A Secure and Resilient Scheme for Telecare Medical Information Systems With Threat Modeling and Formal Verification

Telecare Medical Information Systems (TMIS) is a highly focused and unique domain providing healthcare services remotely, the development and advancement in the realm of information and communication technologies boosted the development of TMIS. Smartphones, IoT devices, Mobile Healthcare Applications (MHA) and hospital servers are the building blocks of TMIS. Emergen Research predicts that IoT based healthcare security market will reach USD 5.52 Billion in 2028. Existing IoT based healthcare solutions are facing many security problems which includes information leakage, false authentication, key loss and are not in compliance with Health Insurance Portability and Accountability Act (HIPAA) regulations as IoT devices and sensors used are prone to Blue Borne, DoS (Denial of Service), DDoS (Distributed Denial of Service) and Reverse-engineering attacks. In addition to these healthcare applications in the IoT devices/sensors and mobile healthcare applications in the smart phone of the patient are vulnerable to repackaging attacks and lacked transport layer protection. This paper proposes a SRSTMIS (Secure and Resilient Scheme for Telecare Medical Information Systems) containing its architecture, a procedure to verify the safety and security of patients credentials and Mobile Healthcare Applications (MHA) and finally proposed a secure protocol. White-Box Cryptography (WBC) ensures the safety and security of the keys in the healthcare applications and in the SE, UICC and TPM. We have threat modeled our proposed healthcare framework using STRIDE approach and successfully verified using Microsoft Threat Modeling tool 2016. Our proposed secure and lightweight authentication scheme has been successfully verified with BAN (Burrows, Abadi, and Needham) logic and Scyther tool, and our proposed protocol overcome DoS (Denial of Service), multi-protocol attack, Blue Borne attack, DDoS (Distributed Denial of Service) attack, reverse engineering, insider, outsider and Phlashing attacks. SRSTMIS overcomes information leakage from sensors during rest and during transit, key loss from healthcare applications of the sensors and smart phone and false authentication and ensures HIPAA regulations. Proposed protocol was successfully implemented in Android Studio. We have compared our proposed work with the existing works and found to better in terms of security, resisting attacks, and in consumption of resources.


I. INTRODUCTION
The rapid advances and development of information and communication technologies boosts the development of Telecare VOLUME 10, 2022 This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/ Medical Information Systems(TMIS) as it is an important domain in modern healthcare which ensures medical services remotely for critically ill patients and elderly people. TMIS is playing a crucial role in the ongoing COVID 19 pandemic as TMIS monitors the patients' health and provides treatment using Smartphones, IoT devices, Mobile Healthcare Applications (MHA) and hospital servers so these are the main building blocks of TMIS. TMIS helps in stopping COVID 19 pandemic's spread as TMIS establishes secure communication among all the entities involved in the TMIS ecosystem. The demand for TMIS is increasing rapidly due to the COVID-19 pandemic. MHAs are extremely helpful in providing an online communication platform reducing physical attendance for unnecessary appointments at the hospitals. Emergen Research predicts that IoT based healthcare security market will reach USD 5.52 Billion in 2028. During COVID-19 pandemic there was a huge rise in the cyberattacks on healthcare systems, while many healthcare systems are inefficient to defend these attacks. Medical data are often very sensitive and needs to be protected. Most of the entities involved in the healthcare framework cannot withstand cyberattacks. Mobile healthcare industry should comply with HIPAA standards in order to regulate data privacy for personal healthcare information. Authors of [27] and [28] proposes Telecare medicine information system (TMIS) frameworks which fail to ensure end to end security. Existing TMIS solutions are facing many security problems which includes information leakage, false authentication, key loss and are not in compliance with Health Insurance Portability and Accountability Act (HIPAA) regulations as sensors used are prone to DoS (Denial of Service), Blue Borne and DDoS (Distributed Denial of Service) attacks. The main targets of Intruders are device, application, data in transit and the data at rest for getting patient's data and credentials (keys) in TMIS. Healthcare applications play vital role in the success of TMIS, but these applications in the sensor and smart phone are not trustworthy as these are prone to reverse engineering attacks from the intruders/attackers. Attackers target these applications for patient's data and credentials (keys) and are often successful in getting patient's data and credentials (keys). In addition to these existing TMIS schemes has more communication and computational cost and are not practical.

Motivation a) IoT security market: Emergen Research predicts that
IoT based healthcare security market will reach USD 5.52 Billion in 2028. b) Application security market: According to market watch, the Application Security Market will cross US$ 11 billion by 2024 globally (Application Security Market) [23]. c) According to marketsandmarkets IoT medical devices are will reach USD 63.43 billion by 2023 globally (Shelly Singh) [24]. d) IoT medical devices are being used by many patients all around the globe as they make the life of patients easy and is evident from the predictions from marketsandmarkets (Application Security Market) [23], but these devices should be made secure right from the manufacturing phase of these devices which is the responsibility of the manufacturer. IoT medical devices use healthcare applications and applications need to be portable and secure, the security of these applications is the responsibility of the hospitals and the government. e) Transport layer protection is absent in the existing healthcare applications and are also prone to repackaging attacks. The remaining article's organization is as follows: In Section II, we present the related work in the realm of TMIS security. In Section III, we propose a SRSTMIS framework. Section IV provides the formal verification SRSTMIS protocol. Section V brings security analysis. Section VI presents threat modeling of SRSTMIS framework. Section VII presents the implementation and performance analysis of SRSTMIS protocol, and Section VIII provides the conclusion of the research work.

II. RELATED WORK
Reference [2] proposes an authentication scheme in TMIS based on Physical Unclonable Function (PUF) and Elliptic Curve Cryptography (ECC) technology. But this solution has no clarity a) How the ECC technology can encrypt the messages in the real time. b) How the healthcare application overcomes reverse engineering attacks? Reference [3] proposes healthcare systems which ensures privacy location, mutual authentication and with less storage and computational costs, but the application and communication security was compromised. Reference [4] proposes an authentication scheme for IoT healthcare based on cloud, but it suffers from repackaging attacks and lacked transport layer protection. Dhillon and Kalra [6] uses ECC (Elliptic Curve Cryptography) algorithm for proposing an authentication scheme for healthcare which monitors patients remotely. The main limitation of this work is medical professional or doctor will be able to access patient's data. Sharma and Kalra [4] proposed user authentication scheme which is lightweight in healthcare based on Cloud. The main contribution of this work is hospital can get the real-time data from the sensor of a remote patient and this data can be stored in the cloud server. But this work has the following limitations listed below a) Data security in the cloud is not ensured b) This work does not ensure non-repudiation and accountability properties A novel authentication scheme is proposed by Kumar et al. [7] in the realm of Wireless Medical Sensor Networks, but we have found that the proposed work is very much vulnerable to insider attack, does not ensure end to end security and is prone to off-line password guessing attack.
The research work proposed by Li et al. [8] cannot withstand impersonation and off-line password guessing attacks.
The work proposed by Wu et al. [9] is prone to multiprotocol, Blue Borne, DDOS, reverse engineering and Phlashing attacks. Salem and Amin [10] proposed a privacy protection protocol based on the El-Gamal cryptographic system to improve the medication security of patients in TMIS. However, the storage cost of this protocol is too high. Xu et al. [11] proposed a PUF-based lightweight RFID security protocol to achieve effective verification of a single tag but this work is vulnerable to desynchronization attacks and secret disclosure attacks. Most of the works discussed in this section use RFID technology with RFID tags and these are the main targets of attackers. Following are the limitations of using RFID in TMIS a) There is no clarity how the credentials are stored in RFID tags b) There is no clarity how the credentials are stored in the gateway if they are stored in the memory of the gateway they are prone to attacks. c) RFID tags can be reverse engineered and there is no mechanism to overcome these (reverse engineering) attacks d) There is no safety and security of keys in RFID tags e) RFID tags are prone to information leakage f) RFID tags are prone to false authentication attack Reference [12] proposes a Tele-COVID application which is both web and Android based telemedicine application monitoring COVID patients. Following are the drawbacks of this research work a) There is no mention how the credentials are generated and stored in the Tele-COVID application b) Tele-COVID application is prone to reverse engineering attack. c) Application security and Communication security is not ensured d) This work is not in compliance with HIPAA regulations Reference [22] proposes a block chain-based healthcare system sharing with cloud-based services the main contribution is Access control mechanism and the drawbacks are Scalability and key management. Following are the drawbacks of block chain based healthcare systems a) The size of Block chain ledger will increase as the time passes which makes the record management difficult for IoT devices such as sensors b) Sensors are resource constrained devices and block chain uses asymmetric encryption algorithms which require more processing power and time thereby consuming more battery. c) There is no central database in block chain to store patient's information so the ledger needs to be stored on the participating sensors as the block chain increases in size which makes it difficult for sensors as they have very less storage capacity. d) Block chain depends heavily on private key if the private key is compromised or lost all the patient's information is lost and moreover when the patient is unconscious doctors cannot retrieve his medical records without patient's private key. e) Immutability property of a block chain hinders the adoption of block chain technology in healthcare as patient cannot erase his own health information/ records. f) Block chain based healthcare solutions introduce latency g) Block chain based healthcare solutions are not practical to store high-volume of healthcare information or data on block chain as this is will degrade the performance. So we haven't adopted block chain technology and proposed this research work which ensures defense in depth security and is resilient, so security is ensured in all the entities involved in the ecosystem and at all the levels. Security is incorporated in the design phase and implementation phase of our proposed healthcare system.
Following are the main limitations in the existing literature a) Mutual authentication between the IoT Sensor (SUCH AS WHRM) and hospital is not ensured. b) Existing TMIS schemes are prone to information leakage, key loss and false authentication. c) Healthcare applications are prone to reverse engineering attacks. d) Existing IoT based healthcare schemes are prone to IoT device specific attacks which includes BlueBorne, DDoS attacks in IoT based healthcare. e) Most of the existing works are not practical f) Very few solutions/schemes in the existing literature were implemented in the real time. g) The communication cost and computational cost of the existing TMIS schemes are more. h) Existing TMIS schemes does not comply with HIPAA regulations. So there is a great need of secure and resilient scheme in TMIS. All the entities involved in the framework should be able to withstand, avoid and recover from attacks targeting patient's keys and confidential data in SE, UICC, TPM and MHA. TMIS framework should ensure security and safety at the 'device level', 'application level', during transit and at rest.
Novelty of our research work: The novelty of our proposed work are: a) As per our knowledge we are the first to address key loss, false authentication and information leakage issues in the realm of TMIS. b) As per our knowledge our proposed work is the only work which ensures the safety and security of keys in IoT sensors, smartphones and healthcare applications. c) As per our knowledge our proposed work is the only work in TMIS which overcomes reverse engineering attacks on Healthcare applications. d) As per our knowledge we are the first to address Blue-Borne, DDoS attacks in IoT based TMIS schemes. e) As per our knowledge we are the first to threat model our proposed SRSTMIS framework using STRIDE VOLUME 10, 2022 approach and successfully verified using Microsoft Threat Modeling tool 2016. f) Our proposed SRSTMIS framework is resilient as the intruder/attacker will not be successful in extracting and manipulating the credentials from any of the devices involved in the SRSTMIS framework. In addition to this SRSTMIS framework ensures the security of data at the device level, application level, at rest and during the transit.

Contributions made:
The contribution made by this work are as follows: a) Proposes a TMIS architecture, a procedure to verify the safety and security of patients credentials and Mobile Healthcare Applications (MHA) and finally proposed a secure protocol. b) Proposed healthcare scheme overcomes information leakage, key loss and false authentication. c) We have threat modeled our proposed healthcare framework using STRIDE approach and successfully verified using Microsoft Threat Modeling tool 2016. d) This research work overcomes information leakage from sensors during rest and during transit, key loss from healthcare applications of the sensors and smart phone and false authentication among the entities involved in the system thereby ensuring HIPAA regulations. e) Proposed secure TMIS scheme ensures confidentiality, integrity, availability, mutual authentication and nonrepudiation properties. f) Proposed secure TMIS scheme overcomes multiprotocol attack, Blue Borne, DoS, DDoS, reverse engineering and Phlashing attacks. g) SRSTMIS's energy, communication, and computation costs are far less than that of the existing TMIS research works. h) SRSTMIS is formally verified using BAN logic [17] and [18], and Scyther tool [19] and [20]. i) We have successfully implemented our proposed SRSTMIS in Android Studio.
Motivated by these solutions, we find no work till date which ensures the safety and security of keys and healthcare applications both in sensor and in patient's smartphone. Existing solutions are prone to information leakage, false authentication and vulnerable to repackaging attacks. We name our proposed framework as Secure and Resilient Authentication Scheme in TMIS (SRSTMIS). (i) An attacker has the capability to access the data stored in the memory of Sensor/Smartphone/Hospital Server. (ii) An adversary has the capability to tamper the patient's data and credentials in the ecosystem. iii Attackers has the capabilities to replay, update remove the data exchanges in the ecosystem. (iv) An attacker can also access the credentials and data sensor/smartphone of a doctor by reverse engineering.  MIS framework resists DoS and DDoS attacks as 'H' detects these attacks by change-point detection, activity profiling and wavelet-based signal analysis detection techniques. In addition to these 'H' installs DoS/DDoS protection tools such as ''Fort Guard Anti-DDoS''.

III. PROPOSED SECURE TMIS FRAMEWORK
In order to collect evidence from 'H', TPA collects the evidence from its networks, firewalls, IDPS (Intrusion Detection and Prevention System) and hospital TPM. In addition to these TPA gets vital evidence of DoS attack attempts from the logs of ''Fort Guard Anti-DDoS'' tool.
f) TPM: Both the Hospital (H) and CA use TPM. TPM adds security and integrity for the Hospital and CA's servers as it protects their credentials such as keys, tokens and ensures the integrity of hardware platforms and host Operating Systems. Hardware controller on the server's motherboard is implemented by TPM of the 'H' and 'CA', so hardware controller acts as a repository for securely storing the credentials which includes passwords, tokens, keys, and digital certificates. TPM is immune to malware and forgery. TPM creates a ''fingerprint'' of the server with its components as it boots, and comparing that baseline against periodic measurements of the system's parameters if there is any deviation it indicates that the server was compromised and the server will not boot. If the hospital/CA server boots successfully with the TPM then the server is not compromised and it can be trusted. MPKI is implemented in TPM. The combination of MPKI, WBC and TPM makes the Hospital (H) and CA servers immune to attacks. to ensure all the security properties in the proposed TMIS we need to adopt MPKI, but the implementation of MPKI in the memory of Sensor (S), TPM and smartphone is suicidal as the keys can be compromised. TPM and UICC generates and stores their credentials. j) UICC and SE: A Secure Element (SE) and UICC are tamper-resistant hardware devices with the capability to host mobile applications. UICC hosts different mobile applications in separate security domains which is controlled by the Owner of the application. UICC implements firewalls among applications which restricts mobile applications from interfering. Patient's anonymity in our proposed SRSTMIS framework is ensured using TAC (Traceable anonymous certificate) [26], UICC and SE implements MPKI and WBC which helps in ensuring the safety and security of keys and healthcare applications.  Our proposed work ensures Application security, Endpoint security and Network security. Self-Signing Restriction, Code Attestation, Control Flow Obfuscation. The main difference between Code obfuscation and WBC, Code obfuscation hides the complete variables, program, flow, but in WBC key is private which is a secret and the algorithm is public so the parameters related to the key cannot be retrieved by an attacker who is in possession of the Medical sensor and smart phone. Our proposed framework overcomes BLE vulnerabilities as our MHA's code is obfuscated by the MHA manufacturer and attested by the Certifying Authority (CA) and imposes self-signing restrictions, in addition to these Sensor (S) transmits encrypted data using the symmetric key shared between sensor's MHA and the MHA of the patient (P). Data encryption prevents MITM and eavesdropping attacks. A secure link is established between the sensor's MHA and MHA in the UICC of the patient ensuring application security (symmetric key) and communication security (using SSL/TLS). MM manufactures and distributes MHA to the hospitals and is responsible for the security of the MHAs. In the process of securing the MHAs from reverse engineering attacks which is one of the dangerous attacks against MHA, MM implements the following countermeasures: i) Logic Obfuscation: Our proposed framework adopts logic Obfuscation which prevents the attackers to know the logic of the healthcare application. ii) Control Flow Obfuscation:Our proposed framework adopts control flow obfuscation, MM reorganizes the control flow of the MHA, injects dummy code, removes functions' makes use of proxy methods to redirect the flow of execution and the process tree. iii) Self-Signing Restriction: MHAs are digitally signed only by CA; no other entity has the authority to sign an a healthcare application. C. PROPOSED SCHEME Figure 1 shows an e-healthcare architecture using TMIS, which consists of four types of entities such as Patient (P), Hospital (H), cloud server (C), Certifying Authority (CA),  generates their credentials in 'UICC' of their smartphones. All the participants generate their credentials which involves a public and private/secret keys. CA verifies the possession of private key for an equivalent public key, after successful verification of private key CA issues a certificate for that participant. All the entities in SRSTMIS has trusted storage, which helps them to securely store their credentials and MHA thereby ensuring the integrity and confidentiality of the credentials and MHA. SRSTMIS framework never allows to export the credentials and MHA to other entities without proper mutual authentication and authorization, in addition to these SRSTMIS allows only 'H' to update the MHA OTA (Over The Air) using a secure tunnel at regular intervals. Security of the keys and MHA relies on the tamper resistance nature of the SE, UICC and TPM and the WBC (White-Box Cryptography), as these devices (SE, UICC and TPM) can securely store keys, generate random numbers, encrypt messages (using both symmetric and asymmetric), perform hashing and implements WBC (White-Box Cryptography).

Resilience of 'SRSTMIS':
Our proposed framework is resilient as the entities/participants has the ability to withstand, avoid and recover from attacks to compromise the keys, confidential patient's data in SE, UICC, TPM and MHA in addition to security of confidential patient's data during the transit thereby ensuring HIPAA regulations. In our proposed 'SRSTMIS' TMIS framework security and safety is at levels i.e. 'device level', 'application level', during transit and at rest. We have used AES-256 algorithm for encrypting the messages exchanged among the participants.

IV. FORMAL VERIFICATION A. FORMAL VERIFICATION OF THE PROTOCOL 1) BAN LOGIC PROOF
BAN logic [16], [17], [18] contains many objects classified as principals, keys (symmetric, asymmetric keys and digital signature keys) and statements. These are represented symbolically as K sh , K ph and K dh are the shared symmetric keys in the SRSTMIS framework. MPKI is adopted in SRSTMIS framework so K S , K P , K H andK D represents the public keys of 'S', 'P', 'H' and 'D' and K −1 S , K −1 P , K b) 'E' sees X: The principal 'E' receives a message containing X. 'E' will decrypt the received message. c) 'E' said X: The principal 'E' believed X when it sent the message d) 'E' controls X: 'E' has jurisdiction over X. The principal 'E' is an authority on X and should be trusted on this matter. e) fresh(X): This means that ''X'' is fresh when 'X' has not been sent in a message at any time before. f) X ↔ Y: X and Y is a shared symmetric key 'K', both 'X' and 'Y' trust 'K'. g) {X } K : Key 'K' is used to encrypt 'X' SRSTMIS protocol is written in Security Protocol Description Language (SPDL); SPDL is a language for the Scyther simulation tool [19] and [20]; it verifies the security of protocols. Scyther tool defines the roles of SRSTMIS framework and all the entities involved face different types of attacks such as authentication attack, integrity attack and confidentiality attack. This tool helps in verifying, falsifying, and analyzing the security properties of SRSTMIS protocol with a unique ability to verify multi-protocol attacks. Attack model: SRSTMIS framework is implemented in SE, UICC and TPM and their credentials are generated and stored in these tamper resistant devices; in addition to these WBC is implemented in these tamper resistant devices along with MHAs ensuring the safety and security of the keys. So all the entities involved in the framework which includes 'P', 'D' and 'H' ensures end to end security.

V. SECURITY ANALYSIS
This section presents the security analysis of SRSTMIS protocol. Table 5 shows the comparative analysis of SRSTMIS with the related work.
1) Proposition 1: The proposed protocol healthcare protocol ensures confidentiality property Proof: Encrypted medical data is exchanged in SRSTMIS framework thereby ensuring confidentiality property.

VI. THREAT MODELLING
The process of threat modeling is divided into the three main phases as following: identifying assets, access points and trust levels, Identify and Rank all the potential threats and Discover countermeasures and build mitigation plan. Table 3 suggests a list of Threats and the countermeasures provided by our proposed framework corresponding to STRIDE.
(1) Identifying assets, access points and trust levels: An asset is a valuable thing which is owned by an entity of SRSTMIS framework, and the intruders/ attackers/adversaries are interested in, and wish to retrieve/access, control or delete it. This step is the most crucial step in threat modeling. Access points are the interfaces through which intruders/attackers/adversaries can interact with the system in order to gain access to assets. The next step is to identify and define boundaries of trust in the system. There are different levels of trust indicating the trust required for accessing a component from a system. The main idea of a trust boundary is that within a boundary, there is a common level of security, so within that boundary the components trust each other and will not question the integrity of each other.
List of Assets in our proposed SRSTMIS framework:Sensor, MHA in Sensor, Smart phone of the patient, MHA in Smart phone of the patient, TPM in hospital List of Access Points (AP) in our proposed SRSTMIS framework:Smart phone of the patient, List of Trust Levels (TL) in our proposed SRSTMIS framework: There are three trust boundaries in our proposed framework they are i) User and Device boundary: User and Device boundary is between Patient and the MHA in the UICC of the Patient's smartphone in which patient credentials are entered and the patient receives response from the MHA. ii) Internet boundary:Internet boundary is between Patient's Smartphone and the Hospital Server (H), Patient encrypts the messages using the shared symmetric key between 'P' and 'H' ensuring application security and the TLS is used in ensuring communication security. iii) CorpNet Trust boundary:CorpNet Trust boundary is between the Hospital Server (H) and the Hospital Database, messages are exchanged between them and protected using IPSec protocol. Hospital has Private Hospital Network (PHN) which hosts hospital and hospital database. PHN is a dedicated network which connects differed sub-entities in the hospital premises, outside traffic is not allowed in the PHN. (2) Identify and Rank all the potential threats: The capabilities and objectives of an intruder which can arise from inside or outside the organization are termed as threats. Threats can be identified by analyzing the assets and access points in the framework which compromise the security properties such as availability, mutual authentication, confidentiality, non-repudiation and integrity. Microsoft STRIDE model classifies threats into six classes.
i) Spoofing is an attempt to gain access to a system by means of a false identity. Patient's smart phone of may be spoofed by an attacker/adversary which leads to data being written to the attacker/adversary's device instead of the patient's smart phone. ii) Tampering is a means of manipulation of data without the consent and permission of the data owner.
Role-Based Access Control (RBAC) is deployed in our proposed healthcare framework., so tampering of data and logs are not possible. Use of SE, UICC and TPM. Tampering Data -patients and doctors intentionally/unintentionally can modify, update, and remove/delete patient's medical data. iii) Repudiation is the ability of authorized users denying of conducting specific actions. iv) Unwanted exposure of confidential information is called Information disclosure. v) The process of making a system or an application unavailable to its legitimate users is called Denial of service. vi) When a user with limited privileges elevates his/her privileges by identity theft in order to gain access to crucial assets of a system. (3) Discover countermeasures and build mitigation plan: After identifying the assets and threats we need to have mechanisms and strategies in order to mitigate these threats. This phase provides a mitigation plan to overcome the identified threats. a) Countermeasures for Spoofing: In our proposed TMIS framework, Spoofing is not possible at Sensor (S), Smartphone (P) and at the Hospital (H) as all the entities involved in the framework have their credentials and MHA on SE, UICC and TPM which are tamper resistant. In addition to these all the entities adopt WBC which ensures the safety and security of keys and applications. Proposed TMIS framework ensures application and communication security. b) Countermeasures for Tampering: SRSTMIS framework ensures integrity of the messages as the messages are encrypted and these encrypted messages contain timestamps and nonce thereby ensuring timeliness, freshness, uniqueness and integrity properties. So tampering of messages is not possible. In our proposed TMIS framework Hospital Employs Logging and Monitoring Manager (LMM) which keeps track of the logging information. CA employs an auditor in the hospital who audits the working of the hospital. c) Countermeasures for Repudiation:In our proposed TMIS framework Hospital Employs Logging and VOLUME 10, 2022 Monitoring Manager (LMM) which keeps track of the logging information. CA employs an auditor in the hospital who audits the working of the hospital. MHAs in the Sensor (S) and Smartphone (P) logs all the information sent and received. d) Countermeasures for Information Disclosure: SRSTMIS framework adopts RBAC (Role Based Access Control). Encrypted medical data is exchanged in SRSTMIS framework thereby ensuring confidentiality property. SRSTMIS framework ensures integrity of the messages as the messages are encrypted and these encrypted messages contain timestamps and nonce thereby ensuring timeliness, freshness, uniqueness and integrity properties. So, SRSTMIS framework ensures the secrecy and integrity of the patient's data, so patient's data cannot be compromised. e) Countermeasures for Denial of service:SRSTMIS framework resists DoS and DDoS attacks as 'H' detects these attacks by change-point detection, activity profiling and wavelet-based signal analysis detection techniques. In addition to these 'H' installs DoS/DDoS protection tools such as ''Fort Guard Anti-DDoS''. f) Countermeasures for Elevation of privilege: SRSTMIS framework adopts RBAC (Role Based Access Control). Attacker will not be able to impersonate any of the entities involved in the ecosystem our proposed SRSTMIS framework ensures application and communication security. In addition to these  SRSTMIS is proposed on SE, UICC and TPM by adopting MPKI and WBC mechanisms.

VII. IMPLEMENTATION AND PERFORMANCE ANALYSIS OF THE PROPOSED PROTOCOL
This section highlights the implementation details and performance analysis of the proposed protocol.

A. IMPLEMENTATION OF THE PROPOSED PROTOCOL
SRSTMIS is implemented in Android Studio using Kotlin language. AES symmetric encryption algorithm is used to ensure all the confidentiality property, Timestamps and Nonce ensures the freshness and timeliness of the messages transmitted. GCM mode is used in AES symmetric encryption algorithm which encrypts the patient's readings

1) PERFORMANCE ANALYSIS OF THE PROPOSED PROTOCOL
The efficacy of the SRSTMIS's protocol is better than the existing TMIS solutions as it only employs symmetric encryption/decryption and one-way hash operations.   According to [14] the time complexities are TH = 0.0004 and TS = 0.1303 where TH is time taken for hashing and TS is the time taken for encryption/decryption, TH and TS are in seconds. SRSTMIS's protocol has better performance as shown in figure 9, table 5 compares the energy consumption of SRSTMIS's protocol with the other existing works in the literature. In order to calculate energy consumption, we used hash and symmetric key operations, according to [15] the energy consumed in generating one encryption/decryption operation based on AES algorithm is 1.21 Micro Joules/byte and for generating one SHA-1 hash (EH) is 0.76 Micro Joules as shown in Figure 10.

VIII. CONCLUSION
This paper proposes a TMIS architecture, a procedure to verify the safety and security of patients credentials and Mobile Healthcare Applications (MHA) and a secure protocol. We have threat modeled our proposed healthcare framework using STRIDE approach and successfully verified using Microsoft Threat Modeling tool 2016. Our proposed secure and lightweight authentication scheme has been successfully verified with BAN logic and Scyther tool, SRSTMIS withstands DDoS and DDoS attacks in addition to multi-protocol and Blue Borne, reverse engineering and Phlashing attacks. Proposed framework overcomes information leakage from sensors during rest and during transit, key loss from healthcare applications of the sensors and smart phone and false authentication among the entities involved in the system thereby ensuring HIPAA regulations. We have successfully implemented our protocol using kotlin language in Android Studio. SRSTMIS is better than the existing TMIS solutions. Safety and security of the keys are ensured by White-Box Cryptography (WBC). Proposed framework overcomes reverse engineering attacks. SRSTMIS's communication, computational and energy costs are far less than the existing TMIS solutions.