Abstract
As a part of the smart city revolution, crowd management is an emerging trend and it can enhance the quality of life. Unmanned ariel vehicles (UAVs) can help in making the crowd management process more efficient and more accurate. UAVs can monitor and collect environmental-related surveillance data and share real-time information with each other and with the decision makers. However, the battery-operated UAVs communicate over the open public channel making the privacy and security of the UAVs a crucial element in mission-critical applications. The weaknesses of the existing scheme pave the way to design a new lightweight authentication scheme for UAV environments. In this article, we present a symmetric key primitive-based scheme and provide authentication among a user and a UAV through an intermediate control center. Due to usage of symmetric key and elliptic curve cryptography, the proposed scheme fulfils the performance requirements of the UAVs. The security of the proposed scheme is substantiated through BAN logic, along with a discussion on security features extended by the proposed scheme. The performance and security comparisons show that the proposed scheme provides adequate security and efficiency and can be practically deployed in real UAV environments.
1. Introduction
With the advancements in software and hardware infrastructure for information and communication technologies, the UAV communication has become a reality, which is making daily life more easy and automated. The UAVs, also called drones, can be deployed in a variety of applications including smart cargo and surveillance. The UAVs can enhance quality of life and can be deployed at remote and inaccessible locations like depths and mountain peaks. In contrast to traditional roadways, the UAV/drone can reach inaccessible locations in a very speedy manner [1–3]. Moreover, in many emergency-like situations, the traditional transportation and surveillance could not respond in a rapid way and the slow response can cause irreparable losses like lives. Initially, developed for military operations, the UAVs can be very beneficial in many applications including smart agriculture, surveillance, goods delivery operations, and so on [4]. UAVs have many properties similar to IoT devices and have sensors, transmitters, and receivers for communication with the other entities including humans [5]. The UAVs collect specific environment data for initiating the decision-making process [6]. Reddy et al. [4] provided a good survey of the usage of UAVs in agriculture and related fields for interested readers. Among many other applications, UAVs can be adopted for crowdsourcing systems [7]. However, very similar to generic IoT devices, the UAVs are battery powered and can be used by dishonest attackers for deceitful intentions, and such unintended usage can be harmful for decision making. The attacker can trace user locations and can effect the quality of services as well as forge the collected data and deviate the UAVs from their designated task [8]. Therefore, the communication among several UAVs and users need to be secured from unintended receivers. The following subsections briefly explain the adopted system architecture and adversarial model for the UAV-based crowdsourcing systems.
1.1. System Architecture
A typical UAV-based crowd monitoring system is shown in Figure 1. The UAVs are deployed at several geographical monitoring locations, where each geographical location can be monitored by one or more UAVs which are connected to a ground station (GS). All GSs are linked with the UAV (trusted) control center . The proposed architecture also contains the cloud edge and a monitoring system to facilitate monitoring-based decision making through human interaction. The UAVs monitor the crowd and send the sensed information to the respective GS. The GS sends these data to the control center and finally this information is sent to the crowd monitoring system. The GS acts as an intermediate entity that accumulates and sends the UAV sensed data, whereas UAV, TC, and the user at monitoring system are responsible for the confidentiality of the sensed data during transmission and authenticity of the communicating entities, i.e., user, TC, and UAV.
1.2. Adversarial Model
The communication among UAVs and other devices including user and is carried out on the insecure public channel. Therefore, UAV communication has an attractive infrastructure for the attackers for initiation of forgery. As per the widely accepted adversarial model, DY model (Dolev–Yao model) [9], the attacker can listen, modify, replay, and stop a message exchanged among two entities of a UAV network [10–14]. The attacker can try to forge any message sent from any of the entities including user, , and UAV. In addition to the DY model, the strong de facto CK adversarial model [15] has got more attention from the researchers. In this paper, we consider the CK model with an adversary having all capabilities of the adversary considered in DY model; in addition, the adversary can compromise any one of the long-term secrets, ephemeral secrets, and session keys. Therefore, in the CK model, the session keys should be formed using long and short-term keys. Moreover, the session keys should be statistically independent of each other, and the compromise of a session key may not affect any other. Moreover, as per [16, 17], if an attacker captures a drone or user mobile device, they can extract parameters stored in the captured/stolen device. In this paper, we consider an extended CK (eCK) adversarial model [18], where an attacker can also launch a key compromise impersonation (KCI) attack.
2. Related Work
To secure communication among the users and UAVs/relevant devices, some access control/authentication protocols are proposed. For securing user-UAV communication, a temporal credential-based framework is proposed using symmetric primitives [19]. The scheme of Srinivas et al. [19] lacks resistance to impersonation based on stolen verifier; also, their scheme lacks anonymity as proved in [20]. The scheme proposed by Ali et al. could not resist ephemeral secret leakage attacks. The scheme proposed by Zhou et al. [21] in a distributed IoT environment using pairing was proved as insecure to IoT device forgery attacks [22]. In 2019, Wazid et al. [23] also proposed a new protocol for securing user-UAV communication through symmetric key primitives. As debated by Hussain et al. [3], Wazid et al.’s scheme is weak against stolen verifier-based forgery of users, UAVs, and trusted control centers. The scheme of Wazid et al. also lacks untraceability. Another scheme was proposed by Zhang et al. in 2020 by using symmetric primitives [24], which also lacks forward secrecy and is weak against insiders and stolen verifier attacks. Zhang et al. managed anonymity using a parameter , which is generic for all users and any dishonest user can break the anonymity and untraceability of the user requesting for login. Moreover, in Zhang et al.’s scheme, the user credentials are not verified appropriately. Therefore, the login request can be sent to the even when wrong credentials are entered. In 2020, Bera et al. proposed a scheme for UAV network using blockchains and certificates constructed upon elliptic curve-based cryptography (ECC) [25]. However, their scheme was debated for having critical weaknesses against man-in-the-middle, replay, and impersonation attacks and lack of anonymity by Chaudhry et al. [2]. In the same year, i.e., 2020, Bera et al. proposed another scheme [26] using blockchains and certificates using ECC. Unfortunately, Bera et al.’s scheme does not extend anonymity owing to the use of static identity of the UAV/user. Likewise, as proved by Irshad et al. [27], there is no verification of the signatures generated by the ground station from the UAV in Bera et al.’s second scheme [26]. It was also debated in [27] that owing to the similar identity for all sessions, the scheme of Bera et al. [26] does not extend UAV anonymity. Another scheme using symmetric key and elliptic curve cryptography was proposed recently by Nikooghadam et al. [28]. The related works are summarized in Table 1, which gives the properties of existing methods along with their weaknesses and cryptographic method used.
2.1. Motivation and Contributions
In the recent past, many authentication schemes were designed for securing the communication among UAV, user, and . However, some of these schemes do not offer performance efficiency and some other schemes suffer from one or many insecurities. In this connection, very recently in 2021, Nikooghadam et al. proposed a symmetric key-based lightweight authentication scheme to secure the communication among users, UAV, and [28]. However, it is proved in the coming sections/subsections of this paper that Nikooghadam et al.’s scheme has many insecurities and cannot be deployed in real-world scenarios. The multi-fold contributions of this study include the following:(i)Initially, we highlight the importance of secure communication among entities of UAV environments.(ii)We revisited the authentication scheme proposed by Nikooghadam et al. and proved that Nikooghadam et al.’s scheme has vulnerabilities against stolen verifier attacks and lacks user anonymity and untraceability. It is also proved that Nikooghadam et al.’s scheme is not practical due to the exposure of secret parameters of the UAVs and the users.(iii)A security-enhanced authentication scheme is proposed in this paper using only the lightweight symmetric key elliptic curve-based cryptography.(iv)We used BAN logic and informal discussion on security properties for proving the security of the proposed scheme.(v)By using MIRACL library, we set up a real-time environment, where we used a smartphone Xiaomi-Redmi Note-8 for replicating the user mobile device, Raspberry Pi3B + Cortex (A53-ARMv8) to replicate a UAV, and HP-EliteBook 8460-P to serve as the control center . We implemented the proposed scheme’s primitives on these three devices for computation of the execution time of an authentication round.(vi)Finally, a comparative study among proposed and related schemes based on performance and security features is conducted.
3. Revisiting the Scheme of Nikooghadam et al
The following subsections provide details on different phases of Nikooghadam et al.’s scheme for extending the authentication and session key among a user and a UAV through the control center. The notations used for the technical details of this paper are defined as per Table 2.
3.1. Initialization
The initialization is performed by the control center by selecting an elliptic curve , where , , and . Now, selects a base point and a private key . also selects a hash function , where . Finally, secretly stores and publishes .
3.2. User Registration
For completion of the registration, selects an identity and passwords and a random number . now computes and sends to . On reception of , the checks the availability of and availability of , and the randomly selects and computes , , , and and transmits to securely, in addition to storing the tuple in its verifier table. On receiving , stores these parameters in the memory of mobile device.
3.3. UAV Registration
For completion of the registration, selects and transmits an identity to . On reception of , the checks the availability of . On availability of , the randomly selects and computes and and transmits to securely, in addition to storing the pair in its verifier table. On receiving , stores the pair in its memory.
3.4. Login and Authentication
For login and authentication, the user inputs the tuple , and the mobile device computes , , and and checks . On success, the following steps as depicted in Figure 2 are performed between , , and . ZDA 1: . generates and computes , , and sends to . ZDA 2: . on receiving checks the freshness of by checking . On proven freshness, the extracts the related tuple and checks . On success, generates and computes and . The transmits to . ZDA 3: . on receiving checks the freshness of by checking . On proven freshness, computes and checks . On proven validity, generates and computes , , , and . Lastly, sends to . ZDA 4: on receiving checks the freshness of by checking . On proven freshness, computes and the session key and checks the validity of session key by verifying . On success, keeps the as session key authenticated with .
4. Weaknesses of Nikooghadam et al.’s Scheme
This section explains the weaknesses of the scheme of Nikooghadam et al. [28] against secret key exposure and stolen verifier attack. Both of these attacks are very critical and common and render the scheme of Nikooghadam et al. inapplicable and impractical.
4.1. Stolen Verifier Attack
In the scheme of Nikooghadam et al., the trusted control center stores two separate verifier tables consisting of tuple , where and , where each for users ( numbers of users) and UAVs ( number of UAVs), respectively. As per the adopted CK adversarial model, the tables stored in the memory of TC can be exposed to the attacker. The whole authentication process can be compromised if these verifier tables are exposed to an attacker. An attacker with a verifier relating to users can impersonate on behalf of all users and an attacker with a verifier relating to the UAVs can impersonate on behalf of all UAVs. Therefore, the scheme of Nikooghadam et al. is not practical and is subject to stolen verifier attacks.
4.2. Lack of Anonymity and Untraceability
In the scheme of Nikooghadam et al.’s scheme, the user sends a pseudo-identity , which remains the same for all sessions. Therefore, Nikooghadam et al.’s scheme lacks anonymity and untraceability.
4.3. Secret Parameter Exposure
In the scheme of Nikooghadam et al., let user be a dishonest user. can initiate the login/authentication request, and for this, transmits to and transmits to . The dishonest user while listening to the channel can receive . can compute the secret parameter of as follows:
It is very clear that through equation (1), the dishonest user has computed the secret parameter of the UAV .
4.3.1. Related Attacks Based on Secret Parameter Exposure
Once the secret parameter of a UAV is revealed to a dishonest user say , the now using can impersonate not only on behalf of but can also use to extract the secret parameter of any user requesting a session with . Therefore, based on the secret parameter exposure attack, the malicious user can impersonate on behalf of any user requesting login with the . Similarly, the malicious user can extract secret parameters of several or even all of the UAVs’ , and this can also lead to exposure of the secret parameters of all the registered users and the malicious user can impersonate on behalf of all the UAVs as well as all the users. Therefore, the scheme of Nikooghadam et al. cannot be practically deployed in any environment due to its weaknesses against the secret parameter exposure.
5. Proposed Scheme
In this section, we propose our improved scheme.
5.1. Initialization
The initialization is performed by the control center by selecting an elliptic curve , where , , and . Now, selects a base point , a private key , and a public key . also selects hash functions , where , and a block cipher encryption/decryption algorithm . Finally, secretly stores and publishes .
5.2. User Registration
For crowd monitoring, a user has to register with the system. After registration, a user at monitoring system can authenticate a UAV through TC and can receive and interpret the crowd management-related data. For completion of the registration, selects an identity and passwords and computes . Now, sends to . On reception of , checks the availability of and availability of , and randomly selects and computes , , and and transmits to securely, in addition to storing the tuple in its verifier table. On receiving , stores these parameters in the memory of mobile device. The registration process of the user is furnished through secure/private channel.
5.3. UAV Registration
For completion of the registration, (the UAV) selects identity and sends it to . On reception of , checks the availability of . On availability of , randomly selects and computes and transmits to securely, in addition to storing the pair in its verifier table. On receiving , stores in its memory and publishes . Like user registration, the registration process of the UAV is furnished through secure/private channel.
5.4. Authentication
For login, the user inputs the tuple , and the reader computes and and checks . On success, continues for authentication phase as depicted in Figure 3 detailed as follows: IDA 1: . Initially, generates and computes , , and using (x-coordinates of ) encrypts , i.e., . Now, computes and sends to the trusted center . IDA 2: . on receiving checks the freshness of by checking . On proven freshness, computes and decrypts and gets . now verifies the validity of , and on proven validity, computes and verifies the validity of . On success, computes and generates . The now computes , , and and transmits to . IDA 3: . on receiving checks the freshness of by checking . On proven freshness, decrypts and gets . Now, verifies the validity of . On proven validity of , generates and computes , , and session key . further computes and transmits to the . IDA 4: . on receiving checks the freshness of by checking . On proven freshness, checks , and on proven validity, computes . Now, generates and computes , . finally sends to . IDA 5: on receiving checks the freshness of by checking . On proven freshness, computes , and the session key . Finally, computes and checks the validity of session key by verifying . keeps the as session key authenticated with .
6. Security Analysis and Discussion
In the following subsections, we prove the security of the proposed scheme formally as well as provide discussion on the attack resilience of the proposed scheme.
6.1. Formal Security Analysis through BAN
We demonstrate the formal security analysis of the contributed work using Burrows–Abadi–Needham logic (BAN) logic [31]. In this model, we conduct the security analysis with the consideration of session key protection and distribution along with mutual authenticity between the legal participants. Some related notations in this analysis are explained below:(i) The principal S believes .(ii) sees .(iii) once said , and S believes it to be true.(iv) has jurisdiction over .(v): is not replayed and is fresh.(vi): or are parts of a message.(vii): using , or are encrypted through symmetric encryption.(viii): the communication among and is secured using as the key.
Some related rules employed in this analysis are given as follows: Rule 1: message meaning rule: Rule 2: nonce verification rule: Rule 3: jurisdiction rule: Rule 4: freshness conjunction rule: Rule 5: belief rule: Rule 6: session key rule:(i)G-1: .(ii)G-2: .(iii)G-3: .(iv)G-4: .(v)G-5: .(vi)G-6: .
The idealized form of exchanged messages is given as follows:(i)M1: (ii)M2: (iii)M3: (iv)M4:
Next, the following premises have been constructed to prove the model.(1)(2)(3)(4)(5)(6)(7)(8)(9)(10)(11)(12)(13)(14)(15)
Now we utilize the above idealizations in the following formulations considering and of the idealized formalization:(i)M1: (ii)M2:
Using seeing rule for M1 and M2, we get(1)(2) According to and message meaning rule, we get(3)(4) Using D3, W1, freshness conjunction, and nonce verification rules, we get(5) Using D4, W2, freshness conjucatenation, and nonce verification rules, we get(6) Using D5, W12 and jurisdiction rule, we get(7) Using D6, W14 and jurisdiction rule, we get(8) Using D5, D7 and session key rule, we get(9) (G-1) Using D5, D7, W6, W8 and nonce verification rule, we get(10) (G-5) Consider M3 of the idealized form:(11)M3: By applying seeing rule for M3, we get(12) Using D11, W7 and message meaning rule, we get(13) Using D12, W3, W13, freshness conjucatenation, and nonce verification rules, we get(14)(15) (G-3)(16) (G-4) Next, considering M4 idealized form, we get(17)M4: By applying seeing rule for M4, we get(18) Using D14, W4, W5, W11 and message meaning rule, we get(19) By using D15, W2, W3, freshness conjucatenation, and nonce verification rules, we get(20) Using D16, W4, W10, W15 and jurisdiction rule, we get(21) According to D17, we apply the session key rule as(22) (G-2) Using D18, W2, W14, we apply the session key rule as(23) (G-6)
This analysis sufficiently proves that the proposed model achieves the specified goals by attaining mutual authenticity among the involved participants.
6.2. Informal Security Analysis
An informal security discussion on the security features of the proposed scheme is provided in the following subsection.
6.2.1. Mutual Authentication
In the proposed model, and entities mutually authenticate one another. This is because when submits the authentication request message towards , the latter verifies the freshness of timestamp , and on successful verification, it computes and to further calculate and match against . If the equation holds valid, confirms the authenticity for . Similarly, after receiving the message checks the freshness of timestamp. Next, it computes using shared secret and compares it against to completely verify . Likewise, communicates the parameter to through . Then, verifies the authenticity of by computing , , and . Then, it further computes and verifies the equality check against . If this holds true, validates the authenticity for . Hence, in the proposed scheme, both entities and mutually authenticate each other with the help of a trusted intermediary, i.e., .
6.2.2. User Anonymity
In the proposed scheme, remains anonymous while submitting the authentication request to . submits the message to without disclosing the identity as plaintext. Since encrypts its identity as using , Ci can only be decrypted by a legitimate . Moreover, the scheme supports untraceability since the authentication messages in our scheme do not bear any message parameter that remains constant among various sessions that could help the adversary in identifying the location of the user. Hence, our scheme ensures anonymity as well as untraceability for the user.
6.2.3. Impersonation
The proposed scheme is immune to impersonation attack. In case an adversary attempts to construct a fake authentication request message with a fresh timestamp , it would not be able to do this until it has access to secret key which can only be computed by valid and can never be guessed or accessed by the adversary under ordinary circumstances. Hence, our scheme can resist impersonation attack.
6.2.4. Man-in-the-Middle Attack
In the proposed scheme, no malicious entity can engage in the ongoing communication session of the legal participants. This is because of the fact that both participants such as and are deployed with secret keys, i.e., , , respectively, during the initialization phase by the trusted . Thus, all participants having possession to those secrets may engage in mutual authentication process and construct the mutually agreed session key on legal basis. Hence, the proposed scheme is very much immune to man-in-the-middle attack.
6.2.5. Session Key Security
The contributed protocol supports session key security because it supports mutual authentication and can resist man-in-the-middle as well as impersonation attacks, and the resistance against man-in-the-middle and impersonation attacks provide sufficient grounds to maintain this fact that our scheme ensures session key security in the face of crafty and malicious adversaries.
6.2.6. Denial of Service Attack
In DoS attack, an adversary may exploit vulnerability in the scheme if it is designed in such a way that the adversary may fabricate multiple authentication requests, while the server gets engaged to entertain each fake request and maintains that session for an indefinite time. In this manner, the attacker could initiate multiple fake requests towards the server and affect its capacity to serve the legal authentication request, which undermines its serving capacity. In our scheme, aborts the session immediately if either the timestamp is not found to be fresh or the equality does not hold true. In both cases, drops the session immediately which helps the protocol to avoid denial of service attack on the ’s end.
6.2.7. Replay Attack
Our scheme employs the feature of timestamp that allows the protocol participants to verify the freshness of the received message. To construct the authentication request , the engenders the latest timestamp and embeds the same in the parameter for submission to . verifies the freshness of timestamp and verifies the calculated against . If true, it confirms the validity of . Likewise, after receiving the message confirms the freshness of timestamp and proceeds further to compute the response message. Upon successful verification, submits the parameter towards via . Then, verifies the authenticity of by computing and verifying the corresponding timestamp . In this manner, the replay attack can be successfully thwarted for the proposed scheme.
6.2.8. Physical Capturing Attack
In our scheme, the control center stores in the memory of a UAV (say ), where . In addition, the stores in its verifier memory. is a randomly selected identity of the , and it is different for each of the drone. Likewise, is also uniquely selected for each of the drones. Henceforth, are computed uniquely for each of the UAV. In case a UAV is physically captured by a malicious entity, it cannot extend any advantage to the malicious entity to successfully forge any response message on behalf of another UAV say , due to the uniqueness of parameters stored in each of and . In this manner, physical capturing is thwarted for the proposed scheme.
6.2.9. Perfect Forward Secrecy
In our scheme, the computation of the session key requires both session specific secrets as well as long-term secrets of both the entities , where and . It is clear that requires private key of the user and requires private key of the UAV. Every session consists of unique short-term and long-term parameters. The leakage of any of the session key or any of the long/short-term parameters may not extend any advantage to the attacker to gain abilities for compromising any future session.
7. Security and Performance Comparisons
In this section, the comparisons of the proposed scheme and related schemes presented in [19, 23, 24, 28–30] are conducted.
7.1. Security Features
Security feature comparisons of the proposed and relevant schemes proposed in [19, 23, 24, 28–30] are depicted in Table 3. All the relevant and compared schemes [19, 23, 24, 28–30] lack one or more security features: as proved in Section 4, the scheme of Nikooghadam et al. [28] has weaknesses against several attacks including stolen verifier, lack of untraceability, user and UAV impersonation based on secret parameter exposure of the UAVs, privileged insider and related attacks. The scheme of Wazid et al. has weaknesses against anonymity and untraceability, privileged insider and stolen verifier attacks, the scheme of Ever [30] lacks forward secrecy and prone to known session key and is privileged insider attacks. Similarly, the scheme of Zhang et al. [24] lacks mutual authentication, anonymity and untraceability, and forward secrecy, and the scheme of Turjman et al. [29] is also insecure against stolen verifier attack. The scheme of Srinivas et al. [19] lacks anonymity and untraceability and is also vulnerable to stolen verifier attacks. Only the proposed scheme provides all security features and resists known attacks.
7.2. Computation Cost
This subsection provides the comparison of the proposed scheme with related schemes [19, 23, 24, 28–30] as per real-time experiment conducted over MIRACL library, where we used Xiaomi-Redmi Note-8 smart phone with RAM size of 4 GB and 2.01 GHz Octa core Max processor. The underlying operating system in smart phone is Android V-9 and MIUI V-11.0.7. For replicating the trusted control center in the experiment, HP-EliteBook 8460-P with 2.7 GHz Intel® Core-TM and 4 GB RAM along with Ubuntu LTS 16 OS was used. Likewise, to replicate a UAV, Pi3B + Cortex (A53-ARMv8) with 64 bit-SoC, 1.4 GHz processor, and 1 GB LPDDR2-SDRAM was used in the experiment. The running times and notations used for each cryptographic operation for each of the user/mobile, , and UAV are depicted in Table 4.
The authentication process in the proposed scheme is furnished between , , and , where, executes operations. The executes operations; in addition, executes operations. Therefore, in proposed scheme, the running time to complete authentication among , , and is . The running time to complete authentication in Nikooghadam et al.’s scheme [28] is . As per the experimental results given in Table 5, the running times of the schemes presented in [19, 23, 24, 29, 30] are , , , , and , respectively.
7.3. Communication Cost
The communication cost comparison of the proposed and related schemes [19, 23, 24, 28–30] is shown in this subsection. In proposed scheme, we utilized with 160 bit size. We used AES as the symmetric cryptographic algorithm with 128 bit cipher size and 192 bit key. The output sizes of asymmetric cryptographic techniques are taken as per NIST recommended sizes which are 1024 bits and 320 bits each for RSA and ECC. The size of random numbers are also taken as 160 for simplicity and the size of a timestamp is fixed at 32 bits. Moreover, all the identities are taken as 128 bits long. sends to , where, , the size of each of the identity is 128 bits, and size of , so total size of , which requires three symmetric encryption operation each of size . Therefore, to encrypt , we require 48 bytes. The total size of is bits. Similarly, requires to send -bits from to . sends , which requires transmission of -bits. Finally, sends to , which also requires transmission of bits. Therefore, total communication cost of the proposed scheme is bits, which is 408 bytes. The communication cost of the Nikooghadam et al.’s scheme [28] is 356 bytes. The communication cost of the schemes presented in [19, 23, 24, 29, 30] is 212, 240, 184, 300, and 192 bytes, respectively.
8. Conclusion
We initiated this article by highlighting the importance of secure communication among entities of a UAV network for crowd management. We revisited and proved that Nikooghadam et al.’s scheme has vulnerabilities against stolen verifier attacks and lacks user anonymity and untraceability. It is also proved that Nikooghadam et al.’s scheme is not practical due to the exposure of secret parameters of the UAVs and the users, which can ultimately lead to the total failure scenario. Therefore, a security-enhanced authentication scheme is proposed in this article using only the lightweight symmetric key primitives and ECC. We implemented the proposed scheme in a real-time environment to extract the running time of each of the entities involved in the authentication process. The BAN logic-based security analysis and a discussion on the security features affirm that the proposed scheme can resist known attacks while raising some computation. In future, we intend to implement the proposed protocol in real world for securing crowdsourcing systems.
Data Availability
No data were used to support this study.
Conflicts of Interest
The authors declare that they have no conflicts of interest.
Acknowledgments
The authors express their appreciation to the Deputyship for Research and Innovation, Ministry of Education in Saudi Arabia for funding this research work through the project number (227).