A SECURED ARCHITECTURE FOR IOT-BASED HEALTHCARE SYSTEM

Healthcare has gradually moved away from the model centered on traditional health centers due to the emergence of highly accurate sensors and Internet of Things (IoT) enabled medical equipment. Ambient intelligence takes whatever actions are required in response to a recognized event in order to enable continuous learning about patient data. The capabilities of IoT-assisted healthcare services might be improved by incorporating autonomous control and human-computer interface (HCI) technologies into ambient intelligence. Major unsolved issues include the privacy and security of information collected by medical IoT devices, both during transmission to and during cloud storage. This research explores different techniques, IoT factors, and features, with an emphasis on the data security concerns connected to data flow in medical IoT. In order to guarantee data security and privacy at all data levels, this study suggests a safe design for the IoT healthcare system.


INTRODUCTION
The Internet of Things principal objective is to connect every device on the earth.Today, IoT is mostly employed in healthcare to provide instant access to information.The IoT is a network of autonomous computing devices, mechanical and digital machines, animals, or humans.It connects everything to the Internet, encourages information exchange, organizes correspondence, and enables item positioning, tracking, administration, and monitoring.It provides information technology (IT) solutions, which employ computers to store, retrieve, transfer, and modify data without requiring human-to-human or human-to-computer interaction.According to its definition, the Internet of Items (IoT) is a "dynamic worldwide interconnected network technology with self-configuring capabilities based on standard and coherent communication protocols where virtual and physical things have identities, physical characteristics, and virtual personalities".Integrating a number of promising technologies will enable the IoT idea to be realized in the real world Jeong et al. [2016] Darshan et al.
[2015] Thakre et al. [2022].IoT may be combined with identification, sensing, and communication technologies.Security concerns, in addition to heterogeneity, scalability, connection, and many other problems, are significant hindrances to the growth of the Internet of Things and must be adequately addressed if IoT is to be successful.Among other security challenges like confidentiality, integrity, etc., authentication of devices participating in the IoT is one of the more significant ones and is the main theme of this article.IoT development has now been supported by a variety of technologies, including NOMA in wireless technology, MEMS, and the Internet Thakre et al. [2022] Kinhikar, et al. [2022].A market worth more than 2.1 trillion dollars is predicted to emerge by 2025 owing to the low cost of sensor devices, resulting in more than 25 billion installed units by 2020.A market worth more than 2.1 trillion dollars is predicted to emerge by 2025 owing to the low cost of sensor devices, resulting in more than 25 billion installed units by 2020.

RELATED WORK 2.1 ARCHITECTURES
In the context of the IoT, the authors Almotiri et al. [2016] presented a mobile health (m-health) system (IoT).The use of mobile devices to obtain real-time health information from patients and store it on network servers connected to the Internet is known as m-Health.

SECURITY ENCURITY MODELS
Privacy protection has been widely researched and accepted as a bottleneck in smart medical healthcare.Goldwasser and Micali introduced a GM encryption system that proved semantic security under the assumption of quadratic residuosity (Shafi & Silvio, 1984).The ideal lattice-based encryption scheme proposed by Gentry [2009]

PROPOSED ARCHITECTURE FOR IOT-BASED HEALTHCARE SYSTEM
The system under the proposed design uses IoT sensors to gather patient data, which is subsequently sent to the hospital's cloud storage.Five parts make up the architecture.

IOT DEVICES IDENTIFIED FOR MEASUREMENT ARE
Blood glucose sensor: An opto-physiological glucose sensor, which measures blood glucose levels using a photodiode and an accelerometer, can be utilized in a non-invasive IoT system for glucose monitoring.
Temperature monitoring sensor: A Raspberry Pi-based temperature monitoring system and a planned wireless network of sensors would alert a doctor if a patient's temperature rose over a certain level.The LM35 sensor is proposed as a device for sensing the body's core temperature.
Healthcare systems for the elderly: An approach for detecting falls among the elderly and informing concerned parties is described.The system detects falls using sensor values from the accelerometer and gyroscope.These should be simple for the monitoring systems to dismiss as false positives and not register as falls.The waist is the right location to implant the sensors to detect geriatric falls, according to an examination of diverse sensor implantation sites on the body using various types of sensors and algorithms for machine learning.

Electrocardiogram heart monitoring systems:
A minimal-cost ECG system is recommended.The ECG sensor in this system is merely a data collection device.Multiple users' ECG values are collected with the ECG sensor and sent via Zigbee to a centralized server.

SENSORS IN IOT-BASED HEALTHCARE SYSTEMS
A minimal cost ECG system is recommended.The ECG sensor in this system is merely a data collection device.Multiple users' ECG values are collected with the ECG sensor and sent via Zigbee to a centralized server.The MMA7260QT accelerometer is specifically suggested for use in detecting body movement.Angular velocity can be measured along the x, y, and z axes using a device with three axes.A gyroscope can help alert medical workers when a person falls by detecting body tilt.A magnetometer aids in determining the relative direction by measuring the magnetic field.A magnetometer can detect a human fall when combined with an accelerometer, gyroscope, and other sensors, and is frequently seen in aged care equipment.Applications used: Raspberry Pi: According to, body temperature monitoring devices are created using the Raspberry Pi platform.A multispectral system that monitors body temperature, respiration rate, heart rate, and movement is developed using a Raspberry Pi as reviewed in Kumar N. [2017].

COLLECTION AND TRANSMISSION
Sensors are fastened to the patient's body to collect and transmit data to the master node in real time.
The data from the master node is received by the mediator device and sent to the cloud.The collection and storage of data in the cloud is an ongoing process.The IoT sensors based on WiFi and Zigbee have the lowest latency, as shown in Table 3.2.Material is delivered to hospital servers using gateway nodes installed in the area for the protocol conversion from ZigBee to TCP/IP.Low-power sensor nodes can now be connected to the internet utilizing the Internet protocol, owing to communication protocols like 6LoWPAN.In order to function, the IPV6 protocol needs a substantial amount of computing power and bandwidth.They foresee "always-on" activity, which IoT devices do not.The study Ge et al. [2016] discovered that delivering these signals as JSON packets rather than XML packets significantly reduces response, processing, and interpretation times.

COMPUTING AND SECURITY
In fog computing, fog nodes (which can be access points, switchers, gateways, routers, etc.) are spread at the network's edge and move toward terminal facilities at a certain location.The fog computing layer is composed of the security, storage, and monitoring layers.Fog computing converts cloud data centers into uniformly spread platforms while aiming to sustain cloud services.As a result, there is a decrease in processing time for data from wireless medical sensors, which improves customer satisfaction and service level.Because storage nodes are low and also have limited power capacity, the public key encryption method cannot be employed for security.The research defines fog computing as the addition of cloud computing to the system's edge, which is a highly virtualized stage of the source pool that delivers computation, storage, and networking resources to local end users.The results indicate that fog computing may achieve more than 89% low latency and bandwidth efficiency.[2020].In this study, a heart monitoring application was developed in which each patient was required to provide a 30-minute recording of their electrocardiogram signal to fog processing units for handling, analysis, and decision-making.The values are decreased so that less energy is utilized in both processors and networking equipment.When compared to the central cloud, the results show that energy may be saved by up to 69%.
To keep information confidential and private, it is necessary to encrypt it before it is delivered across a public channel and stored on fog nodes or cloud servers.After receiving information and instructions from the user and the cloud, they are responsible for filtering the raw data and uploading it to the cloud for long-term storage or additional analysis.

Steps for Homomorphic encryption
Step 1: The inquirer describes the data collected as a vector Z = (z1,z2,z3,.....,z8).After that, the secret eight-order invertible matrix M is generated.MZ is then computed and sent to the server.
Step 2: On the server, an intermediate vector V represented by V = (v1,v2,v3....v8) is stored.The greatest (denoted as Mx) and smallest (denoted as Mn) values of each physiological item's normal zone are likewise kept, and Li = Mx -Mn is discovered, with xi set to be 1/Li.In addition, we create matrix A, the elements that form the leading diagonal are equal to Mid, where Mid= -Mn/Li.https://doi.org/10.17993/3cemp.2022.110250.222-230 Step 3: When the medical result MZ arrives at the server, it is multiplied by V to get MZV.MZV and matrix A are then returned to the enquirer by the server.The enquirer then obtains ZV by left multiplying it by M-1.Finally, Di is obtained by multiplying matrix A by ZV.
Step 4: If Di is in the range [0, 1], the data item remains normal.If Di is more than one, the data item appears too high; otherwise, it displays too low.
Even if the attackers are capable of stealing data passing through the communications platform, such as MZ and A, the data vector Z is left multiplied by M. If the proposed homomorphic encryption mechanism is used, hackers will be unable to access genuine data because they do not know M.
Fig (3.4) shows that the proposed encryption algorithm outperforms RC5 and RC6 in terms of efficiency.Several experiments were carried out to demonstrate its speed.

ANALYSIS OF DATA
Following the acquisition, the acquired data must be handled, filtered, and compressed to eliminate extraneous information.We can use the ontology technique, as proposed in the article Kumar V.
[2015].Clinic data for patients is specified as a source with a specific URL address in the recommended approach.To enable information transmission via Ontology data access, mapping with both previously collected and newly acquired records should be carried out after the acquisition of patient records.The doctor will create a treatment strategy for the patient after considering these variables, which will include the drugs and dosages that will be used.In this instance, the relationship is represented by the Protégé tool.Run a sparql query to get data or an owl/rdf file from the database.
Because the fog nodule has a specified limited size, it cannot contain a large number of activities each second Yi et al. [2014], Kumar Y. et al. [2019].Fog computing in health care was the focus of Isa et al.
Gia et al. [2014]elligence algorithms evaluate m-health data to find trends and raise different alarm levels based on the status of the observed patients.An m-health system's information technology architecture is a component of the IoT architecture, which is multi-layered and includes data collection, data storage, and data processing layers.In , T. N.Gia et al. [2014]the 6LoWPAN architecture, which is composed of low-power wireless area networks (LoWPANs)2, which are IPv6 stub networks for IoT networks, is described.6LoWPANhasemergedasthe favored option.Sensor nodes employ the 6LoWPAN protocol stack, represented in Figure2, to transmit data to the network.The sensor data is encapsulated in a 6LoWPAN datagram and delivered over an IEEE 802.15.4 frame to the edge router or gateway.The packets are converted to IPV6 packets by the gateway and sent to the server for additional data processing through the standard IPV6 network.In Fig2depicts the endto-end architectural features of the 6LOWPAN-based healthcare system. https://doi.org/10.17993/3cemp.2022.110250.222-230

Table I .
Comparison of available wireless communication technologies for smart healthcare.