Merging RFID and Blockchain Technologies to Accelerate Big Data Medical Research Based on Physiological Signals.

The proliferation of physiological signals acquisition and monitoring system, has led to an explosion in physiological signals data. Additionally, RFID systems, blockchain technologies, and the fog computing mechanisms have significantly increased the availability of physiological signal information through big data research. The driver for the development of hybrid systems is the continuing effort in making health-care services more efficient and sustainable. Implantable medical devices (IMD) are therapeutic devices that are surgically implanted into patients' body to continuously monitor their physiological parameters. Patients treat cardiac arrhythmia due to IMD therapeutic and life-saving benefits. We focus on hybrid systems developed for patient physiological signals for collection, storage protection, and monitoring in critical care and clinical practice. In order to provide medical data privacy protection and medical decision support, the hybrid systems are presented, and RFID, blockchain, and big data technologies are used to analyse physiological signals.

1. Introduction

The medical applications are continually increasing. For handling physiological signals efficiently, specific technologies, such as data gathering using RFID protocols, infrastructures, and distributed information storage based on blockchain frameworks, are required. The hospitals applications are adopting physiological signals to realize a quicker way to visit these records. The physiological signals are responsible to offer patient care, enhance the clinical performances, and promote the clinical data research [1-5].

Since the fog computing solves the secure storage issues of big data in the clinical data research with minimal cost, the fog computing technology is customizable and economical and offers infrastructure, platform, and software. Physiological signals' analysis and migration have been proposed for accessing and sharing physiological signal data by different research labs and health-care experts, which can enable exchange of physiological signals more rapid and suitable by using RFID technologies and smart phone app platforms. The advantages of RFID protocols [6-9], the fog computing, and blockchain in the medical applications provide security and privacy protection for storing and sharing physiological signal records. It can provide doctors with collaboration ways through IMD [10] and RFID to help patients in case of emergencies mode. The new model based on blockchain can support medical background rural healthcare and analyse data for medicines and medical research [11-15].

It is urgent for different research institutions to share the encrypted physiological signals. Therefore, privacy and security problems of physiological signals are the data owners and research institutions' primary focus, when the physiological signals include a lot of sensitive information and the attackers are continually trying novel approaches to steal the physiological signals. In order to handle these problems, the medical databases adapted blockchain, and fog computing are proposed [16, 17]. The medical application ecosystems allow the regulators to share and exchange physiological signal data in Figure 1. The introduction of the blockchain-fog-RFID based on data ecosystems ensures that the individuals take control over physiological signal information. The proposed sharing data-driven economy shares the physiological signals for research and commercial purposes in Figure 1.

Figure 1

The flow of data from the individuals to the companies and research institutions.

In the paper, we protect cardiac IMD against security threats by presenting a security scheme. First, we verify and classify the IMD's major security attacks. Second, we introduce blockchain and the RFID systems to extend the IMD architecture [10] and discuss the structures of the interoperability in the medical environment, as shown in Figure 2.

Figure 2

Blockchain in the medical environment.

The motivation of the blockchian-fog-RFID method for accelerating big data medical research based on physiological signal is as follows: the method is becoming more common due to the application of powerful computers and the availability of physiological signals from various sources. However, although the complexity of physiological signals makes the complex methods particularly applicable, their application of physiological signals is generally considered earlier than in other fields. Big data has become a buzzword in medical innovation. Rapid advances in artificial intelligence particularly promise to reform medical practice from the resource allocation to the complex diseases' diagnosis. However, big data brings huge risks and challenges, including major questions about patient privacy: the importance of fairness, consent, and patient management in data collection based on RFID; data storage based on fog computing; and dealing with data breaches by using blockchian. In the future, we will discuss the method's applications in physiological signals research: basic research; disease management; aetiology; detection and diagnosis; health services research; treatment development; and treatment evaluation. The possibilities of the blockchian-fog-RFID method for accelerating big data medical research in physiological signals are enormous.

The paper contribution consist of four parts as follows:

The security scheme is a low energy cost RFID system in IMD. The applied authentication protocol is implemented on the RFID circuit without energy.

The applied energy harvesting scheme uses the enhanced WISP, which performs computational functions and uses the harvested energy to go beyond passive RFID tags.

The presented authentication protocol enables the authorized health-care professionals to obtain the access permission to cardiac IMD securely in the regular and emergency model which are determined according to the patient's ability to supply valid credentials, thanks to a biometric key distribution scheme implemented.

The schemes generate and share a master key securely based on the physiological sets of the patient collected by IMD. Monitoring and ensuring data integrity during clinical trials is not always feasible in current research systems. Blockchain makes the data collected immutable, traceable, and probably more trustworthy during clinical trials. We also improve the way we currently report adverse events.

In conclusion, we argue that the blockchain can improve the management of clinical trial data, enhance trust in the clinical research process, and simplify regulatory oversight of trials. Finally, we evaluate the security solution's security and performance.

The proposed model covers the many aspects of the health industry such as doctors, patients, and pharmacies to insurance suppliers and government. The paper shows the applications of using RFID, blockchain technologies, and fog computing for storing and managing the physiological signal data. A blockchain model for sharing physiological signals is proposed. In the next section, the combination of blockchain, RFID, and artificial intelligence (AI) technologies is suitable for collecting, storing, and handling heterogeneous physiological signal. The proposed model can be used for physiological signals management.

2. Related Work

The industry of healthcare has changed dramatically because of the boom in clinical research for physiological signal data sharing. We summarize the healthcare studies including physiological signal data, patient information obtained by fog computing, and improvements to blockchain technology. The health-care applications of physiological signal data adopt big data and deep learning technologies and provide with data confidentiality and identity authentication, so as to maintain patients' privacy. In order to more conveniently serve big data medical analysis, Rajan and Rajan [1] and Faust et al. [2] proposed the importance of medical big data privacy and the impact of data analysis on medical care.

Rajan and Rajan [1] proposed a physiological signal monitoring scheme by using the Internet of Things (IoT). Our schemes use IoT to improve the access method of physiological signals and the real-time dynamic monitoring method of the remote monitoring system, which enhances the efficiency of the remote monitoring systems. Faust et al. [2] summarized the application of deep learning algorithms in physiological signals and pointed out that deep learning methods performed better than classical analysis and machine classification methods for large and diverse datasets. Shanthapriya and Vaithianathan [3] proposed the health monitoring system for human regional network. The steganography technologies monitor patients' health safety and provide patients with data confidentiality and identity authentication. Orphanidou [4] reviewed big data applications of physiological signals, pointed out how the applications use physiological signals to provide real-time support for medical decision making in both clinical and family settings, and need to be overcome in clinical practice. Tartan et al. [5] proposed a heart rate monitoring system based on mobile devices and geographical location, which can monitor physiological signals and send alarm information when abnormal heart rate changes.

The health-care systems [6-9] are data-distribution domains where many physiological signals are generated, stored, scattered, and accessed daily by using RFID. Yuri álvarez et al. [6] described that the contribution of RFID technology can improve medical services, can offer hospital tracking of patients, drugs, and medical assets, and can improve the efficiency and safety of electronic medical applications. Martinez Pérez et al. [7] used RFID technology in the ICU (information management system) to track ICU patients' admission, nursing plan, life monitoring, prescription, and drug management process, improving the quality of patients' care during hospitalization. Adame et al. [8] proposed the monitoring systems for intelligent healthcare which provides location status and tracks patients and health-care assets. Omar et al. [9] proposed the reliable, secure, and privacy-based medical automation and organizational information management system that can provide real-time monitoring of vital signs of patients during hospitalization for intelligent patient management.

The literatures [11-15] have been tremendous concentration in blockchain applications. Xu et al. [11] provided a decentralized resource management framework based on blockchain by studying resource management issues. Aiqing and Xiaodong [12] proposed a blockchain-based security and privacy protection sharing protocol to improve the diagnosis of electronic health systems. The private blockchain is responsible for storing personal medical information (PHI), while alliance blockchain keeps the secure index record of PHI. Dubovitskaya et al. [13] proposed a framework for sharing EMR data for cancer patients based on the blockchain and implemented. Lebech et al. [14] used multisignature blockchain protocol for diabetes data management and access control, as well as sharing and encryption. The new approach helps to share diabetes data more effectively in different institutions. Yue et al. [15] proposed the medical data gateway (HGD) architecture based on blockchain, which enabled patients to safely own, control, and share the data without infringing privacy.

When different research institutions share the physiological signals, the issues of privacy and security are the primary focus of research institutions because the physiological signals include the sensitive information, and the attackers are continually trying novel approaches to steal information. In order to meet the privacy needs and deal with the security problems, medical databases which use blockchain and fog computing technology are proposed.

The enhanced trusted sharing physiological signals model features highly secured data encryption and decryption schemes. The model requires permission from the blockchain network to share patient information among medical staff. The proposed model encrypts and analyzes the physiological signals through the blockchain network, big data analysis technology, and AI technologies. Kamel et al. [16] pointed out that blockchain technology is becoming more and more important in the research of medicine and medical care, proposed eight solutions of blockchain application in medical care, and predicted that blockchain and AI solve various medical problems in the future. Jen Hung et al. [17] used blockchain in the drug supply chain to create transparent drug transaction data, prevent counterfeit drugs, and protect public health.

The abovementioned research findings do not apply blockchain to RFID systems. However, the protocol [18] proposed the RFID system based on blockchain and did not apply fog computing to medical fields. It is our innovative work to propose RFID protocol based on fog computing and block chain technology in medical systems.

RFID protocol framework based on fog computing and blockchain is used for medical big data collection and data privacy protection [19-21]. Gu et al. [19] proposed a security and privacy protection solution for fog computing, which designs a framework for security and privacy protection using fog computing and a privacy leakage based on context-based dynamic and static information to improve health and medicine infrastructure. Silva et al. [20] proposed a medical records management architecture based on fog computing. The architecture used blockchain technology to provide necessary privacy protection and to allow fog nodes to execute authorization processes in a distributed manner. Guan et al. [21] discussed data security and privacy issues in fog computing. They pointed out that the data security and privacy challenges posed by fog layers and data protection technologies in cloud computing cannot be directly applied to fog computing. Patel added the fog computing in the original blockchain medical data sharing sequence model [22]. Tang et al. [23] proposed a new game theory framework to improve the mining efficiency of blockchain network and maximize the total benefits of blockchain network. In order to improve the diagnosis of an electronic medical system, Zhang and Lin [12] proposed a security and privacy protection based on the blockchain PHI sharing (BSPP) scheme. The consensus mechanism (private blockchain and joint blockchain) is constructed by designing a blockchain data structure.

3. Mutual Authentication Protocol Using IMDs

The presented mutual authentication protocols for the WISP have two modes: the regular mode shares the IMD and the same credentials; the emergency mode is initiated when one of the following status appear. The IMD credentials are not shared by the programmer; the patients cannot communicate with the shared credentials; and the credentials configured are expired.

3.1. The Threats and Its Influence on the Medical Record

The threats and its influence on physiological signals are as follows: privacy, equity, consent, and patient governance in health information collection; discrimination in information applications; and handling data breaches.

Because of newly developing data collection and storage technologies to collect and analyse vast amounts of data, the technologies (RFID, blockchain, and artificial intelligence) enable more human experience. While strict clinical testing is still required for handling data breaches, the technologies will fuel a new age of precision medicine in various methods, as shown in Table 1.

Table 1

Audiences and influence functions of medical record.

Audiences	Influence functions
Patients	Promote diagnoses and identification of physiological signals, facilitate preventive care, and reduce costs
Doctors	The rigorous diagnosis, treatment choices, monitoring disease progression, therapy response, and patient susceptibility
Researchers	Perform large-scale disease modelling and efficacious therapies
Clinics	Risk estimation, forecasting relapse possibility, designing criteria for discharge/readmission, predicting mortality, and conveying potential crisis episodes

3.2. Physiological Signals Data Privacy Rules

While physiological signals are the lifeblood of today's digital society, numerous people are not fully aware of appropriate data collection and processing. The privacy issues are the concerns in the process of generating data. It is more significant to be considered privacy protection in healthcare, where personal physiological signals consist of a large percentage of the data. The rules and regulations guide the process of data generation, transmission, access, and exchange. The privacy storage rules are as follows: entitles patients more control over physiological signals; establishes boundaries of physiological signals' use and release; protects the privacy of physiological signal; enables patients to make choices wisely; and enables patients to be aware of methods for preventing data leakage. It is completely important to maintain the security and privacy of physiological signals by using RFID, fog computing, and blockchain.

3.3. Security Attacks and Requirements for IMDs

This part shows IMDs' main security attacks [10] and discusses the security requirements in Figure 3. Table 2 explains the symbols and definitions of all the authentication protocols.

Figure 3

Security attacks and requirements for secure IMDs.

Table 2

Symbols and definitions of the enhanced RFID system privacy protection authentication protocol.

Symbols	Definitions
C i; TIDTi; COUNT	Challenge from the DB to reader; temporary identity; count
R i; Ri∗; Ns	Response for the reader; Ri ⊕ Ns; random number generated
Ress; CRP (CiRi); Ki	h(COUNT+1‖Ri‖Ri∗)); ith challenge-response; ith key
PUFT;h (.);  ⊕; ‖(|)	PUF for the tag T; one-way hash function; XOR; concatenation
K H; KP; KD	Hospital; patient; doctor

3.4. Mutual Authentication Scheme in the Emergency Mode

The IMD and programmer can securely produce and offer the major key which is extracted from the patient's data by executing the presented mutual authentication protocol's emergency mode in Figure 4.

Figure 4

Mutual authentication protocol in the emergency mode (protocol 1).

Step1: the reader initiates the presented mutual authentication protocol's emergency mode by transmitting the synchronization request M1 = (IDR, NR, and flag) to the IMD.

Step2: WISP computes features V = RandPermute (FW ∪ F′ W) and sends V to the reader.

Step3: the reader computes Kbio = H (Q) and sends M3 = (IDR, I, HMAC (Kbio, I|Q|IDR)) to WISP.

Step4: if the number of matching characters is greater than the predefined threshold, the WISP calculates K′bio = H (Q), and verifies K′bio ? = Kbio. If the key is successfully confirmed, WISP generates NW and computes K = H (Kbio | NW) and K′ = H (K | NW). WISP admits the reader by transmitting M4 = ((NW, IDW)Kbio, HMAC (Kbio, NR|NW|IDW)).

Step5: in order to determine (NW, IDW), the reader decodes the message's first part using Kbio. After that, it verifies the authenticity of (NW, IDW) by employing HMAC function and comparing the result to the received message's second section. If they are equal, the reader calculates K = H (Kbio |NW) and K′ = H (K | NW) and then sends M5 = (Seq1, HMAC (K′, NW |Seq1)). The reader sends messages (K′, Seq1) to the programmer.

Step 6: WISP verifies the session keys' equality. IMD collects the key of session and the relevant sequence number.

Two modes (emergency mode and regular mode) have the same shortcomings. First, neither model talks about how to store large amounts of data on the database. Second, both models have secret key leakage attacks and tracking attacks. Third, neither model uses cloud storage technology or blockchain technology.

3.5. Attacks for Mutual Authentication Protocol in the Emergency Mode

3.5.1. The Reader Impersonation Attacks

The reader computes Kbio = H (Q) and then sends M3 = (IDR, I, HMAC (Kbio, I|Q|IDR)) to WISP.

In order to simplify the analysis steps, the steps 3–6 in Figure 4 are omitted here. The tracing attacks in the emergency mode have three phases.

The testing phase: the attacker chooses the target tag R, monitors the first round (1M1, 1M2, 1M3) to R, and obtains the outputs keys 1Kbio = H (Q), and the reader applies 1M3 = (IDR, I, HMAC (Kbio, I|Q|IDR)) to WISP.

The reader impersonation attacks phase: the attacker (the counterfeit reader R′) chooses the monitored information 1M1. The attacker monitors the output information (2Kbio = H (Q), 2M3 = (IDR, I, HMAC (Kbio, I|Q|IDR))) in the second round.

The decision phase: the adversary obtained the values (1Kbio, 1M3) and (2Kbio, 2M3). If (1Kbio, 1M3) ≠ (2Kbio, 2M3), and the attacker confirms that R is not R′ with the probability 1; if (1Kbio, 1M3) = (2Kbio, 2M3), the attacker makes sure that R is the counterfeit R′. Therefore, the protocol does not meet the weak indistinguishability property and suffers from the reader impersonation attacks.

3.5.2. Reducing the Calculation Cost of Reader and WISP

In order to reduce the computation of the whole systems, the HASH computational expense of the reader and WISP are high, the proposed protocol uses the PRNG function to replace HASH function.

3.6. Mutual Authentication in the Regular Mode

The regular mode ensures the secure data exchange, as shown in Figure 5.

Figure 5

Mutual authentication in the regular mode (protocol 2).

Step1: the reader sends M1′ = (NR, IDR, flag, HMAC (K, NR |IDR)) in the regular mode.

Step2: WISP can confirm the received request's freshness and the reader's authenticity. If the organized primary key has not run out, the received request from the keys is authenticated by the WISP. By contrary, the WISP rejects access by sending the denial message.

Step3: WISP computes K′ = H (K | NW), and sends M2′ = ((Nbr, NW, IDW) K, HMAC (K, NR| NW|IDW)) to reader.

Step4: when receiving the messages, the reader decodes the first part of the messages to obtain (Nbr, NW, and IDW).

Step5: after verifying successfully, the reader calculates the key value K′ using NW and sends the messages M3′ = (Seq1, HMAC (K′, NW | Seq1)).

Step6: WISP can confirm the message's freshness and the keys' equality computed on both sides. WISP increments the Nbr parameter which represents the total number of session keys which originated from the primary key.

Step7: WISP delivers the messages (K′, Seq1, Nbr) to awaken IMD antenna.

The attacks for mutual authentication protocol in the regular mode.

3.6.1. Secret Key Disclosure Attacks

The attackers monitor the delivery messages and reveal the secret keys as follows:

In Step1, M1′ = (N, IDR, flag, HMAC (K, N |IDR)), the attacker discloses IDR

In Step3, M2′ = ((Nbr, NW, IDW) K, HMAC (K, NR| NW|IDW)), the attacker discloses IDW

In Step7, (K′, Seq1, Nbr), the attacker discloses K′

3.6.2. The Tracing Attacks

In order to simplify the analysis process, the steps 3–6 in Figure 5 are omitted here. The tracing attacks have three phases.

The testing phase: the attacker chooses the target tag T. Then, she/he monitors the first round (1M1, 1M2, 1M3, 1M4) to T and obtains the outputs keys (1IDR, 1IDW).

The tracing attacks phase: we assume that the tag set (T0, T4,… T) includes T and the counterfeit tag T′. The attacker monitors the keys (2IDR, 2IDW) in the second round.

The decision phase: the adversary obtained the values (1IDR, 1IDW) and (2IDR, 2IDW). If (1IDR, 1IDW) ≠ (2IDR, 2IDW), the attacker confirms that T′ is not T with the probability 1; if (1IDR, 1IDW) = (2IDR, 2IDW), the attacker makes sure that  T is T (the counterfeit tag T′). Therefore, the original protocol in the regular mode does not meet the weak indistinguishability property and suffers from the tracing attacks.

3.6.3. Medical Framework Based on RFID, Blockchain, and Artificial Intelligence

At present, amounts of patients have the comprehensive datasets which consist of clinical history (the genetic, lifestyle data, drug, and blood biochemistry). In addition, the consumer companies and the pharmaceutical are willing to pay much money for the vast personal physiological signal data applied to train their AI model via using the machine learning. We proposed the medical framework based on RFID, blockchain, and artificial intelligence, as in Figure 6.

Figure 6

The medical framework based on RFID, blockchain, and artificial intelligence.

Previous researches based on RFID, blockchain, and artificial intelligence mainly focused on the medical application, respectively. The studies improve the time proficiency of physiological signal data processing and contribute to medical data management by combining three technologies. The effectiveness of the medical framework involves low resource usage, large computation time, more energy, less power, and low memory consumption (Algorithm 1).

Algorithm 1

The suggested mutual authentication protocol in the emergency mode.

4. Security and Performance Analysis of Protocol 3 and Protocol 4

The protocol 3 and protocol 4 are more suitable to store physiological signals in medical applications.

4.1. Security Analysis for Protocol 3

Scheme 3 overcomes the weaknesses of protocol 1, and the protocol 4 overcomes the weaknesses of protocol 2.

4.1.1. The Reader Impersonation Attacks Resistance

In order to resist the reader impersonation attacks, the reader calculates K′ = PRNG (Q||N) using N. Even if the attacker monitors the output information (2K = PRNG (Q||N′), 2M3=(IDR, I, HMAC (2K, I||Q||IDR))) using the new nonce N′ R in the second round, the attacker cannot counterfeit the original reader.

4.1.2. Key Leak Attack Resistance

In order to resist the key leak attacks, WISP calculates B =  N ⊕ ID; the reader calculates K = PRNG (Kbio||N); and K′ = PRNG (K||N).

4.1.3. Provision of Data Integrity Verification

In order to meet data integrity, the protocol 3 has used HMAC hash calculation to protect the integrity of messages (K1, Seq1).

4.1.4. Provision of Scalability and Efficiency

In order to satisfy the scalability, each tag identifier does not match the corresponding key in DB. Therefore, the identifications of tag keys do not match one by one in DB of the improved protocol, which guarantees the efficiency of tag authentication and satisfies the scalability property.

4.1.5. Replay Attacks Resistance

The attacker replays the messages to authenticate by monitoring the previous information. In order to resist replay attacks, all messages are encrypted by using the random numbers (Nbr, N, and N) and combined with PRNG function.

4.1.6. Provision of Data Integrity Verification

In order to achieve the property of data integrity, we have used PRNG calculation K′ = PRNG (K|N) to protect the integrity of K′.

4.2. Security Analysis for Protocol 4

4.2.1. Secret Key Disclosure Attacks Resistance

In order to achieve anonymous and privacy requirements in improved protocol 4, the protocol uses the XOR function to encrypt the transmitted keys as follows:

B = ID ⊕ N, K1 = K′ ⊕ NR, K2 = K′ ⊕ N.

4.2.2. Tracing Attacks Resistance

The key updating mechanism K′ = PRNG (K|N) involves the i keys and the nonces (N, K). The i key K cannot be cracked by the (i+1) keys K and the i sessions. The reasons are that PRNG functions protect the parameters by the encrypted messages. Therefore, the enhanced protocols resist the tracing attacks.

4.2.3. Availability and Desynchronization Attacks Resistance

In order to provide anonymity, the communication components (tag and DB) update the shared messages after completing the conversation. If the opponents destroy the updating process, the authentication scheme is subjected to desynchronization attacks. In order to guarantee the confidentiality and anonymity of K, the messages synchronously should be updated. In addition, the attacker knows the shared key K′ during the updating processes, which is protected by the random numbers (N, N). The improved protocol is desynchronization resistance.

4.3. The Comparisons of Security and Performance Analysis

Table 3 lists the computational cost for five protocols. The computational costs of tags in protocol 3 are 3PRNG + Xor, and the computational costs of tags in protocol 4 are 2PRNG + Xor. The safety performances of the enhanced protocols are superior to other schemes. Compared with the original protocol 1 and protocol 2, the improved protocols support the security enhancements and ensure the function such as integrity, efficiency, and user privacy.

Table 3

The comparisons of the performance analysis and safety performance.

Performance	Protocol 1	Protocol 3	Protocol 2	Protocol 4	Protocol 5
F0	No	Yes	No	Yes	No
F1	3H + Xor	3PRNG + Xor	2H + Xor	2PRNG + Xor	1PRNG+2Xor
F2	No	Yes	No	Yes	Yes
F3	No	Yes	No	Yes	Yes
F4	No	Yes	No	Yes	Yes
Attack types	Protocol 1	Protocol 3	Protocol 2	Protocol 4	Protocol 5
R1	No	Yes	No	Yes	No
R2	Yes	Yes	Yes	Yes	Yes
R3	Yes	Yes	Yes	Yes	Yes
R4	Yes	Yes	Yes	Yes	Yes
R5	No	Yes	No	Yes	Yes
R6	No	Yes	Yes	Yes	Yes

F0: provision of scalability and efficiency; F1: storage cost (tag); F2: blockchain-enabled; F3: cloud computing-enabled; F4: fog computing-enabled. R1: key leak attacks resistance; R2: replay attacks resistance; R3: desynchronization attacks resistance; R4: reader impersonation attacks resistance; R5: tracking attacks resistance; R6: tag impersonation attacks resistance.

5. Blockchain Framework for Security and Privacy Storage and Sharing

A framework is developed to share physiological signals' cross domain and build the radiological studies' ledger and patient-defined access permissions by applying the blockchain as the distributed data store. Relative disadvantages of the framework include the privacy's complexity and security models. Ultimately, the large-scale feasibility of the approach remains to be demonstrated.

The peculiar health-care technologies are required, such as parallel processing, distributed data network, scalable storage, frameworks, and infrastructures. The fog computing is economical and customizable, since fog computing handles these complex problems in the virtual environment and only needs to pay for the used services and resources.

The sharing physiological signals systems are important in different medical institutions, but the current infrastructure for transmitting physiological signals relies on the trust third-party intermediaries. We propose the framework of cross-domain sharing image where the blockchain is used as the distributed data storage to establish patient-defined access rights. The blockchain framework is verified to eliminate the access permission of the third-party to protected physiological signal information, meets many standards of the interoperable medical system, and easily generalizes to fields beyond physiological signal. We summarize the framework based on blockchain to allow patients to securely grant electronic access permission to their physiological signal data and describe the advantages and disadvantages of the approach.

The actual transmission of physiological signals requires the physiological signals receiver who transmits the signed request to the URL endpoint. The individual service is the requesting entity that the access permission of the physiological signals study is authorized to by the owner (patient). The studies of all patients' physiological signals result in the huge blockchain, far too large to download, store, and validate for nodes running on mobile devices. The size of the blockchain has been proven to be the limiting element for chains storing the transactional data.

Considering all of these factors, sharing the physiological signals by using blockchain helps the interoperable health system and has greater ability to access patients' physiological signals electronically.

5.1. Physiological Signals Data Sharing Model Based on Blockchain [22]

Intelligent contract based on blockchain is used to promote the security analysis and management of medical sensors. Intelligent device invokes intelligent contract and writes records of all events on blockchain. The intelligent contract systems support real-time patient monitoring and medical intervention by sending notifications to patients and medical professionals. The provider of medical records can modify the physiological signals, but it needs patient's consent, and the patient can assign access authority to medical records.

When applying blockchain to the construction of the credit system, we promote the collection and supervision of credit information in the medical field and build the new relationship platform. It is significant to the improvement of the credit system construction. According to the unified evaluation criteria, the credit rating is evaluated, the result of the rating level is publicized on the platform of block chain, the credit rating is rewarded, and the violation of credit is punished, so as to strengthen the construction of the credit system in the medical field in the real sense.

The asymmetric information encryption methods need two keys: public key and private key. After the physiological signals are encrypted with public key, only the corresponding private key can be used for decryption. On the contrary, if the private key is used to encrypt data, only the corresponding public key can be used for decryption. If the blockchain can be grafted, scientific research institutions understand the probability of disease occurrence, the occurrence of accidents, the level of hospital management, and claims cases and other actual situations.

Use the fog-based blockchain and fog warehouse to store medical data, as shown in Figure 10.

Figure 10

Blockchain based on fog warehouse.

List of medical research and patients in each institute.

Patients are authorized to access the entity set of each study. The entities are represented by the common part of the asymmetric key pair on the blockchain.

Definition study: the transaction builds the patient as the master of a UID which is the specific unique identifier and the source as the creator. Tuples stored in block chains are transactions with double signatures, similar to documents with signatures from patients and hospital representatives. The patients claim that the definition study has received the medical diagnosis in the hospital, which confirms the statement and promised to provide the study in the previous block. The patient's signature declaration is obtained through the mobile application, which shares and stores the values required allowing access to the transaction in the future. Then, the hospital signs the follow-up information of the patients and broadcasts the transaction to the blockchain.

Allow access: the transaction allows the owner of the medical information research to authorize the other party to retrieve its medical data. Patient KP signs a transaction to grant the function to doctor KD. The signed verification blocks are embedded in blockchains. As shown in Figure 11, patients publish the transaction after verifying the key with the doctor through the APP platform. The patient can be authorized to the legitimate doctor or institution, and the doctor can associate any medical information received with the correct local medical record number.

Figure 11

Blockchain medical data sharing sequence diagram based on fog computing.

The middle column (Block Chain Medical Data Sharing Sequence) describes the interaction between entities and judgments in each stage and reflects the sharing medical information by supporting distributed block chains and out-of-block transactions.

The actual medical data transmission requires the medical data receiver to deliver the signature request to the medical source's URL endpoint which creates the research. Both requests and responses are transmitted through the secure link of the transport layer to prevent eavesdropping. The effective blocks are generated in the timely manner by generating the distributed database with access permissions and stimulating the block generator in some way. Only those nodes with security deposits can participate in the expansion of the chain, and any node with misconduct will be forced to abandon its investment. The nature of blockchain provides the direct audit of the activity of each node such as the number of blocks generated and the failure status of the blocks generated. The node operator can prove the node ownership by using the private key which is corresponded to the identity public key of the node to sign the message. The enhanced model adds the fog computing in the original blockchain medical data sharing the sequence model [22], which is used to construct the blockchain for medical data sharing.

We have showed the technology fundamentals of blockchain and provided a summarization of the blockchain application that can be used as a tool to allow the patient-controlled, physiological signal's cross-domain sharing without the central authority. In particular, we highlighted the way blockchain satisfies many requirements of the interoperable health system. However, these technologies also have several important limitations, and the relative merits of existing alternatives must be considered before any large-scale and blockchain-based application for sharing physiological signals.

When receiving query request, the physiological signal data source verifies the correctness of the signature, ensures that the hashed data matches the previously published data for KP-owner via Block B, and confirms that the KP-owner has allowed the requestor access to these physiological signal data via Block C. If meeting all the conditions, the response containing the physiological signal study is returned from the source. In order to prevent eavesdropping, the requests and responses are sent to prevent eavesdropping. The specific steps of blockchain medical data sharing sequence diagram are as follows:

Step 1: for hospital (KH-owner), KH-owner will service physiological signals retrieval requests at https by using on-blockchain transaction and off-blockchain communication.

Step 2: for the patient (KP-owner), the physiological signals are acquired for KP-owner and KH-owner.

Step 3: for hospital, KP-owner's assertion is accrate and KH-owner shares the physiological signals at the established endpoint.

Step 4: for physician (KD-owner), KD-owner reviews the physiological signals from the hospitalization.

Step 5: for patient, if the patients agree, they are KP-owner and will allow access.

Step 6: for patient, the patient permit KD-owner to access the physiological signals that were acquired by KH-owner.

Step 7: physician uses the information in blocks (A, C) to submit the query request for physiological signals, signed by KD.

Step 8: hospital valid at the physician's signature, uses the data in blocks (B, C) to confirm authorization and transmits the physiological signals study in the query response. The requests are sent by the KD-owner at timepoint D.

The ecosystem is consisted of the blockchain nodes and fog storage. For example, one of the main reasons for incorporating fog storage technology into the ecosystem is to supply the offline storage solution, especially for large physiological signals. For security and privacy, the client side would encrypt the physiological signals uploaded to the fog storage. With the maturity of the fog storage, personal storage may be replaced by it.

Most significantly, blockchain technology can create the physiological signal-driven marketplace, where patients can get real return by offering their data to research institutions, pharmaceutical and consumer companies, the application development community, and producing new physiological signal data.

6. Conclusions

We extend the architecture of the IMD with blockchain, RFID, and WISP, which increases the physiological signal data's confidentiality and authenticity. The enhanced RFID protocols provide protection against tracking attacks, readers' impersonation attacks, and secret disclose attacks.

The physiological signal records have proved the importance for the patients, and sharing and acquiring physiological signals is essential for intelligent and advanced medical services. The blockchain application of e-commerce has proven that trusted and auditable transaction in peer-to-peer networking is possible. In the paper, we have introduced a blockchain-based architecture model for physiological signal data on fog computing environment. Our contributions are mainly consisted of the proposed solution and introduction to future medical data directions in blockchain. The paper proposes the outline to show the framework and schemas for dealing with heterogeneous physiological signals. Once the hybrid technologies are integrated, big data systems and AI technology have the potential to offer privacy protection and data sharing and transform healthcare management. In the future, we will focus on heterogeneous physiological signal data issues through fog computing, blockchain, and AI technology in the realistic medical environment.