Cloud-based Secure Smartcard Healthcare Monitoring and Tracking System Kartik Moudgil, Ria Maheshwari, Harshal Bharatkumar Parekh, Kailas Devadkar harshalparekh@outlook.com, kartik@kartik.one, riamaheshwari2507@gmail.com, kailas_devadkar@spit.ac.in Sardar Patel Institute of Technology, Andheri, Mumbai Abstract— As technological advances make strides into many fields, including that of healthcare, the onus is on us to take advantage of these. Technology has been adopted by the healthcare industry since long; however, it retains a secondary role as far as patient data management and connectivity is concerned. The aim of this paper is to devise a novel architecture and framework to make available the patient data in a concise, secure and accurate manner. Our architecture is based on a secure cloud system as backend, using smartcard as authenticators. Our proposed architecture consists of: i) A secure hybrid cloud system, based on the open-source Eucalyptus software ii) A secure smartcard called MediCard to securely authenticate users iii) An Android application that serves as a user-friendly interface to the patient data, from the perspective of the healthcare provider as well the patient. iii) A website that allows easy management of the patient data such as medical reports, prescriptions etc. by the healthcare provider. These three components allow a wide variety of applications that can be undertaken by this healthcare system, which are detailed ahead. This model seeks to help the healthcare providers such as hospitals, physicians, diagnostic labs, and chemists by electronically managing all the patient data they need access to in a secure, and efficient manner. Keywords—healthcare, cloud, hybrid, NFC, data, smartcard I. INTRODUCTION The healthcare industry is one of the few sectors that has a substantial economic as well as social impact on a society. The true benefits of health cannot be gauged in pure economic terms. For this industry to grow at its current rate, the challenges it is beginning to face must be addressed. As our population increases and more people require access to higher standards of healthcare, the huge volumes of patient data that is inevitably generated must be managed with high levels of integrity, security and availability. So, an efficient system and framework is needed to manage the global health care system. The healthcare providers are now realizing the importance of their data as a way to keep up with these drastic changes. For example, the highest level of practices are maintained by the hospitals which are looking for insight into patients, care teams, and services to gain opportunities for improving the quality. As technology permeates the healthcare sector further, the masses too are taking advantage of it, trying to make smart and informed decisions on physicians, as well as other healthcare providers. Crowd-sourced information is being utilized on a real-time basis for these kinds of decisions already. Naturally, the availability of their own medical records, stored in a secure location, backed by secure methods to track and access the same is the next logical step. The data about patients, physicians, services, and locations is currently fragmented across multiple systems; it is quite difficult to build data-management relationships among all these entities, without the presence of an external, centralized entity. The challenge for healthcare providers is existence of incomplete, inconsistent master data. This is an informationbased challenge. In a country like India, patient records are still maintained primarily on paper, and only by larger hospitals it is stored digitally. Hard copy records are extremely unreliable at times, being slow and cumbersome to access, as well as less accurate and generally unreliable. Adoption of our architecture data management system for healthcare will solve this problem by maintaining a centralized, secure bank of all patient records. Using this data, healthcare organizations achieve a better understanding of patients, and may efficiently decide procedures and facilities best for them. The centralized information obtained about doctors/hospitals/labs relationships, making it easier to compare their performances, identify and replicate optimum practices from in-network physicians. Irrespective of the location of the patient or the healthcare provider, availability of the data is paramount for both, patient satisfaction as well as improved clinical outcomes. Cloud technologies facilitate this data trend. Clinics, hospitals, diagnostic labs, etc. all benefit from quick access to computing and large storage facilities in a digital setting. Electronic medical records, data analysis and health information exchange are some of the features that the model aims to implement, backed by this hybrid cloud-based data management model. Compared to other verticals of cyber-crime victims, healthcare had the highest percentage of incidents from data theft/loss. [1] With online cloud-based record system personal health, payment information, also intellectual property i.e. prescription and report details need to be made private. Hence, the security measures taken for the Smartcard users are very important, to prevent the occurrence of any untoward incidents involving very private user data. 978-1-5090-3239-6/17/$31.00 © 2017 IEEE The paper is divided into nine sections. Section II pertains to the details of the existing work done in this area, both theoretical and practical schemes. Section III comprises of the details of the underlying NFC technology behind the smartcard this model uses, including specific information about our implementation model. Section IV is composed of the specialized NFC card model/family proposed to be used and all the details that are associated with its capabilities, including the very important anti-cloning measure. Section V details the application scenarios for our model; the functionality it provides and the systems it is supported by. Section VI pertains to the specific hybrid cloud system proposed in this paper. Section VII brings to the forefront the security concerns that are related to the choice of the software and hardware base used in our model; it addresses those issues as well as their resolution. Section VIII details the implementation of our model architecture practically, at the various user stratum. The final section, Section IX, deals with the goals this model seeks to achieve, and the various refinements that are proposed for implementation in the future. III. NEAR FIELD COMMUNICATION Two electronic devices can communicate using Near-field communication (NFC) which defines a set of communication protocols, to establish communication between two devices out of which one is a portable device like smartphone by bringing them in a proximity of 4 cm i.e. 1.57 inches of each other. NFC is a subset of the RFID family where radio waves are used to uniquely identify items operating at 13.56 MHz frequency. NFC is designed to be a secure and convenient form mechanism to exchange data; NFC-capable devices include the functionality of an NFC reader and an NFC tag. Contactless payment systems which are similar to credit cards and electronic ticket smartcards is a common application of NFC allowing mobile payment to replace/supplement these systems. NFC-enabled devices can act as electronic identity documents and key cards. It also offers a minimum setup method to establish a connection that can be used configure more capable wireless connections in turn. Each full NFC device can work in three modes: II. RELATED WORK The European Health Insurance Card [2] (2004) was designed to cover the cost of healthcare, state-provided, for British travellers in certain other European countries. To obtain a card online, the applicant provided a personal NHS number, allocated upon registration at a General Practitioner’s office. Migrants were able to charge the full cost of medical treatment in their home countries to the UK. The card, valid for 5 years, was being replicated to be used for more than this term. The government incurred a heavy loss for two years before they acted on this. This model will tackle this kind of fraud, by ensuring that one citizen gets just one Smart card, which is issued on that person’s Aadhaar Card number (or equivalent). Even if misplaced or damaged, another Smart Card with the previous one’s credentials and records can be issued. This will ensure that a new card is not issued for the same person twice. Also, use of anti-cloning technology in the smartcard itself will prevent such any cloning fraud with our model. In 2009, in Karachi, Pakistan- a mobile phone and NFCbased patient presence alert system was bought up by Interactive Research and Development (IRD) is a non-profit organization for the detection of early pneumonia. Patients enrolled in the study would wear an NFC bracelet/anklet or carry an NFC card, which the physician scanned with a cell phone (Nokia 6131) on a medical encounter. [3] It worked very well in the city area for a long time. The issues it eventually faced were: keeping the data secure on each device, as NFC cards can be read by a lot of devices. Also, they were not successful in securing the GPRS communication and central database. It got expensive to scale. The entire success of this project was based on the physician’s phone. A lot of security measures have been taken in the proposed model to rectify these issues. Our hybrid cloud architecture takes care of all issues above and more that may arise in the future, by using a scalable and flexible base. [4] NFC card emulation: Exposes device functionality emulating a passive NFC smartcard, allowing users to perform transactions such as payment or ticketing using the device in lieu of the smartcard. NFC reader/writer: The reader/writer is the smartphone serving as the active device and the card is the NFC tag serving as the passive device. NFC peer-to-peer: Information can be exchanged when two NFC enabled devices communicate in ad-hoc fashion. Fig. 1: NFC Tag architecture [5] NFC tag type definitions: There are four basic tag types are given designations 1 to 4 and each having different format and capacity. These NFC tag type formats are based on ISO 14443 Types A and B (international standard for contact-less smartcards) and Sony FeliCa which conforms to ISO 18092. The different NFC tag type definitions are as follows: 1. Tag 1 Type: These NFC tags are RW capable; users may configure the tag to be read-only. Memory availability is 96 bytes expandable up to 2 Kbyte. The communication speed of this NFC tag is 106 Kbit/s. 2. Tag 2 Type: Type 2 is similar to Type 1 except the memory size of this tag type is only 48 bytes although this can be expanded to 2 Kbyte. 3. Tag 3 Type: The NFC Tag 3 Type is based on the Sony FeliCa system. It has 2 Kbyte memory and communications speed is 212 Kbit/s. 4. Tag 4 Type: Pre-configuration of these tags at manufacture is possible to either read / re-writable, or read-only. The tag memory can be up to 32 Kbytes and communication speed between 106 Kbit/s and 424 Kbit/s. [6] Passive NFC tags do not possess their own power. A small amount of power is drawn by the NFC tag from the reader/writer to power the tag. The tag is then enabled for data transfer. ISO /IEC 7816-4 commands. The most important characteristics of the device is DESFire where DES stands for Data Encryption Standard i.e. it indicates high level security using 3DES or AES; FIRE stand for Fast, Innovative, Reliable and Secure IC for contactless proximity transactions. [7] The data transfer rates supported by the device are: 106 Kbit/s, 212 Kbit/s, 424 Kbit/s, 848 Kbit/s with an input frequency of 13.56 MHz The MIFARE DESFire EV2 features non-volatile memory with the capacities: 2 kB, 4 kB or 8 kB, EEPROM (Electrically Erasable Programmable Read Only Memory) with up to 25 years of data retention. The read and write operations are very quick typically 1ms for each transfer. Each device has a unique 7-digit serial number having a facility of “RANDOM” ID (optional) to enhance the level of security along with 3 pass mutual authentication according to ISO/IEC 7816-4. The key management for the device supports up to 15 different keys: 1 master key and up to 14 application keys. The hardware DES uses using 56/112/168 bit keys and AES uses a 128-bit key encrypting key on RF-channel. MIsmartApp is available for the application programmers as a delegated application management. Several checks are performed by the device such as i) Proximity check for protection against Relay attacks. Fig. 2: NFC Reader architecture [5] Read speed: Within range, it is necessary for the NFC tag to pass all its data over. Block transfer is allowed by NFC tag type 1 maintaining read performance. Die size: Lower cost is possible with a smaller die and also the NFC tag being less obtrusive. Unit price: The NFC tags are aimed to have a low unit price. The NFC tags can be made read-only makeReadOnly() API function, making the tag irreversibly read-only. IV. MIFARE DESFIRE EV2 The model proposes to use MIFARE DESFire EV2 or equivalent, a part of the MIFARE DESFire family. Smartcard anti-cloning should be one of the characteristics of the tag card used in the model. The MIFARE DESFire EV2 possesses a similar security certification as the banking cards or electronic passports i.e. Common Criteria EAL5+. The device is based on air interface and cryptographic methods and is complaint to ISO/IEC 14443A specifications supporting an optional Fig. 3: MIFARE DESFire EV2 architecture [7] V. APPLICATIONS A. Prescriptions A prescription is information or a set of instructions given by the doctor to the patient regarding purpose of the visit, symptoms, and it also serves as the document which allows the patient to purchase medication whose availability to the general public is restricted (Schedule H drugs as per Indian Penal Code). The proposed model serves the feature of writing a prescription which is non-volatile i.e. replacing the traditional paperwork and digitalizing the prescriptions. The doctor can write the prescription using the application on his mobile device or through the web services of the system. The patients can easily change the doctor whilst ongoing treatment and the changed doctor can quickly and easily understand the situation by reading all the older prescriptions and reports of the patient and start the medication. The prescription consists of the date of the appointment, time of the prescription, identification of the doctor and the patient, the symptoms, detected disease and the medication. The system authorizes the doctor or patient by the MediCard. Current date and time are tagged in the prescriptions and it is used to sort the prescriptions in order, to keep track of duration of treatment. The prescriptions are stored on the hybrid cloud storage. The prescriptions once stored, cannot be altered. The patients can view the prescriptions and are thus authorized to buy the medicines. The application also provides an interface to order the medication directly, using third-party medicine stores available online. The stores can be filtered based on location, availability, discounts offered. The model is also extended to the secondary healthcare providers, such as chemists. Integrated into this framework, they are given limited credentials to be able to verify the validity of the prescription shown by the patient. This feature is discretionary and may be utilized in a limited manner; for the certain medications, which are liable to be misused, so that their use is heavily monitored and can be easily tracked. B. Reports The process of determining a condition which explains the symptoms of a patient or determining a disease is diagnosis. The outcome of this process (text reports, codes, images, atomic results) are amalgamated together to form a report. The diagnostic laboratories add the reports to the hybrid cloud for each patient. The reports added by the diagnostic centre may have deletion or modification restricted. The reports are uploaded in the standard PDF format for compatibility. All the types of laboratory, pathology, radiology reports such as Xrays, ECG, blood test, etc. are stored in the cloud as PDF files. Thus, it is possible for the patients as well as the doctors to view the reports at their convenience. Thus, the overhead of maintaining reports of a long term chronic disease for a patient is diminished drastically in a systematic manner. C. Personalised Treatment Programs This proposed model also allows doctors to start a program for a particular patient bearing a particular problem. Example: a physiotherapist can start a program for a patient having tennis elbow (a condition causing pain in the elbow due to overuse of arm muscles), or chronic back pain. Such a patient needs to visit the physiotherapist daily. The doctor can set the duration of the program and add prescriptions to it. The program allows to order the medication altogether and book the appointments for the tenure. This application is advantageous as the linked appointments and prescriptions are grouped together in a program in an intelligent manner. After the successful completion of the program, an analysis report will be created using all the prescriptions given during the program course. The report will include the graph and statistics indicating the improvement or deterioration in the patient’s health. In case the patient switches doctors in the meantime, this analysis helps the new doctor to understand the case quickly rather than going through all the prescriptions of the program. Personalized programming is a methodical and meticulous approach in system designing. D. Payments NFC-based e-wallets are being adopted by the masses, along with merchants and vendors for the purpose of quick, cashless transactions. They have proven to be largely secure and fast. The mobile e-wallet of the individual is used to execute the payment. They are secured using the most modern methodologies, and as the underlying hardware is up to date, the security of this method is almost guaranteed. Various technology giants around the world have rolled out their own versions of this payment model, such as Apple Pay and Android Pay. Within our proposed model, multifarious payment options are available such as linking with the mobile wallet, credit cards, debit cards, and third party applications, all with the help of such payment avenues. The first step of the payment procedure is to authorize the doctor’s account. The doctors need to have their account registered and approved as a trusted entity. The availability of the third-party applications for payments depends upon the doctor’s registered accounts, or the preferred modes of transactions. To make a secured payment, the doctor has to bill the amount and tap the patient’s card again to the mobile device. The patient’s account is then validated and the preferred payment mode is chosen. The verification is completed after the patient authorizes the transaction by means of a secure code, OTP or biometric authentication, depending on the mode of payment selected, and the third-party provider’s implementation. The application also keeps the track using the log event register and patients can track the payments sorted according to time, location and doctor. This application is advantageous as the usage of mobile payments increases. E. Live patient monitoring The records for the patients admitted into the hospital are critical and has to be kept secured and accurate. Rewritable NFC tag cards can be used in these cases where a card can be used for a patient to store all the data in the database and after the discharge, the ID in the card can be re-written to provide it to another patient thus reusing the resources. The data of the patient is not deleted from the database and a relation is established between that data and the patient's’ application ID. All the hospital admissions can also be tracked in this way. The scope of this application is not limited to storing the data in the database but the doctors can also access the card and know the updates about the patient immediately and speedily. This application is advantageous in organizing the data in nonvolatile manner and linking it as a whole to the patient's’ application ID. The updates that can be stored may include the blood pressure of the patient, heart rate, temperature, saline levels, electrolyte levels, IV medication, as well as other relevant physiological parameters. F. Patient Medical Data Tracking & Analysis The hybrid cloud system provides an interesting set of opportunities regarding data analysis. Usually, the patient data is scattered across inaccessible, closed data storage modes such as physical records. Having it in a centralized manner allows easy access to the data for various purposes, such as data analysis. As this data, may be obtained in an aggregated format easily, it is easy to run statistical models on these datasets. Using only the non-personally identifiable patient information, the privacy of the patient can be respected, as well as run simulation and analytical processing models on the datasets. This raises the prospect of real-time, live medical monitoring. The progression of any virulent diseases can be tracked much faster, as patient records get updated in realtime. Other secondary health factors such as propensity to a chronic health condition, grouped by a patient factor, locationwise or any other factor, can be easily tracked. Such observations can lead to quicker response time from the concerned authorities, as well as targeted actions to ensure the quick dispersion of such threats. This information, however must be very carefully managed; it can be very personal if the identity of the person involved is revealed somehow. Hence, the data processing models must be run within the hybrid cloud itself, and care must be taken to ensure that only statistical, aggregated queries are run on the data, with personally identifying information such as an UID being inaccessible to such statistical programs. turning digital. Loss of such data is way too frequent and can lead to bigger problems as this concerns our health. The aggregation of data, loss or invalidated records, lack of centralized structure calls for a better solution. Cloud computing lets healthcare organizations focus on health care rather than worrying about data centers, digital real estate to house them, and skilled professionals to maintain and operate them. Hybrid Cloud-based applications can easily scale up or down as demand changes; from a development perspective, they’re flexible and accessible. A Hybrid Cloud based architecture extends flexibilities to collaborate which is the key to this application. Eucalyptus by Amazon Web Services (AWS) is a private, compatible and hybrid cloud computing environment. Our proposed data storage architecture is backed by this platform. It features pooling compute, storage, and network resources that can be scaled up dynamically as workloads grow. So as the amount of data increases, corresponding to the number of patients and their records, the database itself must grow and configure itself to the data load it is being subject to. VI. HYBRID CLOUD MODEL The healthcare industry has already begun to transition towards full digitalization, with all its benefits being apparent in the ease of access to healthcare data and its management. Currently, a limited local network backed by a local database server is commonly used to store and manage the patient data. This effectively limits the accessibility of the data, to both the patient as well as the other healthcare providers who might need access to that data. Compounding this problem, overall digitization and improvement in the healthcare standards has effectively increased the magnitude of data collected exponentially. Thus, the problem of data management in healthcare is with the unofficial term “huge data” is acceptable as to be a big data problem. And the only way to deal with big data lies in the modern, non-traditional methodologies. The costs and issues associated with obtaining equivalent functionality and security as a hybrid cloud architecture, from legacy systems is too prohibitive and may not well be possible. In such a scenario, the scalability of cloud systems needs to be able to keep up with the data trends. Thus, the migration into cloud systems is inevitable. The traditional architecture of medical records on local computers of hospitals, labs, clinics and hard copies with individual patients is rendering obsolete as everything else is Fig. 4: Eucalyptus architecture overview [8] Our architecture consists of a set of instances which are fixed collections of software modules when put to use at runtime. Cloud instances are virtual servers running off a hybrid cloud network, which are provisioned by using specific machine images, that include all the necessary software modules and information needed for the node to be handle the cloud functionality. Using instances makes cloud computing dynamic; it allows adding of resources (processing nodes/servers) to be done on demand, enabling maximum throughput to be maintained across the hybrid cloud. As demand reaches beyond the limits of the current servers, instances allow a new server to be configured and added to the hybrid cloud in real time. This abstraction allows seamless transitioning of data and resources across the hybrid cloud, which is the reason behind its scalability and flexibility. Eucalyptus provides a mechanism to firewall off an instance using just an IP address and functionality port block or allow. Instances are isolated at TCP/IP layer 2. In the absence of this, any user could manipulate the networking of records and access nearby data violating the basic archetype of instance isolation and separation. Another feature of Eucalyptus is that identities can be grouped together for common access control. Common Hybrid cloud security concerns tackled: 1. Data breach - A data breach is when sensitive, protected or confidential information is released, viewed, stolen or used by an individual who is not authorized to do so. Passwords, sensitive medical reports and appointment details and prescriptions are a patient’s intellectual property that needs securing. The standard way to combat this is by format-preserving encryption (FPE) to encrypt structured data and Tokenization - the process of replacing sensitive data, such as the patient ID, with a no sensitive equivalent(token). 2. Insufficient Identity, Credential and Access ManagementPatient, doctor or lab credentials and cryptographic keys won’t be embedded in source code. A well-secured public key infrastructure (PKI) is utilized here, as the cryptographic base. Multifactor authentication systems – smartcard, OTP (planned), and phone authentication helps address password theft, and enable access to resources without user consent. The cryptographic keys are periodically rotated, including TLS certificates, keys used to protect access and encrypt data. 3. Advanced Persistent Threats (APTs) are cyberattacks that infiltrate systems to establish a foothold in the computing infrastructure. Direct hacking systems, USB devices delivering attack code, penetration through partner networks, use of third-party unsecured networks that act as common points of entry for APTs. APTs customize tools, to stop these attacks blocking ways for them to communicate out of the program and unusual SQL activities are handled. [9] 4. Denial-of-service (DoS) attacks prevent users of a service from being able to access their data. They cause a system slowdown, asymmetric DoS (application level) attacks can take advantage of vulnerabilities in a web server, databases of cloud resources to take out an application with a single extremely small attack payload. To prevent this, techniques used can be: firewall implementation, flow level filter used to detect the low rate DoS attack, application of attack surface reduction and false positives reduction. [10] TLS (Transport Layer Security) [11] is an updated, more secure, version of SSL - Secure Sockets Layer which has been the standard technology for keeping an internet connection and sensitive data that is being sent between two systems secure. It works by making data transferred between users and sites, user two systems impossible to read using encryption algorithms to scramble data in transit. This framework is proposed to be administered into the data storage. The tenets of TLS 1.3 are utilized, a working draft that has added features like: i) Requiring digital signatures even when a previous configuration is used; ii) Removing support for weak and lesser-used elliptic curve encryptions, MD5 and SHA-224 cryptographic hash function and more VII. SECURITY Many organizations, and hence software systems, seek to utilize the benefits of cloud; however, data security remains a prime concern. Maintaining effective data security using encryption in the cloud is the objective, which is addressed through a number of techniques. For data security expansion at an asymmetrical rate, there arise a number of issues to be addressed. While encryption is the fundamental technology that drives security, encryption in the cloud is a relatively newer, and as yet complex issue. One of the factors making this decision hard is the availability of so many encryption techniques and algorithms; thus, any decision to select the particular ones to use can be quite hard. In legacy systems, encrypted data was stored on the local servers. As applications move to the cloud, software as a service (SaaS) systems can be used that will manage the encryption and decryption keys as well as any encryption metadata. The latter approach is used in our hybrid cloud architecture. It uses Transport Layer Security (TLS) [12] encryption to create a secure link between the cryptography management server and the cloud storage application. Once the data reaches the cloud application' servers, it can encrypt/decrypt the data securely. Managing encryption keys is the most important objective in this model. For this to be achieved, separating the encryption key from the encrypted data is paramount to keep the data secure. The Cloud Storage Application might store the keys in memory while they are being used. This must be managed securely on fulfilment of the operation, using secure wipe techniques to make any possible unauthorized retrieval of the keys impossible. Encryption keys will also be kept on a separate server or storage block. A backup of all the keys is also envisioned to be kept in an offsite location in case of disaster. This backup must be audited every couple of months. It is very important to carry out these tasks regularly and efficiently, as any encrypted data is almost useless without the encryption keys. There are three basic states of the data across the hybrid cloud network: in transit, in use and at rest. Protecting data at rest is essential. Sensitive data will be protected at any layer, from the local layer to the data centre/cloud layer. As data is added to the database, security is an integral part of process, so that security moves with the data. Secure Socket Layer (SSL) approach, standard for years, has fallen out of favour since the discovery in 2014 of the POODLE (Padding Oracle On Downgraded Legacy Encryption) attack, a man-in-the-middle exploit that effectively was designed into the SSL code. The model has implemented the TLS 1.3 draft where handshake and record protocols provide authentication and security. PSK (pre-shared key) system resumption has been replaced with cipher suites. Digital signatures are checked even though a configuration is previously used. Changes in Ed448 digital signature algorithms (Ed25519) and message authentication codes (Poly1305) are made. Cloud Security Alliance recommends: Use of approved algorithms and long, random keys Encryption before transmission from the lower layers to the cloud provider Encrypted state is maintained in transit, at rest, and in use The cloud data storage must never have access to the decryption keys In our model, processing of sensitive data takes place in the cloud, and users take advantage of the cloud's economy of scale and elasticity. The data remains encrypted up to the moment of use and that both the decryption keys and the decrypted versions of the data are available in the clear only within a protected transient memory space. It is paramount that the keys and any cleartext data is securely wiped, without insecure disk copies being created by processes such as logs or any other persistent records. B. Android security The android application used in our model has access to the highly-secured information belonging to the patient, as it is one of the possible endpoints of the data communication flow. This means that the data must be unencrypted at the device, for it to be viewed and used by the patient. Also, any new data records may be added as a result of the functionality provided in our application, such as when the patient seeks to pay thru the e-wallet functionality. Here, the patient’s actions lead to a record being written to the patient’s medical records, with the date, time, purpose and case details being logged. Here, some data originates from the device, and hence it must be securely transmitted to avoid eavesdropping, malformation or loss of integrity of the data. Hence, our proposed model features TLS connections, using public-key encryption scheme: ECC for session key agreement, while using RSA 2048-bit private key for authentication. The objective is to implement the following cipher suite: TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256 1) ECDHE: ephemeral Elliptic Curve Diffie-Hellman key agreement 2) RSA based authentication The application does not store any of the data unencrypted; the cache level data is stored encrypted, and whether it is data fetched from the server or from the cache, decryption is performed on-the-fly. The use of root permissions will not be effective in breaking the data security in any way; without the card, the key cannot be obtained, as it is stored nowhere on the device itself. Thus, the only exploit avenues left are memory dumping while the decrypted data and the key is held in the memory, and the use of malicious snooping applications acquiring global hooks on events. To be immune from this issue, our application utilized the new Android API SafetyNet. SafetyNet is an API allowing the evaluation of the hardware and software of the device to verify if it’s characteristics match those with which the application was developed. Put simply, it will block out the rooted/modified devices, where it is not possible to be entirely sure about the chain of trust. The value contained in ctsProfileMatch variable, from the response received from Play Services, indicated if the device has passed the Android compatibility testing, and hence can be allowed to execute further. VIII.IMPLEMENTATION To deploy this model architecture, an application for Android devices using Android APIs (SDK 21 and above) and an interface to it thru a normal web server using PHP and MySQL as backend was made. The android application authenticates the user via the NFC MediCard or manual login using the unique MediCard ID and password, which keeps record of all test reports also downloadable offline, appointments, prescriptions-authorized by physician and a doctor search engine. It has a pop-up notification functionality for reminder of appointments. This application has been tested on Google Nexus 5 device. The android.nfc.tech package provided by Android has all required methods and classes to enable interaction with NFC tags. The card can be accessed from an authorized Android application installed using custom platform APDUs (application protocol data unit) for communication between card and card reader. The physician can send new and view old prescriptions and reports. A payment gateway is also integrated into the application. The web server has been implemented using advanced PHP and MySQL for the hybrid cloud based database set-up Eucalyptus (Version 3.3.2). The cloud was set up on two machines. One hosted the cloud controller - Walrus, the cluster and the storage controller and the other acted as the node controller that ran on a KVM hypervisor. Two simple desktops were used, each equipped with a 500GB hard drive, a 4 GB RAM, and a Virtualization Technology enabled processor on-board. [13] Fig. 5: Proposed Eucalyptus Architecture [14] Fig. 6: A demo set-up [14] The website developed has a dedicated patient, diagnostic, hospital and pharmacist profile. The patient logs using his MediCard ID and password and can keep track of his medical information, prescriptions, reports, appointments, personalized programs and a doctor search engine. Diagnostics and hospitals can add and authorize prescriptions and reports in a PDF format. A database of all medicines is maintained and they can add to this list. The card used for this implementation is the MIFARE DESFire EV2 [7]. It is a Common Criteria EAL5+ security certified card with backup management system and mutual three pass authentication system on-chip. DES indicates the high level of security and hardware cryptographic engine-DES using 56/112/168 bit keys. A 7 bytes long serial number acts as a unique identifier that generates an ID required. It has flexible, non-volatile memory. There is the mutual authentication protocol according to ISO/IEC 7816-4 on application level. These new features make this the ideal technology to use for secure communication and data storage such as our applications. IX. CONCLUSION AND FUTURE WORK This paper has proposed applications on an NFC based smartcard for improving healthcare, to secure medical object data and to implement a system of patient Smartcard or mobile device itself. The applications to this technology are significant, namely prescription management, conveyance and storage of reports, personalized doctor-patient routines, easy payment integration and live patient monitoring. Implementation of this system will decrease the load on hospitals. The model will benefit the patients and medical professionals equally. security. Periodic data access could get validated using the patient’s finger print to minimize leakage of data. Prescription and reports authentication by doctors and labs too. To facilitate the usage of this MediCard in developing countries and all sectors of society including those who do not have access to mobile internet at all times, offline saving of prescriptions, reports, programs, charts and basic emergency medical information about a patient; and uploading of data can be implemented with periodic (monthly) updating of the online cloud database. Another application to this is the NFC Smartcard integrated with Bluetooth that can unlock channels and send large quantities of data such as a timed electrocardiogram strip. This is a way in which live patient monitoring speed will be enhanced. REFERENCES [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] Another application of this technology could be embedding NFCs into patients’ medical charts at hospitals/labs. Doctors and nurses could quickly see procedures due and completed along with the patient’s prescriptions and medical history, to administer the best possible care. At the patient front, NFC in healthcare can be extended to a centralized (offline) mobile phone controlled interface. This will provide the added layer of security namely biometric [13] [14] Verizon Enterprise, source: http://www.verizonenterprise.com/verizoninsights-lab/dbir/, December 2016. Hugh Pym “NHS vulnerable to health card fraud”, source: http://www.bbc.com/news/health-33843758, December 2016. Aam Marcus, Guido Davidzon, Denise Law, Namrata Verma, Rich Fletcher, Aamir Khan, Luis Sarmenta, “Using NFC-enabled phones for public health in Developing Countries”, First international workshop on Near Field Communication, 2009, IEEE. Divyashikha Sethia, Daya Gupta, Huzur Saran, “NFC based Secure Mobile Healthcare System”, IEEE, 2014. RFwireless World, source: http://www.rfwirelessworld.com/Terminology/NFC-tag-vs-NFC-reader.html, December 2016. Radio Electronics, Ian Poole, source: RFwireless World, source: http://www.rfwireless-world.com/Terminology/NFC-tag-vs-NFCreader.html, December 2016. MF3D(H)x2; MIFARE DESFire EV2 contactless multi-application IC Data sheet, source: http://www.nxp.com/documents/short_data_sheet/MF3DX2_MF3DHX2 _SDS.pdf, December 2016. Eucalyptus Architecture Overview, source: https://en.wikipedia.org/wiki/Eucalyptus_(software), December 2016. 2016 Cloud Security Alliance “The Treacherous 12 Cloud Computing Top Threats in 2016”, source: https://downloads.cloudsecurityalliance.org/assets/research/topthreats/Treacherous-12_Cloud-Computing_Top-Threats.pdf Subramaniam. T. K, Deepa. B, “Preventing Distributed Denail of Service Attacks in Cloud Environments”, International Journal of Information Technology, Control and Automation, Vol. 6, No. 2, April 2016. Transport Layer Security, source: https://en.wikipedia.org/wiki/Transport_Layer_Security#TLS_1.3_.28dr aft.29, December 2016. Transport Layer Security Protocol, E. Rescorla, source: https://tlswg.github.io/tls13-spec/, December 2016. Arunkumar Goge, Magesh S., “Implementation of Database as a Service in a Private Cloud using EUCALYPTUS”, source: https://www.researchgate.net/publication/262375919_Implementation_o f_Database_as_a_Service_in_a_Private_Cloud_using_EUCALYPTUS, April 2013. Yohan Wadia, “Build Your Own Private Cloud with Eucalyptus”, source:http://opensourceforu.com/2014/03/build-private-cloudeucalyptus/, March 2014.