Multiaccess Edge Computing Empowered Flying Ad Hoc Networks with Secure Deployment Using Identity-Based Generalized Signcryption

Hamdard Institute of Engineering & Technology, Islamabad 44000, Pakistan Department of Information Technology, Hazara University, Mansehra, Pakistan Department of Electrical Engineering, National University of Sciences and Technology (NUST), Islamabad 44000, Pakistan Department of Computer Science and Information Systems, Islamic University of Madinah, Madinah 400411, Saudi Arabia Department of Electrical Engineering, Air University, Islamabad 44000, Pakistan Descon Engineering Limited, Lahore, Pakistan Department of Electronic Engineering, ISRA University, Islamabad 44000, Pakistan


Introduction
Unmanned Aerial Vehicles (UAVs) have earned recognition in multiple domains owing to their versatile applications for surveillance, agriculture, health services, traffic monitoring, inspection, public safety, etc. [1]. Multiple small UAVs, as a flying ad hoc network (FANET), can combine and accomplish the assigned tasks efficiently in an autonomous manner [2,3]. In FANETs, small interconnected UAVs synergize and exchange data with one another and with the ground stations [4]. ey are characterized by high mobility, easy deployment, and self-organizing behavior [5]. However, such distinctive features, for efficient and effective deployment, demand the compliance of stringent guidelines [6]. For instance, it is mandatory to assure security and Quality of Service (QoS) when choosing a FANET system for on-time data communication services. Moreover, the networks must deploy an efficient networking architecture complemented by an efficient security scheme in order to allow a reliable exchange of information between UAVs and the ground stations.
FANETs can either be deployed independently or they can be integrated with the traditional networks via satellite or cellular communication links. e topic allures experts from the industry as well as academia. Most of the relevant research studies propose to integrate multiple-UAV systems with the traditional networks to assure Quality of Service (QoS), unhampered security, and sustained reliability. erefore, it is imperative to identify loopholes in existing solutions. is can pave the way for solutions that support high throughput and a secure data communication regime. e envisioned Fifth Generation (5G) of wireless cellular communication systems is expected to offer higher capacity, enhanced data rate, and lower latency [7]. Besides, 5G offers multiaccess edge computing (MEC) architecture, which is characterized by cloud computing functionalities. us, 5G, when integrated into a UAV environment, by leveraging MEC, can relieve the resource-constrained UAVs from processing the computational tasks. Instead, the computationally intensive tasks will be offloaded to the edge of the network.
Generally, the small UAVs are not designed with security considerations and are, therefore, prone to security and privacy pitfalls [8]. UAV's sensing portion is also worth consideration. For instance, in the worst case, a sensor might transmit wrong information and that can result in UAVs making erroneous decisions. Similarly, the case of the faulty sensor is far more sinister. A damaged sensor can severely hamper the UAV's attempt to obtain information and might result in an event of a crash. Furthermore, a strong communication link is essential to allow the exchange of information between a UAV and a Base Station. An insecure and vulnerable link, on the other hand, is susceptible to attacks [9]. e concerns of confidentiality and authentication can be addressed by employing encryption and digital signature, respectively. And, in case both the attributes are desired, a hybrid version, the sign-then-encrypt approach, is utilized mostly.
However, the stringent constraints associated with a flying ad hoc network (FANET), such as limited onboard energy and limited computing capability, do not permit complex cryptographic operations. Moreover, undertaking computationally intensive tasks may result in slow response time which can, in turn, deteriorate the performance of FANETs. Fortunately, such deficiencies can be resolved by employing an amalgamated scheme, named "signcryption" [10]. It is a public key cryptosystem that performs the function of encryption and digital signature simultaneously. It is far more efficient and cost-effective than each of the alternates, i.e., encryption and digital signature. To simplify the key management process and to allow flexibility, Han et al. [11] presented an extension of the signcryption scheme, i.e., generalized signcryption (GSC). Not only does GSC offer encryption and digital signature in one go, but it also has the option to offer them separately, if demanded. Such feature is helpful in case either of the two key attributes, confidentiality or authenticity, is required.
In the public key cryptosystems, two basic approaches, Public Key Infrastructure (PKI) and Identity-Based Cryptography (IBC), are used to authenticate public keys [12]. In the PKI environment, it is crucial to ensure a trustworthy unforgeable link between the identity of the participant and its public key. is further stipulates the need for a signature Certificate Authority (CA) that assigns the link a unique signature. In the certification stage, the CA bounds the public key as the identity of a participant with certificates.
e Public Key Infrastructure (PKI) approach encounters issues with certificate distribution and storage. On the other hand, an identity-based cryptosystem is used to reduce the cost of public key management [13]. In ID-based systems, a trusted third party named private key generator (PKG) computes private keys from a master secret and users' identity information. It then distributes these private keys to the users participating in the scheme. is eradicates the necessity for certificates as used in a conventional PKI. e security and efficiency of the aforementioned security schemes are based on computationally hard problems. e RSA cryptography [14,15] is based on a large factorization problem, which utilizes a large key, parameter certificate, and the identity stretches as much as 1024 bits [16].
is is not suitable for resource-constrained networks, or FANETs, because small UAVs lack onboard processing resources. Furthermore, bilinear pairing is 14.31 times worse than RSA [17], due to huge pairing and map-to-point function computation. In order to eliminate the discrepancies accompanying RSA and bilinear pairing, a new type of cryptography called the elliptic curve was introduced [18]. e elliptic curve cryptography is characterized by smaller parameter size, smaller public/private key size, smaller identity, and smaller certificate size. Moreover, unlike bilinear pairing and RSA, the security hardiness and efficiency of the elliptic curve cryptography scheme are based on 160bit small keys [19]. e 160-bit key is, still, not suitable for and affordable by resource-hungry devices such as small UAVs. us, the hyperelliptic curve, a more modern version of the elliptic curve cryptography, was proposed [20]. e hyperelliptic curve uses an 80-bit key, identity, and certificate size and, at the same time, promises the security features assured by the elliptic curve, bilinear pairing, and RSA [21,22]. erefore, the hyperelliptic curve is a cogent choice for energy-constrained devices.

Authors' Motivation and Contributions.
To reap the extensive benefits of multi-UAV systems, the underlying technical challenges need to be addressed. For instance, the small UAVs have limited onboard energy, which restricts the flying time to a specified period and the UAV's limited computational capability does not permit complex cryptographic operations. erefore, there is a need to harness a state-of-the-art communication architecture with a lightweight security mechanism, which can, significantly, stabilize the battery lifetime, offer limited computation cost, and provide better connectivity.
Motivated by such objectives, for FANETs, the authors, here, suggest an identity-based generalized signcryption scheme. e very scheme makes use of multiaccess edge computing (MEC) and is based on a much advanced version of the elliptic curve, i.e., the hyperelliptic curve (HEC). HEC is characterized by a smaller key size and, at the same time, promises security comparable to that of the counterparts, i.e., elliptic curve, bilinear pairing, and modular exponentiation. Incorporation of HEC reduces power consumption and improves the device's performance, thereby making it suitable for a wide range of devices, ranging from sensors to UAVs.
Some of the salient features signifying the contribution of our research work, in this paper, are as follows: (i) We introduce a new architecture for flying ad hoc networks (FANETs) leveraging multiaccess edge computing (MEC) facility, where the primary UAV acts as a MEC node in order to provide computational offloading services for the member UAVs having limited local computing capabilities (ii) We propose an efficient and provably secure identity-based generalized signcryption scheme for the architecture using the concept of a hyperelliptic curve (iii) e proposed scheme is potent enough to thwart attacks, both known and unknown, and the validation results using the Automated Validation for Internet Security Validation and Application (AVISPA) tool second such notion (iv) Moreover, upon doing a comparative analysis with the extant schemes, it is revealed that our proposed scheme is superior, particularly, in terms of computational and communication costs 1.2. Structure of the Paper. e rest of the paper is organized as follows. In Section 2, we provide a brief about the related work. Foundational concepts of the research work are presented in Section 3. Section 4 is dedicated to present the two system models, i.e., network model and threat model. In Section 5, we explain the salient features of the proposed scheme. Informal security analysis is provided in Section 6. Section 7 presents the practical deployment of the proposed scheme. For performance evaluation, the proposed scheme is compared with the existing schemes in Section 8. Section 9 contains a brief about a case study in which the scheme is applied for precision agriculture. Finally, Section 10 concludes the work.

UAV-Enabled Multiaccess Edge
Computing. Owing to the promising features of on-demand communication services and flexible deployment, UAV-enabled multiaccess edge computing capabilities have received much attention in recent years. So far, various studies have been conducted to examine the usability of edge computing for UAVs [23,24]. However, the studies do not address the topic of security. Garg et al. [25] aimed to answer the surveillance-related concerns by proposing a framework based on probabilistic data structures. e framework treats UAVs as intermediate aerial nodes that offer a cyberthreat detection mechanism complemented with a real-time analysis. Four major elements of the framework are as follows: UAV, dispatcher, aggregator, and edge devices. e UAV is responsible for capturing and validating the data. e processing tasks in the edge computing devices are scheduled by the dispatcher. e aggregator assures the secure transmission of data. And, the edge devices analyze the data.
In [26], the authors extend the concept of network slicing to the case of UAV-based 5G network deployment and investigate the feasibility of a backhaul of an aerial node utilizing a UAV. e LTE signals are monitored to evaluate the suitability of UAVs in two scenarios: network capacity enhancement and increasing network coverage. e methodology proposed by Christian et al. [27] increases the system reliability and reduces the end-to-end source-actuator latency. eir work intends to broaden the 5G network edge by making the FANET UAVs fly close to the monitoring layer. For enhanced operations, the UAVs follow a policy of mutual help and are accoutered with MEC facilities. However, the work fails to address the issue of the limited battery duration of the MEC-UAVs. In [28], the authors proposed a UAV edge-cloud computing model that utilizes a UAV swarm to provide the users real-time support.
e end data are stored in the cloud server. In [29], the authors presented an architectural design of a slice orchestrator that enables new application models where the Internet of ings related functions can be applied on small Unmanned Aerial Vehicles, thus paving the way for implementing these functions on the edge network.

Security Mechanisms in Flying Ad Hoc Networks.
e primary security mechanisms for FANETs emphasize authenticity, confidentiality, and integrity of data via cryptography. A well-designed data protection mechanism can significantly reduce the probability of the data get compromised, irrespective of the devilish technique involved.
ere are a few studies dedicated to investigating the data protection issues for UAV Networks. In a secure communication scheme proposed by He et al. [30], the requirement of an online centralized authority is waived off. e UAVs manage the area themselves and the authorized devices can obtain a broadcast key. e scheme is characterized by employing hierarchical identity-based broadcast encryption and a pseudonym mechanism, whereby the devices can, anonymously, broadcast the encrypted messages and decrypt the legal ciphertext. e work done seconds the notion that the very scheme, satisfactorily, addresses the four important security concerns: confidentiality, authentication, partial privacy preservation, and resistance to Denial of Service (DoS) attacks. However, it inherits a restriction in the registration phase, i.e., the concern of finding a hash value's preimage persists.
ree communication scenarios have been described by Won et al. [31,32] to propose cryptographic protocols for drones and smart objects. e first scenario, i.e., one-to-one, implies a certificateless signcryption tag key for facilitating an authenticated key agreement and for providing nonrepudiation and user revocation. One-to-many, or the second scenario, enables a UAV to broadcast privacy-sensitive data to multiple smart objects using a certificateless multirecipient encryption scheme. e third scenario is termed "many-to-one" and is characterized by UAVs capable of collecting data from multiple smart objects. However, for such protocols [31,32], transmitting encrypted messages and assuring privacy simultaneously are too difficult to undertake. Such novel cryptographic mechanisms are efficient and secure. However, they are supposed to be used in group communication where nodes are of equal computational capability. In 2019, Asghar et al. [33] proposed a blind signature scheme for flying ad hoc networks in a certificateless setting. e scheme is suitable for authentication; however, it does not offer confidentiality and authentication simultaneously.

Identity-Based Generalized Signcryption Schemes.
Lal et al. [34], in 2008, introduced the first identity-based generalized signcryption scheme and proposed a security model for it. However, Yu et al. [13] pointed out that the security model presented by Lal et al. [34] scheme is incomplete and proposed a new scheme, which is efficient in terms of computation and is secure. Later, in 2011, Kushwah et al. [35] simplified the security model introduced by Yu et al. [13] and proposed a more efficient identity-based generalized signcryption scheme. Wei et al. [36], in 2015, presented an identity-based generalized signcryption scheme, which demonstrated to be secure enough in the random oracle model. Shen et al. [37], in 2017, proposed an identity-based generalized signcryption scheme in the standard model. Nevertheless, the proposed scheme is based on bilinear pairing that is computationally expensive. In 2019, Waheed et al. [38] analyzed the work done by Wei et al. [36] and suggested an improved scheme that is far more secure and cost-effective. Lastly, in 2019, Zhou et al. [39] proposed an identity-based combined public key scheme for signature, encryption, and signature (IBCSESC). Under the premise of ensuring the confidentiality, integrity, authentication, and nonrepudiation of data, the combined cryptosystem reduces the key management work, saves storage space, and offers decreased computational consumption.

Hyperelliptic Curve Cryptography (HECC)
. HECC is the advanced form of elliptic curve cryptography (ECC), and it is used to exchange keys and facilitate secure communications between two parties with very small size keys and incur lower computational and communication costs. For instance, an encryption activity done using RSA with a 1024bit key and ECC with a 160-bit key is equivalent in performance to HECC encryption with an 80-bit key [40].
Suppose that Iq is a predetermined set and presume z as the genus of hεc having order as z ≥ 2.
us, hεc of genus z ≥ 2 over Iq is set of points (v,) Iq * Iq as shown in It forms the divisors which are the formal sum of finite integers like d � x i z i where x i ∈ Iq and z i ∈ hεc. Further, it forms a Jacobian group I hεc (Iq) having the following order:

Hyperelliptic Curve Discrete Logarithm Problem
(hεc − dlΡ). Assume that d is the divisor that is publicly available in the network and L is a randomly picked private number from I t . Upon recovering L from d 1 � d, L is said to be (hεc − dlΡ).

System Models
To elaborate on the operation and applicability of the proposed scheme, two models are used.

reat Model.
e proposed scheme employs the Dolev-Yao (DY) threat model [42]. e model indicates that an untrustworthy nature prevails between the end-point entities and that there is an insecure open channel between the parties. us, for an attacker, it eases the task to eavesdrop and delete/modify the exchanged messages. Far worse is the scenario when a drone, while hovering over a hostile area, is physically captured and the data is compromised. Recently, the widely accepted "Canetti and Krawczyk's adversary model (CK-adversary model)" [43] becomes the "current de facto standard model in modeling authenticated key exchange protocols." According to the CK-adversary model, "the adversary can not only deliver the messages (as in the DY model), but can compromise the secret credentials, secret keys and session states a well, particularly, when stored in the insecure memory." erefore, it becomes an essential requirement that "the leakage of some forms of secret credentials, such as session ephemeral secrets or secret key, should minimally effect the secrecy of the communicating participants" [33].

Proposed Identity-Based Generalized
Signcryption Scheme

Syntax of Identity-Based Generalized Signcryption
Scheme. A formal model of identity-based generalized signcryption scheme consists of the following four algorithms [13,37]: setup, key extraction, generalized signcryption, and generalized unsigncryption. e notations used in the proposed scheme are illustrated in Table 1.
(i) Setup. In the setup phase, the private key generation (PKG) generates the public parameters, randomly selects their master private key, and computes the master public key with the input of security parameter. (ii) Key Extraction. When each of the participated contestants transmits their respective identities (ID ps ) to the PKG, PKG generates the private (A pc ) and public (B pc ) keys for each of them and delivers them using the private network. (iii) Generalized Signcryption. e sender performs this process for producing generalized signcryption of a message (m). It initially takes the input parameter such as the identity of the sender and receiver (ID cs , ID cr ), message (m), the private key of the sender (A cs ), the public key of the receiver (B cr ), and a fresh nonce (n cs ). (iv) Generalized Unsigncryption. e receiver performs this process for recovering a message (m) and verifying generalized signcryption text ψ. It takes the input parameter like generalized signcryption text ψ, the identity of the sender and receiver(ID cs , ID cr ), the private key of the receiver (A cr ), the public key of the receiver(B cr ), and the public key of the sender (B cs ).

Construction of the Proposed Identity-Based Generalized
Signcryption Scheme. It includes the following four subphases [13,37]: Setup: in this phase, the private key generation (PKG) center performs essential steps. It

Mobile Information Systems
Generalized signcryption: given a message (m), the private key of the sender (A cs ), the public key of the receiver (B cr ), the identity of the sender and receiver (ID cs , ID cr ), and a fresh nonce (n cs ), the sender performs this process for producing generalized signcryption by undertaking the following steps (a) It selects a number in an irregular manner as φ ∈ [1, 2, . . . , (q − 1)] and calculates Δ � φ · D (b) It calculates β � φ · B cr · ID cr (c) It computes η � e β (m//ID cs //ID cr //n cs ) (d) It calculates σ � h b (m//ID cs //ID cr //n cs ) (e) It computes z � (ID cr · φ − σ · Δ · A cs · ID cs ) mod q (f ) It produces the final generalized signcryption text for the receiver as ψ � (z, σ, η, Δ) Generalized unsigncryption: given a generalized signcryption text ψ � (z, σ, η, Δ), the private key of the receiver (A cr ), the public key of sender and receiver (B cs , B cr ), and the identity of the receiver (ID cr ), the sender performs this process for verifying the signature, and recovering a plain text (m) by undertaking the following steps: (a) It computes β � z · B cr + ID cs · Δ · σ · B cs · A cr (b) It decrypts (m//ID cs //ID cr //n cs ) � d β (η) (c) It computes σ ∧ � h b (m//ID cs //ID cr //n cs ) (d) It compares σ ∧ � σ, if holds, then accept ψ otherwise generate the error symbol ╨ Note that, in the above algorithm, if ID cs � null and ID cr ≠ null, then generalized signcryption proceeds in an encryption process. If ID cr � null and ID cs ≠ null, then generalized signcryption will run in the signature mode. And, if ID cs ≠ null andID cr ≠ null, then generalized signcryption will run in signcryption mode.

Informal Security Analysis
is section is dedicated to spotlight the proposed scheme's contribution in upholding basic security including resistance to replay attack, confidentiality, integrity, and unforgeability. Each of the characteristics is briefly analyzed in the following sections.

Confidentiality.
e proposed scheme ensures confidentiality. In case an intruder wants to steal the original contents of a message or the secret key, he/she must have beforehand information about the key as β � φ · B cr · B ID cr . In order to determine β, it is required to compute φ from Δ � φ.D, which is the discrete log problem in the hyperelliptic curve. e scheme offers replay attack resistance. Each session implies a fresh key (β) and a nonce (n cs ) i.e., η � e β (m//ID cs //ID cr //n cs ). erefore, it is, literally, not possible for an intruder of a session to penetrate another session with the same session key. Besides, the receiver is required to run a check for ascertaining the freshness of a message at every instance of reception. An obsoleteness, if spotted, renders the message useless.

Integrity.
e sender takes the "hash value" of the message before sending the message, i.e.,: σ � h b (m//ID cs //ID cr //n cs ). e "hash" exhibits a property of being an irreversible function. For the confirmation if either of the ciphertexts is altered or not, the receiver performs the following steps: it first decrypts (m//ID cs //ID cr //n cs ) � d β (η) and computes σ ∧ � h b (m//ID cs //ID cr //n cs ). After it compares σ ∧ � σ, if it holds, then it accepts ψ; otherwise, it generates the error symbol ╨.

Unforgeability.
In our proposed scheme, if the intruder tries to generate a valid signature, then he/she is, first of all, required to compute z � (ID cr · φ − σ · Δ · A cs · ID cs ), and to do so, the intruder needs to find φ from Δ � φ · D and A cs from B cs � A cs · D. is equates to solving two hard problems with commensurate efforts. us, it is ensured that our designed approach offers resistance against the signature forging attack.

Deployment of the Proposed Scheme
In this phase, we provide the practical deployment of our proposed technique in the UAVs network for precision agriculture that involves monitoring of crop health in a cultivated field. e proposed scheme includes three subphases that are initializations, registration, and data transmission and verification, respectively. Figure 2 illustrates the initialization process, in which the PKG first calls the setup algorithm; i.e., it first selects a security parameter κ, picks a hyperelliptic curve (HEC) of the genus, chooses a parameter q where the length is equivalent to 80 bits, selects a finite field f q , where its order is q, picks a divisor D of order q, select two one-way hash functions, i.e., h a and h b , chooses a number uniformly for its private key as δ ∈ [1, 2, . . . , (q − 1)], computes its public as Λ � δ · D, produces all the public parameter E � [q, h a , h b, f q , κ, Λ, HEC, D], and published it to the network. Note that, in this subphase, we used ID mec , ID mbs , and ID m−uav for the identity of MEC-UAV, MBS/SBS, and M-UAV. Figure 3 illustrates the registration process in which the PKG first calls the key extraction algorithm; i.e., when each of the participated contestants transmits its identity (ID pc ) to the PKG, then PKG generates the private and public keys as follows: it computes the private key for identity (ID pc ) as A pc � δ · h a (ID pc )mod q, and then it computes public key for identity (ID pc ) as B pc � A pc · D Finally, PKG delivers the pair of public and private keys (B pc , A pc ) to the participated contestants with its identity (ID pc ) by using the private network.note; in this subphase, we used (A mec , B mec ), (A mbs , B mbs ), and (A m−uav , B m−uav ) for the private and public keys of MEC-UAV, MBS/SBS, and M-UAV. Figure 4 illustrates the data transmission and verification of the proposed scheme. In this phase, MEC-UAV performs the following process for generating a signcrypted ciphertext: it first selects a number in an irregular manner as φ ∈ [1, 2, . . . , (q − 1)] and calculates Δ � φ · D. It also calculates β � φ · B mbs · D mbs and computes η � e β (m//ID mec //ID mbs //n mec ).
In the above process, if ID mec � null and ID mbs ≠ null, then MEC-UAV performs the encryption process. If ID mbs � null and ID mec ≠ null, then MEC-UAV performs the signature method. If ID mbs ≠ null and ID mec ≠ null, then MEC-UAV performs the signcryption mode.

Computational Cost.
For evaluating the effectiveness, the proposed scheme is compared with five existing schemes proposed by Yu et al. [13], Kushwah et al. [35], Wei et al. [36], Shen et al. [37], and Zhou et al. [39]. e major findings obtained from the comparison are depicted in Table 2. e five existing schemes utilize elliptic curve scalar multiplication and bilinear pairings, both of which are costlier options. erefore, we apply the hyperelliptic divisor multiplication. From the observations, it has been revealed that the time taken for processing a single scalar multiplication varies considerably: Elliptic Curve Point Multiplication (ECPM), 0.97 ms; bilinear pairing, 14.90 ms; pairing-based point multiplications, 4.31 ms; and modular exponentiation, 1.25 ms [44]. In order to measure the performance of the proposed scheme, the Multiprecision Integer and Rational Arithmetic C Library (MIRACL) [12] is used. It tests the runtime of the basic cryptographic operations for about 1000 times. For testing the simulation results, a workstation Mobile Information Systems 7 having the following specifications is used: Intel Core i7-4510U CPU @ 2.0 GHz, 8 GB RAM, and Windows 7 Home Basic 64-bit Operating System [42]. Owing to a smaller key size of 80 bits, the Hyperelliptic Curve Divisor Multiplication (HCDM) is assumed to be of 0.48-millisecond duration [45,46]. From the findings in Tables 2-4 and Figure 5, it is evident that our approach is far more efficient in terms of computational costs.

Communication Cost.
is section is dedicated to discuss the comparison results in the perspective of communication costs. e proposed approach is compared with the existing five schemes presented by Yu et al. [13], Kushwah et al. [35], Wei et al. [36], Shen et al. [37], and Zhou et al. [39]. In the comparative analysis, the variables used along with the respective values are shown in Table 5 [40].
It is assumed that each of the schemes has associated communication costs as shown in Table 6.
From Figure 6, it is evident that a decision to opt for our proposed scheme results in a significant reduction in the associated communication costs. Table 7 depicts the percentage reduction in communication costs.

Flying Ad Hoc Network-Based Precision Agriculture: A Case Study
To further assess the practicability, the proposed scheme is applied to a precision agriculture case that involves FANETs for monitoring the health of the crops. Small UAVs are used to capture the images, which are, in the next step, processed to extract useful information. Values from the Normalized Difference Vegetation Index (NDVI) are computed to differentiate healthy plants from the nonhealthy ones. is is done by measuring the chlorophyll content. It further helps in the localization of the area under stress. e images captured by the M-UAVs are transferred to the MEC-UAV, which, utilizing the onboard microcontroller, generates the respective tasks to be carried on by the Decision Support Engine (DSE). For value addition and versatility, the M-UAVs can have additional gadgets, such as cameras, IMU, sensors, and GPS units. e web portal contains a variety of services such as visualization of historical/ real data, NDVI mapping, and the correlation functionality.
For our future work, we aim to complement the research work by including other aspects of formal analysis, such as the Real-Or-Random (ROR) model and Random Oracle Model (ROM). Moreover, we also intend to incorporate a computational offloading and scheduling mechanism, in which the M-UAVs will be able to offload and schedule the computing tasks to the MEC-UAV for improved processing power and faster execution.