (2).

$\displaystyle\rho\left({\left|{\left|{X-c_{i}}\right|}\right|}\right)=\exp% \left[{-\delta\left|{\left|{X-c_{i}}\right|}\right|^{2}}\right]$ (2)

The Gaussian property of the obtained PRBF is obtained from Eq. (2), Here $\delta$ is the mean value of the gaussian vector, $c_{i}$ is the central neuron vectors from the hidden layer for the respective neuron $X$ is the weight vector, and the final localized vector is indicated as follows in Eq. (3).

$\displaystyle\rho\left({\left|{\left|{X-c_{i}}\right|}\right|}\right)=\exp% \left[{-\delta\left|{\left|{X-c_{i}}\right|}\right|^{2}}\right]\ast a_{i}\rho% \left({\left|{\left|{X-c_{i}}\right|}\right|}\right)$ (3)

The final localized vector efficiency is obtained from Eq. (3), here $c_{i}$ is the central neuron vectors from the hidden layer for the respective neuron $a_{i}$ and $i$ with weight vectors $X$ . $\delta$ is the mean value of the Gaussian vector. Consider a training network of PRBF of input pairs to the target of variances like $i\left(t\right),j\left(t\right):t=1\ldots T$ , where the input pairs are $i\left(g\right)$ , and the target output pairs are $b\left(g\right)$ . The initial setup is that targeting the $c_{i}$ as the center of the considered hidden layer of the PRBF function.

The various data instances collected are randomly sampled and ten to fit the linearity of the unbiased function as the second target using the vector coefficients $\sigma_{i}$ . The minimum least square expression of the unbiased function is given by Eq. (4), and the signal to noise ratio is obtained by Eq. (5).

$\displaystyle B\left(f\right)=\sum_{t=1}^{T}B_{t}\left(f\right)$ (4)

Where,

$\displaystyle B_{t}\left(f\right)=[j\left(t\right)-\varphi\left({i\left(t% \right),f}\right)]^{2}$ (5)

The minimum least square expression of the unbiased function $B\left(f\right)$ , $B_{t}\left(f\right)$ is obtained from Eq. (4), and the signal to noise ratio is received by Eq. (5). The dependent weights limit $j\left(t\right)$ are the least square for optimal fitting accuracy $\varphi\left({i\left(t\right)f}\right)$ . As well as the accuracy, the smoothness is achieved by using the weight functions $t$ . As final, it is essential to obtain a normalized objective notation as in Eq. (6).

$\displaystyle D\left(f\right)=B\left(f\right)+\gamma S\left(f\right)=\sum_{t=1% }^{T}D_{t}\left(f\right)$ (6)

Where,

$\displaystyle S\left(f\right)=\sum_{t=1}^{T}S_{t}\left(f\right)$ (7) $\displaystyle D_{t}\left(f\right)=B_{t}\left(f\right)+\gamma S_{t}\left(f\right)$ (8)

The standard objective notation $D\left(f\right)$ is obtained from Eqs (7) and (8), the minimum least square expression of the unbiased function is represented as $B\left(f\right)$ . The smoothness is enhanced by the increasing function $D_{t}\left(f\right)$ , and $\gamma$ is termed as the parameter of regularization. The smoothness is achieved by using the weight functions $t$ .

Further, the classification has been performed by the Bayesian network with probabilistic functions. Every class has the same type of Probability Density Function (PDF), and its respective Gaussian function is noted as an ideal standard function. The covariance of such standard matrices is referred to as transverse of the same. The three units constitute the input, type, and the class, which is represented as $p, q, r$ , respectively. Using the vector of weight index, the normalization pattern of the product is given as $p=P^{t}p$ . The type obtained in the middle type layer is primarily utilized for integrating the combination and vector layer. The functions of each layer produce n identity matrix, which satisfies the class layer according to the ideal standard distribution.

The proposed algorithm is given by considering components $M_{i}$ and $N_{j}$ as the gaussian weights of the nth type for $i=1,2,\ldots n$ and $j=1,2,\ldots,n$ . The training and classification by PRBFN are as followed in Algorithms 1 and 2.

Algorithm 1: Training

Initialize

i=

n=

entire types

Set

i=i+1

Standardize

O_{ij}=\frac{O_{ij}}{\mathop{\sum}\nolimits O_{ij}^{2}}

Train

M_{i}

and

N_{j}

wrt to

O_{ij}

If M

=

N, then

O_{ij}=

Repeat train until Lj

=n

End

Where, $O_{ij}$ is the unit type used for estimating classification based on preliminary estimates. A maximum of preliminary estimates represents the testing side of the category.

Algorithm 2: Classification

Initialize M

=

0, N

=

Types

Set

m=m+1

Standardize using Eq. (7)

If M

=

=

1 then

\exp\left({m.n}\right)

Repeat until M

=

Return class

End

The non-linearity in the least-squares has been reduced using the algorithms. The Gaussian property is always as similar to the descent algorithms and solved the challenges through the consecutive tasks. Hence, the gradient descent of the performance is represented by

$\displaystyle m_{i+1}=m_{i}-\textit{GD}+m_{i}-1$ (9)

The gradient descent of the performance is obtained from Eq. (9), Where, $m_{i+1}$ is the weight function of the gradient descent (GD) with $m_{i}-1$ as the minimum weight. The input of a person $i_{1}\left(g\right)$ , $i_{2}\left(g\right)$ , $i_{l}\left(g\right)$ with the Gaussian weights $N_{1}$ , $N_{2}$ with the output of the network is shown in Fig. 5.

Figure 5.

Classification by PRBF.

4. Results and discussion

The IoT-SWS has been validated based on accuracy and efficiency in predicting sports person health to improve the performance in playing. IoT-SWS are validated by mHealth datasets (https://archive.ics.uci.edu/ml/datasets/MHEALTH+Dataset) for the collection of data efficiency. The IoT sensors gather all the information and are placed in the database for further confirmation. The fitness systems are used effectively and without hesitation to track sports people’s fitness. The classification product is calculated in conjunction with the given data function as a likelihood term. IN IoT-SWS, the ECG sensor has been introduced, which detects impulses from the sensors that would be in the player’s structure with the developed ECG Monitoring Device. The ECG ensuring data is then sent to the ECG accounting systems and is eventually measured. The classification accuracy of IoT-SWS is shown in Fig. 6.

Figure 6.

Classification accuracy of IoT-SWS.

The identification of the critical point suits the various sensitivities algorithms of multiple forms of intrusion from different characteristic point markers. If there is a minimal interruption, the indicators are essentially the same, or even other hands are produced when there is a significant interruption. In the afternoon and at night, the device receives the heart rate signals primarily due to external disturbance causes. The efficiency of the pulse is calculated by comparing the various effects of R/P wave detection of the training phase. Taking into account the computing power and the real-time device requirement, the PRBF must take both into account Precision and sophistication. A portion of ECG lead and art signaling that interact at various points are derived, while the consistency of the lead signal always remains remarkably high. The results of this operation are shown in a manner compatible with the shift in the referential heart rate pattern, while the cardiac rate fluctuated considerably after fusion. The significant irregular variations caused by intervention are stopped. The efficiency rate of IoT-SWS is shown in Fig. 7.

Figure 7.

Efficiency rate of IoT-SWS.

IoT-SWS were used by subjects to carry out the sequential purchasing experiment in everyday sports behavior. The error in the system’s cardiac rates estimation is determined according to the leading information with the maximum signal-to-noise proportion as the standard ECG signal and Heart Rate Comparison of the approximate cardiac rates obtained with the sample datasets. The error in cardiovascular estimates and describes as the mean value, the mean root value of the discrepancy between the cardiovascular estimates. An accurate measurement of cardiac frequency can be derived from ECG and stress phase impulses in static conditions. The waveform response is more stable in the playing condition than the ECG signal due to an interference shift in the electrodes touch component while moving, but this has no impact on the pulse detector. The movement to noise ratio of IoT-SWS is shown in Fig. 8.

Figure 8.

Signal to noise ratio of IoT-SWS.

The performance of the IoT-SWS is better when the sportsperson’s health-based intensity is better when compared with the ECG signal. That’s because the acceleration of the movement of the sportsperson in detection the health in the form of less error rate is implemented. The heart rate error in the combination of heart rate is lesser when used in a static state. The regular movement spectrum and the root mean square error is smaller than the failure from the ECG or pulse wave signals alone. This shows that a smart touch ring tracking method has a high right-winged rating such that the device can stilt. IoT-SWS is a kind of intelligent, portable hand ringing health tracking system based on the Internet of Things using internet technologies, embedded products, sensors, and so on to track human heart rate in the course of practicing athletes, student sports, marathons, and so on. The error rate of IoT-SWS is shown in Fig. 9.

Figure 9.

Error rate.

The actual processing of data via the monitoring devices of data can be transmitted to the central tracking portal through the Internet of things of communications infrastructure technologies from primary sign factor data such as heart rate, respiratory rate, blood pressure, etc. during sports practice. The suspicious data are alerted after data collection and examination to ensure that the cardiac condition of the sportsperson is tracked during the practice phase. The creation and implementation of the Internet-based monitoring devices surveillance system are very relevant in safe living. It is not a smart control system, but a requirement for fitness and healthy living, and the prediction level of sports person health can be obtained by IoT-SWS. The mobile application has an attractive attraction between human and machine. A sportsperson can start or finish the test or save dynamic ECG and history view on display. The theoretical framework is used for the minimum process; the cell phone application detects ECG data through background modeling pre-treating after the test has begun. The prediction rate of IoT-SWS is shown in Fig. 10.

Figure 10.

Prediction rate.

IoT-SWS achieves the highest classification accuracy with less error rate when compared to other existing methods Wearable heart rate monitoring (WHRM), Cloud Assisted Recurrent Neural Network (CA-RNN), the wearable sensor based on the IoT (WS-IoT).

5. Conclusion

This paper presents IoT-SWS to track individual sports behaviors in a safe way. The difficulties in the sports world are to play professionally and safely. Wearable devices are used to monitor images of sports in real-time. The aggregation of the wearable device and IoT technology information can allow companies to refine online campaigns, recognize vulnerabilities of competitors and make wiser decisions and sportsmen’s trading choices. The information that is needed obtained by various sensors is then generated in the database through the IoT system and then used during sports events to track fitness. The experimental result shows that IoT-SWS achieve the highest accuracy of 98.22% and the highest predicting rate the sports person’s health to improve sports person performance reliably.

Footnotes

Conflict of interest

None to report.

References

Pieles

Stuart

. The adolescent athlete’s heart; A miniature adult or grown-up child? Clinical Cardiology.2020 Aug; 43(8): 852-62.

Thota

Sundarasekar

Manogaran

Varatharajan

Priyan

. Centralized fog computing security platform for IoT and cloud in healthcare system. In Fog computing: Breakthroughs in research and practice. IGI global, 2018; pp. 365-378.

Papa

Mital

Pisano

Del Giudice

. E-health, and wellbeing monitoring using smart healthcare devices: An empirical investigation. Technological Forecasting and Social Change.2020 Apr 1; 153: 119226.

Manogaran

Varatharajan

Lopez

Kumar

. Sundarasekar

Thota

. A new architecture of Internet of Things and big data ecosystem for secured smart healthcare monitoring and alerting system. Future Generation Computer Systems.2018; 82: 375-387.

Abdel-Basset

Manogaran

Gamal

Chang

. A novel intelligent medical decision support model based on soft computing and IoT. IEEE Internet of Things Journal.2019; 7(5): 4160-4170.

Gardašević

Katzis

Bajić

Berbakov

. Emerging Wireless Sensor Networks and Internet of Things Technologies – Foundations of Smart Healthcare. Sensors.2020; Jan; 20(13): 3619.

Wang

Zhao

Yuan

Hsu

. Edge server placement in mobile edge computing. Journal of Parallel and Distributed Computing.2019; 127: 160-168.

Amin

Kumar

Biswas

Iqbal

Chang

. A light weight authentication protocol for IoT-enabled devices in distributed Cloud Computing environment. Future Generation Computer Systems.2018; 78: 1005-1019.

Veningston

Kadry

Kalash

Balamurugan

Sathiyaraj

. Intelligent social network based data modeling for improving health care. Health and Technology.2020; 10(1): 321-332.

10.

Al-Turjman

. 5G-enabled devices and smart-spaces in social-IoT: an overview. Future Generation Computer Systems.2019; 92: 732-744.

11.

Rath

. Big data and IoT-allied challenges associated with healthcare applications in smart and automated systems. InData Analytics in Medicine: Concepts, Methodologies, Tools, and Applications2020; (pp. 1401-1414). IGI Global.

12.

Kurdi

Alsalamah

Alatawi

Alfaraj

Altoaimy

Ahmed

. Healthybroker: A trustworthy blockchain-based multi-cloud broker for patient-centered ehealth services. Electronics.2019; 8(6): 602.

13.

Kumar

Krishnamurthi

Nayyar

Sharma

Grover

Hossain

. A novel smart healthcare design, simulation, and implementation using Healthcare 4.0 processes. IEEE Access.2020 Jun 25; 8: 118433-71.

14.

Winkelmann

Eberman

Games

. Telemedicine experiences of athletic trainers and orthopedic physicians for patients with musculoskeletal conditions. Journal of Athletic Training.2020 Aug; 55(8): 768-79.

15.

Alabady

Al-Turjman

Din

. A novel security model for cooperative virtual networks in the IoT era. International Journal of Parallel Programming.2020 Apr; 48(2): 280-95.

16.

Indrakumari

Poongodi

Suresh

Balamurugan

. The growing role of the Internet of Things in healthcare wearables. In Emergence of Pharmaceutical Industry Growth with Industrial IoT Approach2020 Jan 1 (pp. 163-194). Academic Press.

17.

Baskar

Shakeel

Kumar

Burhanuddin

Sampath

. A dynamic and interoperable communication framework for controlling the operations of wearable sensors in smart healthcare applications. Computer Communications.2020; 149: 17-26. doi: 10.1016/j.comcom.2019.10.004.

18.

Kwon

Kim

Lim

Mahmood

Yeo

. A skin-conformal, soft material-enabled bioelectronic system with minimized motion artifacts for reliable health and performance monitoring of athletes. Biosensors and Bioelectronics.2020 Mar 1; 151: 111981.

19.

Selvaraj

Doraikannan

. Privacy and Security Issues on Wireless Body Area and IoT for Remote Healthcare Monitoring. Intelligent Pervasive Computing Systems for Smarter Healthcare.2019 Jul 22: 227-53.

20.

da Costa

. Engineered nanomaterials in the sports industry. In Handbook of Nanomaterials for Manufacturing Applications2020 Jan 1 (pp. 309-320). Elsevier.

21.

Nguyen

Liu

Nguyen

Chou

, Cyber Security of Smart Grid: Attacks and Defenses, ICC 2020–2020 IEEE International Conference on Communications (ICC), Dublin, Ireland, 2020, pp. 1-6, doi: 10.1109/ICC402772020.9148850.

22.

Janarthanan

Doss

Baskar

. Optimized unsupervised Deep learning assisted reconstructed coder in the on-nodule wearable sensor for Human Activity Recognition. Measurement.2020; 108050. doi: 10.1016/j.measurement.2020.108050.

23.

Jung

S-S

Simon

. Chang

P-R

. Treatment of Students with ADHD without Medicines, But with Sensory-Motor Integration Training (SMI-Tx)(1) Combing Latest qEEG and Neurofeedback Come to the Shortest and the Best Results. Int J Psychiatr Res. 2020; 3(5): 1-9.

24.

Xiao

Han

. Wearable heart rate monitoring intelligent sports bracelet based on the Internet of things. Measurement.2020; Nov 1; 164: 108102.

25.

Zhou

Montenegro-Marin

Hsu

. Wearable IoT based cloud-assisted framework for swimming persons in the health monitoring system. Current Psychology.2020 Jun 14: 1-0.

26.

Ardalan

Hosseinifard

Vosough

Golmohammadi

. Towards smart, personalized perspiration analysis: An IoT-integrated cellulose-based microfluidic wearable patch for smartphone fluorimetric multi-sensing of sweat biomarkers. Biosensors and Bioelectronics.2020 Nov 15; 168: 112450.

27.

Butkevičiūtė

Bikulčienė

Sidekerskienė

Blažauskas

Maskeliūnas

Damaševičius

Wei

. Removal of movement artifact for mobile EEG analysis in sports exercises. IEEE Access.2019 Jan 1; 7: 7206-17.

28.

Yang

Forsman

Lindecrantz

Seoane

Ekblom

Eklund

. Evaluation of physiological workload assessment methods using heart rate and accelerometry for a smart wearable system. Ergonomics.2019 May 4; 62(5): 694-705.

29.

Huifeng

Kadry

Raj

. Continuous health monitoring of sportsperson using IoT devices based on wearable technology. Computer Communications.2020 Jul 1; 160: 588-95.

30.

Wen

Liu

Chen

Zhang

Lee

. Advances in chemical sensing technology for enabling the next-generation self-sustainable integrated wearable system in the IoT era. Nano Energy.2020 Jul 26: 105155.

31.

Johnson

Mian

Robinson

Verheul

Lloyd

Alderson

. Multidimensional ground reaction forces and moments from wearable sensor accelerations via deep learning. IEEE Transactions on Biomedical Engineering.2020 Jun 30.

32.

Zhang

Zhu

Sun

Shi

Zhu

Dong

Yuce

Lee

. Deep learning-enabled triboelectric smart socks for IoT-based gait analysis and VR applications. npj Flexible Electronics.2020 Oct 26; 4(1): 1-2.