Cross Domain Heterogeneous Signcryption Scheme with Equality Test for WBAN

The rapid development of wireless sensors has accelerated the popularity of wireless body area network (WBAN). WBAN use multiple sensors to collect the patient’s body data, and the data is transferred to the medical cloud for processing and analyzing. In order to protect the data in the medical cloud, some heterogeneous signcryption schemes that support equality test have been proposed. However, we observe that these schemes use the same cryptographic parameters in different cryptographic systems. In addition, most of these schemes cannot resist the replay attack (RRA) or know session temporary key attack (RKSTKA). To deal with these problems, this paper presents a cross domain heterogeneous signcryption scheme with equality test (CDSCET) for WBAN. In CDSCET, the ciphertexts are from certificateless cryptographic system to public key infrastructure, where two different cryptosystems use different cryptographic parameters. CDSCET can realize confidentiality, integrity, authentication, RRA and RKSTKA. Moreover, compared with three latest schemes, CDSCET has reduced the total computation cost by at least 56.46%.


Introduction
Wireless body area network is an advanced medical branch of wireless sensor network, which can help the doctor to monitor the physical condition of patients, analyze the body data and establish instant communications [1]. Normally, the cloud-assisted WBAN generates and uploads a great deal of data to the medical cloud (MC) [2,3]. However, the data in the MC is suffering many security problems, such as data tampering, eavesdropping, and so on. On the one hand, if any attacker invades into the WBAN system, the patient's private data will be exposed and causes economic losses. On the other hand, if the doctor receives tampered data, it will lead to misjudgment of the patient's disease, which in turn will endanger the life safety of patient. To address this challenge, several data transmission schemes and authentication protocols are proposed [4][5][6], which improve the security of WBAN.
In order to ensure the security of WBAN data, an effective method is to encrypt or signcrypt the WBAN data and upload it to the MC. However, this situation makes the data cannot be searched. To remove this obstacle, Boneh et al. [7] introduced the public key encryption scheme adopting keyword search (PKE-KS). The PKE-KS scheme makes the ciphertext searchable through the use of keywords. However, PKE-KS has a drawback that it needs to encrypt the plaintext under the same public key. To address the challenge, Yang et al. [8] designed the public key encryption scheme adopting equality test (PKE-ET). In PKE-ET, different public keys can be used to encrypt different ciphertexts, and the equivalence between them can be learned through the corresponding trapdoors. For the WBAN applications, Ramadan et al. [9] formulated a PKE-ET scheme that achieves low computation cost.
Authentication is vital for PKE-ET, and one of the ways to realize authentication is to use the digital signature. As proposed by Zheng et al. [10], signcryption allows digital signatures and data encryption to be performed at the same time, which greatly improving the efficiency. Subsequently, a signcryption scheme that supports equality test was formulated by Xiong et al. [11]. Besides, WBAN usually consists of different cryptographic systems. Hou et al. [12] proposed a heterogeneous signcryption scheme supporting equality test (HTSC-ET) from PKI to CLC for the Internet of Things. Xiong et al. [13] proposed a HTSC-ET scheme from PKI to Identity-based cryptographic system (IBC). However, the existing HTSC-ET schemes use the same cryptographic parameters in different cryptographic systems. Moreover, some of these schemes cannot realize RRA or RKSTKA (RRA means that the adversary cannot obtain the result of equality test by resending the previous ciphertext and corresponding trapdoor to the MC; RKSTKA means that the adversary cannot obtain the plaintext from the ciphertext and the session temporary key). To deal with these problems, a cross domain heterogeneous signcryption scheme supporting equality test with different cryptographic parameters that realize RRA and RKSTKA is required.

Related Works
A PKE-ET scheme for cloud-assisted IOV that realizes temporary delegation was proposed by Li et al. [14]. A PKE-ET scheme that improves efficiency and supports partial authorization was proposed by Lin et al. [15]. Deverajan et al. [16] formulated a PKE-ET scheme towards the IIOT, which uses the Near-Ring. Lin et al. [17] designed a pairing-free PKE-ET scheme with authorization. Recently, some researchers have integrated identity-based cryptographic system with PKE-ET (IBE-ET). An IBE-ET scheme with authorization for mobile applications was formulated by Hassan et al. [18]. An IBE-ET scheme towards the cloud medical service was designed by Xu et al. [19]. The IBE-ET scheme proposed by Xu can be resistant to off-line keyword guessing attacks. Furthermore, a certificateless public key encryption scheme with equality test for IIOT was designed by Elhabob et al. [20] that realizes fine-grained access control.
Alornyo et al. [21] introduced signcryption scheme with equality test into IBC. A latticed-based signcryption scheme with equality test under the standard model was proposed by Le et al. [22]. To apply signcryption and equality test functionality into the heterogeneous environment, Hou et al. [12] proposed an HTSC-ET scheme from PKI environment to CLC environment for the Internet of Things applications. Xiong et al. [13] presented an HTSC-ET scheme from PKI environment to IBC environment. In addition, Xiong et al. [23] designed an HTSC-ET scheme from IBC environment to PKI environment, which enables a flexible switch between public key encryption to heterogeneous signcryption. However, as far as the author knows, there is no HTSC-ET scheme that uses different cryptographic parameters in different cryptographic systems.

Our Contribution
The CDSCET is formulated in this paper. The main advantages of CDSCET are as follows: 1. CDSCET achieves cross domain heterogeneous with different cryptographic parameters.
However, the existing HTSC-ET schemes such as [12,13,23] use the same cryptographic parameters, which are not suitable for the cross domain heterogeneous WBAN environment. 2. CDSCET not only realizes confidentiality, integrity and authentication but also achieves RRA and RKSTKA. However, the existing HTSC-ET schemes [12,13,23] do not fully realize these security attributes. 3. In the signcryption and unsigncryption phase, CDSCET does not need any pairing operation. Compared with [12,13,23], the total computation cost of CDSCET is reduced by at least 56.46%.

Bilinear Pairing
Let G 1 be an additive cyclic group and G 2 be a multiplicative cyclic group. Suppose that G 1 and G 2 have the same prime order q . Following are the properties of the bilinear pairing ê ∶ G 1 × G 1 → G 2 : 1. Bilinearity For any R, S ∈ G 1 and x, y ∈ Z * q , ê(xR, yS) =ê(R, S) xy . 2. Non-degeneracy There exists an E ∈ G 1 that ê(E, E) ≠ 1 G 2 . 3. Computability Given any R, S ∈ G 1 , ê(R, S) can be calculated in polynomial time. Figure 1 shows the network model of CDSCET. Five entities make up the network model, including the KGC and WBAN nodes in the CLC environment, CA and doctor in the PKI environment, and MC. Note that every KGC and CA generates different cryptographic parameters. The WBAN nodes collect the body data from the patients and signcrypt it. After that, the signcrypted data is transmitted to the doctor and MC via the wireless networks. When MC receives the corresponding trapdoor from the doctor, it returns true to the doctor if the equality test is passed. Otherwise, MC returns false. The data transmission satisfies confidentiality, integrity and authentication, and is secure against RA and KSTKA.

Generic Model of CDSCET
Setup KGC generates its secret key a , its public key P pub and outputs the cryptographic parameters CGP 1 . Similarly, CA outputs the different cryptographic parameters CGP 2 .
CL-PPKG Given user's identity ID c in CLC, KGC generates the user's partial public key ppk c .
CL-PSKG Given user's identity ID c in CLC, KGC generates the user's partial private key psk c .
CL-ASV Given CGP 1 , a user in CLC selects its secret value x c . CL-PKG Given x c , a user in CLC generates its public key pk c . PKI-KGN Given CGP 2 , a user in PKI outputs its public key pk p and private key x p . SC A sender in CLC executes this algorithm to signcrypt a plaintext m to a receiver in PKI.
USC The receiver in PKI performs this algorithm to unsigncrypt the ciphertext. Trapdoor The receiver in PKI performs this algorithm to generate a trapdoor td. Test Given a medical record and a search content with the corresponding trapdoors, this algorithm returns true if the equality test is passed. Otherwise, this algorithm returns false.

Security Model of CDSCET
This section describes the security model of CDSCET. Let B denotes the challenger. Two kinds of adversaries E i(i=1,2) are defined. KGC's master private key cannot be obtained by E 1 , but E 1 can replace the public key. Meanwhile, KGC's master private key can be obtained by E 2 , but E 2 cannot replace the public key. Definition 1 If in the following game, each polynomially bounded adversary E i could win with a negligible advantage, CDSCET owns indistinguishability against adaptive chosen ciphertext attacks (IND-CDSCET-CCA2).

Game 1
Initialization B executes Setup algorithm and sends the cryptographic parameters Phase 1 E i makes the following queries: CL-PKG queries Given an identity ID , B performs CL-PPKG and CL-PKG algorithm to return E i the public key (ppk ID , pk ID ).
CL-SKG queries Given an identity ID , B performs CL-PSKG and CL-ASV algorithm to return the private key psk ID to E i .
CL-RPK queries Given a valid public key, the corresponding public key is replaced. PKI-PKG queries Given an identity ID , B performs PKI-KGN algorithm to return the public key pk ID to E i . PKI-SKG queries Given an identity ID , B performs PKI-KGN algorithm to return the private key x ID to E i . SC queries Given the plaintext m and ( ID s , ID r ), B performs SC algorithm to return to E i . USC queries Given ( ID s , ID r ) and a ciphertext , B performs USC algorithm and returns the result to E i .
Trapdoor queries Given ( ID s , ID r ) and , Trapdoor algorithm is executed by B and the trapdoor is returned to E i .
Challenge E i sends identities ( ID * s , ID * r ) and two plaintexts (m 0 , m 1 ) to B . B chooses ∈ {0, 1} , performs SC algorithm with m and returns * to E i . Phase 2 E i makes the same queries as those in Phase 1. However, B rejects if receiving a PKI-SKG query of ID * r or a USC query of ( * , ID * s , ID * r ).
Definition 2 If in the following game each polynomially bounded adversary A could win with a negligible advantage, then CDSCET is existentially unforgeable against any adaptive chosen message attacks (EUF-CDSCET-CMA).

Game 2
Initialization B executes Setup algorithm and sends the cryptographic parameters Probing The queries are the same as those in Definition 1. Forgery A sends identities ( ID * s , ID * r ) and * to B . If the following conditions are hold, A wins the game: The SC query of * is not performed.

CDSCET
This section demonstrates the concrete scheme and its correctness. Figure 2 shows the CDSCET.

Construction
Setup KGC picks G 1 , G 2 with the same prime order q 0 . Let P 1 denotes the generator of G 1 .
Then KGC generates its secret key a ∈ Z * q 0 , its public key P pub = aP 1 , and defines five hash functions: After that, KGC outputs the cryptographic parameters , pk m = x m P 2 and sends {x m , pk m } to MC. Then CA outputs the cryptographic parameters CGP 2 = {G � 1 , P 2 , pk m , q 1 }. CL-PPKG Given the identity ID c of a user in CLC, KGC selects b c ∈ Z * q 0 and computes the user partial public key ppk c = b c P pub .
CL-PSKG After performing CL-PPKG algorithm, KGC computes psk c = ab c c , where as its secret value. CL-PKG After performing CL-ASV algorithm, the user calculates pk c = x c P 1 as another part of its public key.
PKI-KGN Given the cryptographic parameters CGP 2 , a user with identity ID p in PKI environment selects its private key x p ∈ Z * q 1 and computes its public key pk p = x p P 2 . SC. A sender in CLC signcrypts the plaintext m to a receiver in PKI as follows: USC Given = (C, S, T) , the receiver in PKI unsigncrypts the ciphertext. If the ciphertext is valid, this algorithm outputs plaintext m . Otherwise, this algorithm outputs failure symbol ⊥. Trapdoor In PKI, receiver chooses Then it computes the trapdoor td = (t 2 , t 3 ) , where t 3 = (f 1 ||ts) ⊕ H 1 (t 1 ) and ts is the current timestamp.
is the medical record in the MC, Z y = S y is the search content, and td x , td y = (t 2 y , t 3 y ) = (t 2 y , (f 1 y ||ts y ) ⊕ H 1 (t 1 y ) are the corresponding trapdoors. To prevent RA, after getting the query (Z y , td y ) from a user, MC calculates (f 1 y ||ts y ) = t 3 y ⊕ H 1 (x m t 2 y ) and checks if ts � − ts y <▵ ts , where ▵ ts is an appropriate period of time and ts ′ is the current timestamp. If so, MC checks the equation . If the condition holds, which means the plaintext m x = m y , then MC returns true. Otherwise, MC returns false.

Correctness
and

Definition 2
Discrete Logarithm Problem (DLP): Given (P, yP) where y ∈ Z * q , P ∈ G 1 , it is difficult to compute y.

Theorem 1 Assume that DDHP is intractable, in ROM CDSCET is indistinguishable against any IND-CDSCET-CCA2 adversary E 1 .
Proof Assume that the instance of DDHP is (P, jP, kP, F) . The process of challenger B uses E 1 to distinguish jkP from F is as follows:

Game 1
Initialization B performs the Setup algorithm, generates the secret key {a, x m , pk m } and outputs the cryptographic parameters CGP 1 , CGP 2 .
Phase 1 B will maintain several lists L i(i=1 ∼ 5) to record H i(i=1 ∼ 5) queries. Meanwhile, B will maintain LK p and LK c to record the private key queries of PKI and CLC, respectively.
H 1 queries Given ppk i as input, B searches L 1 for (ppk i , i ) . If the tuple is in L 1 , B answers i to E 1 . Otherwise, B selects i ∈ Z * q , inserts (ppk i , i ) to L 1 and returns i to E 1 . H 2 queries Given (f 1 , f 2 , S i , m) as input, B searches L 2 for (f 1 , f 2 , S i , m, k i ) . If the tuple is in L 2 , B answers k i to E 1 . Otherwise, B selects k i ∈ Z * q , inserts (f 1 , f 2 , S i , m, k i ) to L 2 and returns k i to E 1 .
H 4 queries Given (f 2 , S i , C i ) as input, B searches L 4 for (f 2 , S i , C i , l i ) . If the tuple is in L 4 , B answers l i to E 1 . Otherwise, B selects l i ∈ Z * q , inserts (f 2 , S i , C i , l i ) to L 1 and returns l i to E 1 .
H 5 queries Given m i as input, B searches L 5 for (m i , h 5 ) . If the tuple is in L 5 , B answers h 5 to E 1 . Otherwise, B selects h 5 ∈ Z * q , inserts (m i , h 5 ) to L 1 and returns h 5 to E 1 . CL-PKG queries: Given ID i as input, B searches LK c for (ID i , x i , pk i , psk i , ppk i ) firstly. If the tuple is in LK c , B returns the public key (pk i , ppk i ) . Otherwise, B picks x i , b i ∈ Z * q , computes pk i = x i P , ppk i = b i P pub , i = H 1 (ppk i ) and psk i = ab i i , inserts the tuple (ID i , x i , pk i , psk i , ppk i ) to LK c and returns the public key (ppk i , pk i ) to E 1 .
CL-SKG queries Assume that E 1 makes a CL-PKG query on ID i previously, so LK c contains (ID i , x i , pk i , psk i , ppk i ) . Given ID i as input, B searches LK c for (ID i , x i , pk i , psk i , ppk i ) and returns the private key (x i , psk i ) to E 1 .
CL-RPK queries Given a valid public key pk * i , B updates the tuple

PKI-PKG queries
Suppose E 1 makes this query q p > 0 times at most. B picks an identity ID ∈ {1, 2, … q p } . Given ID i as input, if ID i = ID , B sets x = ⊥, pk = jP . If ID i ≠ ID , B searches pk i from LK p . If LK p does not contain the tuple (ID i , x i , pk i ) , B picks x i ∈ Z * q and calculates pk i = x i P . Finally, B inserts (ID i , x i , pk i ) to LK p and returns the public key pk i to E 1 .
PKI-SKG queries Assume that E 1 makes a CL-PKG query with identity ID i previously, so LK p contains (ID i , x i , pk i ) . B searches LK p for (ID i , x i , pk i ) and returns the private key x i to E 1 .
SC queries Given (ID s , ID r ) of sender and receiver, and a plaintext m , B searches (x c , psk c , pk p ) from LK c and LK p , performs SC algorithm and returns = (C, S, T) to E 1 .
USC queries Given (ID s , ID r ) of sender and receiver, and = (C, S, T) . B searches (x c , ppk c , x p , c ) from L 1 , LK c and LK p , performs USC algorithm and returns the result to E 1 .
Trapdoor queries Given ID r of receiver, and = (C, S, T) . B searches x p from LK p , performs Trapdoor algorithm and returns the result to E 1 .
Challenge E 1 outputs (ID * s , ID * r ) and two plaintexts (m 0 , m 1 ) . Note that the private key of ID * r cannot be queried during Phase 1.
. Finally, B returns * = (C * , S * , T * ) to E 1 . Phase 2 E 1 makes the same queries as in Phase 1. But B rejects the PKI-SKG query of ID * r and the USC query of ( * , ID * s , ID * r ). Guess E 1 outputs ′ . If � = , E 1 wins the game, and B can get the solution of DDHP as F = f * 2 = jkP . So E 1 can break DDHP with a non-negligible advantage. However, so far there does not exist any efficient algorithm that can solve DDHP. Therefore, CDSCET can achieve confidentiality.

Theorem 2 Assume that DDHP is intractable, in ROM CDSCET is indistinguishable against any IND-CPDSPHS-CCA2 adversary E 2 .
Proof E 2 and B play a game similar to that of Theorem 1, but E 2 is not allowed to make CL-RPK or CL-SKG queries. If E 2 wants to obtain , it needs to compute the encryption key f * 2 . Because E 2 does not know the private key x c of sender, it is also facing DDHP. Therefore, CDSCET is indistinguishable against any IND-CPDSPHS-CCA2 adversary E 2 .

Theorem 3 Assume that DLP is intractable, in ROM CDSCET is existentially unforgeable against every EUF-CDSCET-CMA adversary A.
Proof Assume that DLP's instance is (P, yP) . Challenger B uses A to get y is as follows:

Game 2
Initialization B performs the Setup algorithm, generates the secret key {a, x m , pk m } and outputs the cryptographic parameters CGP 1 , CGP 2 .
Probing A makes some queries including H 1 , H 2 , H 3 , H 4 , CL-RPK, CL-SKG, PKI-PKG and PKI-SKG queries as those in Theorem 1. The CL-PKG queries are as follows: CL-PKG queries Suppose A makes this query q c > 0 times at most. B picks an identity to LK c and returns the public key (pk i , ppk i ) to A.
Forgery A outputs (ID * s , ID * r ) of sender and receiver and * = (C * , S * , T * ) . Note that the private key of ID * s cannot be queried and * cannot be generated by SC query. If ID * s ≠ ID , B aborts. Otherwise, B computes f * 2 = x * r S * and l * = H 4 (f * 2 , S * , C * ) . Using the forking lemma 12, another valid signcryption � = (C � , S � , T � ) is generated and B can get the answer of DLP just as follows: Hence, A owns a non-negligible advantage over DLP. Until now however, there hasn't been an efficient algorithm that is able to solve DLP. Therefore, CDSCET can achieve unforgeability.

RRA
Assume that an adversary intercepts the ciphertext = (C, S, T) and the trapdoor td = (t 2 , t 3 ) = (t 2 , (f 1 ||ts) ⊕ H 1 (t 1 )) , it submits with td to the MC as a RA. If MC lacks a verification of timestamp, the result of equality test will be sent to the adversary. In CDSCET, MC will check the timestamp ts and reject this malicious query. Therefore, CDSCET can realize RRA. However, HHC [12] and XHH [23] cannot realize RRA. In HHC [12], the cloud server directly performs the equality test after receiving the ciphertext C and trapdoor td . Similarly, in XHH [23], the cloud server performs the equality test algorithm without verification.

RKSTKA
Assume that an adversary gets the temporary key d 1 and the ciphertext = (C, S, T) . The encryption key is H 3 (f 2 , S) , where f 2 = x c d 1 (pk p ) . The adversary cannot calculate the encryption key because it cannot calculate the secret value x c . Therefore, CDSCET can realize RKSTKA. However, HHC [12] XZH [13] and XHH [23] cannot realize RKSTKA. In HHC [12], the encryption key is H 3 (J 1 , t 1 PK i,1 ) , where J 1 =ê(P pub , U i,1 ) t 1 . Because U i,1 , PK i,1 , P pub are public values, if the adversary can get the temporary key t 1 , it can compute the encryption key H 3 (J 1 , t 1 PK i,1 ) . The encryption key of XZH [13] is H 3 (r 1 ) , where Because g is public, if the adversary can get the temporary key x 1 , it can compute the encryption key H 3 (r 1 ) . In XHH [23], the encryption key is Because t is public, if the adversary can get the temporary key u 1 , it can compute the encryption key H 4 (U 2 ).

Efficiency Analysis
In this section, we compare the performance and security of our CDSCET with HHC [12] XZH [13] and XHH [23]. For convenience, Table 1 demonstrates the meaning of different symbols. Besides, the experiment platform is similar to 13: a PC running Windows-10 system, PBC library, 3.60 ghz CPU and 8 gb memory. In addition, Table 2 illustrates the computation cost of different operations.  We compare our scheme with [12,13,23] in terms of computation cost and communication overhead in Table 3. Figure 3 demonstrates the computation cost of different schemes. From Fig. 3, in the signcryption phase, when compared with [12,13,23], our scheme reduces the computation cost by 90.67%, 43.23% and 81.1%, respectively. In addition, in the unsigncryption phase, our scheme performs better than [12,13,23], which reduces the computation cost by 89.4%, 92.57% and 92.4%, respectively. Moreover, in the equality test phase, compared with [12,13,23] our scheme reduces the computation cost by 50.87%, 6.6% and 51.7%, respectively. From Tables 2 and 3, the total computation cost of different schemes is shown below: HHC [12]:   In conclusion, the communication overhead of our scheme is reduced by at least 34.9% when compared with [12,13,23]. Table 4 illustrates the comparison of security between different schemes. Let "Y" denotes that the scheme has achieved the corresponding security attribute, and "N" indicates that the attribute is unrealized. From Table 4, compared with HHC [12] and XHH [23], our scheme achieves RRA. Besides, our scheme is the only scheme that realizes RKSTKA when compared with [12,13,23]. The specific analysis of RRA and RKSTKA is described in Sect. 4.3 and 4.4. In addition, our scheme is not affected by the key escrow problem. Because in XZH [13] the user uses IBC system, and in XHH [23] the sensor and user use IBC system, both of them are suffer from the key escrow problem. Moreover, only our scheme uses different cryptographic parameters in different cryptographic systems. In [12,13,23], the sender and receiver use the same cryptographic parameters. Because of their limitations, each security domain cannot independently control its parameters and has to negotiate and share parameters with other domains, which diminishes their practicality.
In conclusion, when compared [12,13,23] our scheme achieves more security attributes and higher efficiency, therefore is more suitable for the cross domain heterogeneous WBAN environment.

Conclusion
This paper presents a cross domain heterogeneous signcryption scheme with equality test for WBAN. In our CDSCET, the WBAN node in CLC environment can signcrypt the body data to the doctor in PKI environment. Meanwhile, MC can execute the equality test to compare different medical records through the corresponding trapdoors and return the result to the doctor. In ROM, CDSCET is able to achieve confidentiality and unforgeability under DDHP and DLP. Moreover, CDSCET achieves RRA and RKSTKA. Through the efficiency analysis, when compared to [12,13,23], the total computation cost of CDSCET has reduced by at least 56.46%. Therefore, CDSCET is more suitable for cross domain heterogeneous WBAN environment.
Author Contributions ML and YP wrote the main manuscript text, MQ prepared Tables 1, 2 Funding This study was funded by the National Natural Science Foundation of China (grant number 62262041) and postgraduate innovation foundation of Nanchang University (No. YC2021-S167).
Data Availability Data sharing not applicable to this article as no datasets were generated or analysed during the current study.

Conflict of interest
The authors have no competing interests to declare that are relevant to the content of this article.

Ethical Approval and Consent to Participate Not applicable.
Human and Animal Rights This article does not contain any studies with human participants or animals performed by any of the authors.

Consent for Publication Not applicable.
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
Ming Luo received the Ph.D degrees in computer application technology from Northeastern University, Shenyang, China in 2010. He is currently a professor and the deputy dean of the School of Software, Nanchang University, Nanchang, China. His research interests cover Internet of Things, Cyberspace Security and cryptography.
Yusi Pei received the B.S. degree from Yichun University, Yichun, China in 2020. He is currently pursuing the M.S. degree with the School of Software of Nanchang University, Nanchang, China. His research interests cover Internet of Things, Cyberspace Security and cryptography.
Minrong Qiu received the Ph.D degree in industrial economics from Wuhan University of Technology, Wuhan, China in 2018. She is currently a associate professor of the GongQing Institute of Science and Technology, Nanchang, China. Her research interests cover Internet of Things, Cyberspace Security and Information System Management.