IEEE Copyright Notice

© 2020 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained
for all other uses, in any current or future media, including reprinting/republishing this material for
advertising or promotional purposes, creating new collective works, for resale or redistribution to servers
or lists, or reuse of any copyrighted component of this work in other works.

Accepted to be published in:

• Proceedings of the 2020 - IEEE International Conference on Advanced Networks and Telecommu-
nications Systems - Workshop on New Advances on Vehicle-to-Everything (V2X) Communications
and Networking), 14-17 December 2020.
 A Survey on Blockchain and Edge Computing
applied to the Internet of Vehicles
Anderson Queiroz, Eduardo Oliveira, Maria Barbosa and Kelvin Dias
Centro de informática
Universidade Federal de Pernambuco
Recife, Brazil
aalq@cin.ufpe.br, ehammo@cin.ufpe.br, mksb@cin.ufpe.br, kld@cin.ufpe.br

Abstract—With the advent of Intelligent Transportation Sys- tography, shared registration, distributed processing, and joint
tems (ITS), data from various sensors embedded into vehicles validation of data and transactions) offered by the blockchain.
or smart cities infrastructure are of utmost importance. This
ecosystem will require processing power and efficient trust Traditional access to the remote cloud may degrade
mechanisms for information exchange in vehicle-to-everything VANETs services due to incurred latency. Nevertheless, with
(V2X) communications. To accomplish these requirements, both the emergence of Vehicular Edge Computing (VEC), process-
edge computing and blockchain have been recently adopted
towards a secure, distributed, and computation empowered
ing and storage are now close to vehicles, that is, at the edge
Internet of Vehicles (IoV). This paper surveys prominent solutions of the network. Recently, researchers proposed solutions for
for blockchain-based vehicular edge computing (VEC), provides joint blockchain and edge computing applied to the Internet
a taxonomy, highlights their main features, advantages, and of Vehicles (IoV). Besides cloud-like services closer to the
limitations to provide subsidies for further proposals. end-user (Edge, Fog 1 , and Cloudlet), IoV also comprises soft-
Index Terms—VANET, Internet of Vehicles (IoV), Fog, Edge warization technologies, such as Software-Defined Networking
and Cloud Computing, Blockchain
(SDN) and Network Functions Virtualization (NFV), as well,
artificial intelligence support for decision making from both
I. I NTRODUCTION autonomous cars and automating service orchestration and
management of vehicular infrastructure.
Vehicular Ad-hoc Networks (VANETs) have emerged as
a key technology to support road safety, comfort, and in- The proposed solutions have points in common and pe-
fotainment applications for passengers and citizens of smart culiarities in terms of architectural design and also distinct
cities. VANETs rely on the following communication modes: characteristics that must be clarified to understand their advan-
Vehicle-to-Infrastructure (V2I), Vehicle-to-Vehicle (V2V), and tages and disadvantages in the scenario of future developments
Vehicle-to-Everything (V2X). These networks aim to pro- of possible solutions in the area. Various approaches involv-
vide communication with any component of the Intelligent ing different edge computing techniques and applications,
Transportation System (ITS) in smart cities, such as other blockchain, and other security mechanisms for exchanging
vehicles, stations, pedestrians, traffic lights, bicycles, and any information and data, have been introduced to the academic
other member participating in this ecosystem. VANETs are community and companies in recent years. This study aims
composed of entities, such as sensors, cameras, and On-Board to analyze and compare these different researches by corre-
Computerized Units (OBU) located in vehicles, Road Side lating the proposals, intending to understand each technology,
Units (RSU), and Base Stations (BS) with various communica- and how they work and deal with the environment in each
tion network standards, such as: 802.11p, 802.15, 4G/LTE, and application [3]–[8].
5G/NR, which have more computational power than vehicle This article examines crucial solutions for edge computing
OBUs [1]. and blockchain applied to IoV, provides a taxonomy, and
The data can be processed and delivered by the base station highlights, through a comparative summary, its main propos-
to the cloud or by vehicles collaboratively. Figure 1 presents als, technologies, and architectures as a way of providing
the various architectural, technological, and communication subsidies to other proposals. Section II shows the background
aspects that embrace the IoV environment in the context of this on Blockchain, Section III presents the taxonomy of the com-
article. However, to be widely implemented, VANETs still lack parative elements of the study, the summary of the solutions,
effective security and privacy mechanisms to minimize false and the discussion of the advantages and limitations of each
and malicious exchanges of information between vehicles. proposal. Section IV presents the conclusion of the article.
Such incidents can cause different types of accidents that
threaten the lives of drivers, passengers, and pedestrians [2].
Considering that even with the growth in the number of
devices carrying out transactions on VANETs, it is possible to 1 This article considers the term Fog to refer to the end nodes, that is, the
reach security levels through the features (immutability, cryp- vehicles
 Fig. 1. A General Framework for Blockchain-based Vehicular Edge Computing

II. F OUNDATION acterized by the centralized control of the nodes and rules
by a single organization. Finally, the consortium blockchain,
A. Blockchain
where a certain group of companies, usually with the same
As Jianbin Gao said in [1] ”Blockchain is a decentralized objective, perform the control and management of the nodes
infrastructure and a distributed computing paradigm that uses and their rules. This last type of implementation, consortium
encrypted chained block structures to validate and store data, or federated, has been widely used in the IoV environment
consensus algorithms to generate and update data and smart since its characteristics are similar to the current model of
contracts to program and manipulate data”. These blocks are vehicular networks as demonstrated in table 1. 3 .
made up of transactions that are verified and validated in a We also need to keep in mind two blockchain concepts, The
distributed manner, that is, without the need for a central trust consensus algorithm and the smart contracts. The consensus
entity. This solution was first proposed by Satoshi Nakamoto algorithm aims to ensure higher trust and security on the net-
2
to define a way to validate transactions carried out on work. Therefore, it verifies the authenticity of the transactions
the internet with the bitcoin cryptocurrency. By its nature, and validates the blocks. However, for this to happen, there
Blockchain management is achieved by several participants, must be a consensus on the network. In other words, the
unlike traditional databases, which are generally managed consensus algorithm is the means to do agreements among
by a single organization. Another important feature of the the network nodes, ensuring that the terms designated by the
blockchain is its ability to easily track and verify data [1]. protocol will be met. The smart contract is a script stored in
Currently, there are three ways to implement blockchain a blockchain network, so it cannot be adulterated and each
on a network: public, private, or consortium. The most com- one has a unique address. This script automatically executes
mon implementation used by cryptocurrencies is the public the rules predefined in a contract or consensus, based on these
blockchain, where any node can participate in some activity rules it performs transactions among vehicles without the need
of the management process. Private implementation is char- for an intermediary.
2 Bitcoin: a point-to-point ATM system, Satoshi Nakamoto - 3 What is Blockchain, SAP - https://www.sap.com/insights/what-is-
https://bitcoin.org/bitcoin.pdf blockchain.html
 Fig. 2. Taxonomy of Blockchain and Edge computing for IoV

III. TAXONOMY OF BLOCKCHAIN APPLIED IN I OV authority.

The study proposal [7] created a scheme for certificateless
The proposed taxonomy is depicted in Figure 2. The taxon-
authentication, applying the session key independently to each
omy contains the building blocks of surveyed architectures
vehicle. They use the blockchain for establishing the V2V
that integrate Blockchain and Vehicular Edge Computing
groups, sharing, and managing all the existing vehicles of the
for IoV environment. It was structured into five categories:
same group.
Access network; Layers of the IoV architecture; Networking
technologies; Execution layer of blockchain tasks, such as Some of the papers utilize software-defined network (SDN)
mining, portfolio, consensus, and finally the Algorithms and or software-defined vehicular network (SDVN). [1] uses the
machine learning techniques applied. combination of blockchain and SDN technologies as a way
A common goal to add blockchain in an IoV environment is of improving the operation and management of the vehicular
to improve the security of the network. [2] presents a solution networks that employ 5G networks and edge computing.
for the adequate use of blockchain in vehicular networks with Being the management responsibilities shared between the
the division of activities and services in a balanced way in blockchain and the SDN. To create a virtualized, collabo-
the three layers of IoV. The blockchain, in the Perceptual rative, and configurable network, [8] proposes a structure
layer, is ensuring network security, the blockchain-related using SDVN along with a paradigm for the blockchain that
tasks are handled by the Edge-layer, and the Cloud-layer has will allow the certification of transactions to guarantee the
a backup of the blockchain storage. On the other hand, some anonymity of the data of the nodes in a totally distributed
articles bring unconventional approaches, like [3]. Aiming to and secure way. In addition, it is proposed a new consensus
solve the problem of data correctness and the information algorithm called distributed miners connected dominating set
integrity attacks in a VANET, the study proposes a non-linear (DM-CDS), which dynamically selects miners nodes.
blockchain. The blocks are driven by an acyclic graph, where Authors in [9] proposes an optimization-based solution for
each block has a single transaction. scaling resources in a virtualized edge computing environment.
Considering the paradigm of Electric vehicles cloud and They use Q-learning technique to determine which server
edge (EVCE), [4] uses blockchain in the distributed consen- would be the best for a given requested activity. To reach
sus activities, where they are based on the frequency and consensus in this network, it was utilized private blockchain.
contribution of data and energy in the system, which are In the paradigm of connected and autonomous vehicles
applied to obtain the score related to a work test (PoW). In (CAVs), [10] is seeking to decentralize the ML learning, cre-
[5] consortium blockchain was used, the authors elaborated ating the possibility for vehicles to share the models, learning
a scheme utilizing a three-weight subjective logic (TWSL) with each other, and at the same time ensuring privacy and
model to secure storage and sharing of data between vehicle security in the network. To this end, they utilize a blockchain-
networks. based collective learning (BCL) technique. The ML algorithms
The paper [6] proposes a many-to-many caching scheme and are used to improve the decision-making process of executions
a trusted access authentication scheme using Blockchain. The activities in the various layers of the IoV network, and the
authentication scheme is like a mechanism used to achieve Blockchain is applied to protect the users from security and
real-time monitoring and promote collaborative sharing for privacy threats.
vehicles. The blockchain is distributed, being the certifying The research work [11] is seeking dynamism and security
in the exchange of information. So, they present Autonomous should only communicate and not do any computation.
Vehicular Social Networks (AVSNs) in the blockchain context. By moving the computation to the Edge, there is less
Besides, they feature a consensus algorithm based on reputa- delay between messages exchanges while maintaining a
tion, Proof of Reputation (PoR). better processing power than the vehicle’s OBU. Both
In order to alleviate the traffic and reduce latency, [12] Single layer (Fog-layer) and this scenario have a big
proposes to use permissioned blockchain to create a content issue on resource allocation since all the processing
caching scheme. Due to the high mobility of vehicles, it is a occurs in a single layer.
challenge to decide when and where to cache. Therefore, the 3) Static multi-layer: In this case, the blockchain-related
study seeks to explore a deep reinforcement learning approach tasks are also in the Edge/MEC-layer [2], [4]–[7], [11].
to design this scheme. In addition, a new metric for the Nevertheless, in this scenario, the Fog-layer and the
selection of the block checker, the Utility Proof (PoU), was Cloud-layer may handle other tasks, some of those tasks
also proposed. could even be blockchain-related. Each of the layers has
The research [13] tried to solve the issue of limited compu- its tasks statically assigned, in some way, this may solve
tational resources in IoV, proposing a brokerage mechanism to the issue of overtaxing the MEC layer, but by failing
assist the validation decision process whether on-site or in an to analyze the context of the user, sending some of
edge or cloud infrastructure. They used Satisfiability Modulo the responsibilities to the cloud may create another set
Theories (SMT) method, besides, the mining of the blockchain of issues, such as the task never completing due to a
can happen anywhere in the network and storage is in every congested network. And by never sending some tasks to
participant node. the cloud, the MEC may be overtaxed in the same way
The surveyed papers mostly adopted machine learning it was in the previous scenario.
algorithms as a strategy for decision making, and heuristic 4) Dynamic multi-layer: This scenario deals with load-
for consensus algorithms. It is worth to mention that most balancing in a hybrid intelligent way [13]. As in the
papers focused on reinforcement learning and deep learning. previous scenario, the blockchain-related tasks may be
Although these techniques were used for different reasons, handled in any layer. Nevertheless, in this case, com-
most researchers have chosen the same techniques when putation resources of each layer and the user’s context
working in an IoV context. It would be interesting to see are taken into consideration to dynamically decide, in
a comparative study of several decision-making algorithms, which layer each task should be performed. However, if
to analyze which one is more appropriate to use in an IoV the decision mechanism chooses an inappropriate layer
context. It’s also worth mentioning that each article used ML to perform one task, it may cause some of the issues
for a different purpose. like offloading [13], resource allocation that were mentioned in previous scenarios
[9], caching [12] or for its application [10].
B. Discussion of Blockchain Solutions in IoV
A. Scenarios definitions On Table I we classified:
We have defined four scenarios to better classify the dif- • The blockchain Type (Public, Private, or Consortium) -
ferent architectures of the surveyed papers. Three facets were It is interesting to highlight that most researches pre-
taken into account to devise these scenarios: ferred consortium blockchains over other types. Public
• Which network layer is responsible for the blockchain- blockchains usually have low scalability, since they are
related tasks; totally decentralized, every node must process every
• What is the role of the Edge in this scenario; transaction. There are strategies to overcome this issue, as
• What is the resource allocation method in this scenario. the architecture proposed on [3]. On Public blockchains
1) Single layer (Fog-layer): In this scenario, the there’s also the issue with storage since every node must
blockchain-related tasks are only in one of the store every block of the immutable blockchain. Private
layers, the Fog-layer (In the vehicle’s OBU) [3], blockchains have their appeal since it does not have
[8]. The Edge/MEC layer exclusively transmits data most of the issues a public blockchain has. A central
between the VANETs and data from the vehicle to the authority decides which nodes can enter the blockchain,
Cloud-layer and vice versa. The Cloud-layer may offer which nodes can be miners, which nodes must have all
some services to the vehicles, but those would not be transactions, etc. But it’s more centralized since a single
blockchain-related. By distributing the blockchain on company manages the whole blockchain. A consortium
the final layer, there is less delay between messages blockchain, is only partially private since several compa-
exchange, since blockchain-related communications are nies share the ownership of the blockchain
carried out in V2V mode. • The blockchain Layer, i.e. where the mining and the
2) Single layer (MEC-layer): In this context, all the consensus are mostly happening. This column is relevant
blockchain-related tasks are in the Edge/MEC-layer [1], for the reader to know which layers were used in a static
[9], [10], [12]. This layer may offer services that are or dynamic multi-layer architecture.
necessary for the user. The Cloud-layer does not have • Architectural Approach, using the previously defined
an important role in this context. As for the Fog-layer, it architecture models. Bringing the blockchain closer to the
 TABLE I
S UMMARY OF SOLUTIONS : BLOCKCHAIN FEATURES .

Paper BlockchainType BlockchainLayer ConsensusAlgorithm Architectural Approach

Jianbin Gao 2019 [1] Consortium Edge PBFT Edge-layer
Shaoyong Guo 2019 [6] - Edge&Cloud - Static multi-layer
Jiawen Kang 2019 [5] Consortium Fog&Edge PoW Static multi-layer
Yueyue Dai 2019 [12] private/permissioned Edge PoU Edge-layer
Heena Rathore 2019 [3] - Fog TangleCV Policy Fog-layer
Youcef Yahiatene 2018 [8] Public&Private Fog Authors created DM-CDS Fog-layer
XiaoDong Zhang 2018 [2] - Fog&Edge&Cloud - Static multi-layer
Chau Qio 2018 [9] private/permissioned Edge PBFT Edge-layer
Haowen Tan 2019 [7] Consortium Fog&Edge&Cloud - Static multi-layer
Vincenzo 2019 [13] - Fog&Edge&Cloud PoS+PoW Dynamic multi-layer
Hong Liu 2018 [4] Consortium Fog&Edge PoEC/PoW + PoF/PoS Static multi-layer
Yuchuan Fu 2019 [10] Consortium Edge BFT-DPoS Edge-layer
Y.Wang 2020 [11] Consortium Fog&Edge PoR Static multi-layer

users decreased latency and increased the overall quality randomization. Usually, rich nodes would always
of service since the information got to the user faster win the miner election, and that may be a secu-
[3], [8]. But not every vehicle is so well equipped that rity issue. This algorithm was only present in the
can do all storage and computation. Some researchers surveyed papers if it was combined with another, as
decided to delegate these blockchain tasks to the Edge it was the case for [5] and [13].
or cloud layer. In fact, most papers have chosen to leave 3) PoU - Proof-of-utility, quite similar to DPoS, in this
the blockchain in the edge or create a hybrid Fog-Edge algorithm we run an election, where the nodes vote
or Fog-Edge-Cloud architecture. This was done to avoid for which mining node has the higher utility. This
overtaxing the vehicles with processing and storage utility can be calculated using multiple variables that
• Consensus Algorithm - Here we specify which consensus change their weight over time, making these vari-
algorithm was used in each surveyed paper. ables non-linear distributed. These variables could
include stake, participation rates, seniority, this can
1) PoW - Proof-of-work, one of the most known
vary as it depends on the implementation. PoU is
consensus algorithms, used by BitCoin. Every node
scalable and safe. [12] authors calculated this Utility
competes with one another to solve a challenge, the
based on the amount of time the node takes to
nodes with the best computational power usually
provide information for the vehicle. The main issue
win the competition and generate the new block.
with this approach is how the Utility is calculated.
This makes it harder for attackers since the cost
Since the number of variables and their weights is
of attacking the chain, trying to pass a new block
all open for the developer to choose, a bad choice
with fraudulent transactions is too high. But that
could cause a major security issue. So, it’s hard to
puts a strain in the mining nodes since for every
implement it safely and correctly.
consensus there is a need for competition. In a
4) PoR - Proof-of-Reputation. The PoR devised in [11]
VANET environment, if a researcher chooses this
is really similar to PoU or DPoS. There is an elec-
method, then it would be best to leave the mining
tion, and the node reputation is directly proportional
outside the vehicles. [5], and [13] both used PoW,
to its voting power. The main difference is that,
but combined with another strategy, so the difficulty
after the election, the elected nodes won’t mine,
of the challenge would be smaller for certain indi-
they will compete in a PoW challenge, and only
viduals. This eases the energy consumption burden
the winner is going to be rewarded. The higher their
since the challenge would be solved faster. [11]’s
reputation, the smaller the difficulty of the challenge
PoR also uses PoW competition in the end. [4] also
is. This combination of DPoS and PoW suppress
used a combination of algorithms, PoW and Proof-
most of PoW shortcomings, since we don’t need all
of-Storage, the nodes that are using the most storage
the network to compete, and those who do compete
gain extra coins each consensus. But the authors
don’t expend that much energy since the difficulty
didn’t change PoW difficulty and its disadvantages
decreases as the reputation increases. But there’s a
remain the same.
risk of centralization
2) PoS - Proof-of-Stake, also a very common consen-
5) PBFT - Proof Bizantine Fault Tolerance is a non-
sus algorithm used by NXT cryptocurrency. This
anonymous, message heavy algorithm. If 7 nodes
algorithm came to try to solve PoW inherent issue
are trying to reach consensus 71 messages are
of energy consumption. This approach chooses the
necessary. If a node is considered in the traditional
miner based on how many coins the node has
sense, as a RSU or a vehicle, then it doesn’t scale.
accumulated. The main issue is the lack of variation
 It is possible, however, in a consortium blockchain the guarantee of safety in the use of integrated technologies of
where each organization elects a single node to ITS. This study sought to identify the types of blockchain, con-
participate in the consensus. The paper [1] does not sensus algorithms, as well as their uses, contexts, and operation
mention this approach. in the environment (cloud, edge, fog, and hybrid). We sought
6) BFT-DPos - Bizantine fault tolerance delegated to categorize the different usage scenarios, taking into account
proof-of-stake. This algorithm puts together two their various positive and restrictive aspects of each proposal,
concepts; the Delegated proof-of-stake is used to be it communication, assistance, security, application, consen-
choose the nodes that will try to achieve consensus- sus, and intelligent decision-making technologies, all of these
based on their stake. After choosing 21 nodes, these aspects are present in Figure 2, then we classified each paper
nodes will continuously produce 12 blocks. And into one of these scenarios. We discussed each of the proposals
for each one they apply the second concept, the of Table 1, delivering a comparative study of how blockchain
BFT to achieve consensus, verifying the blocks. and edge computing are being used in IoV, how they are being
This is a scalable and fast way to generate blocks. designed, which algorithms and technologies were used, listing
Since the users have a tiny chance to influence the their advantages and disadvantages. In summary, this survey
result, in a real-life scenario, is likely that the users aimed to guide and assist researchers, analysts, developers in
would deposit their money on an exchange, so the their decision making of current solutions or future proposals
exchange can vote for them. That can cause voting for the IoV context. In the future we will carry out an analysis
centralization, this issue is commonly seen in DPoS. on delay and energy efficiency of those consensus algorithms
7) TangleCV - To understand the tangleCV policy, in an unified simulation framework, so they can be properly
it’s necessary to explain the tangle architecture. compared.
It’s based on blockchain, but it takes a different
R EFERENCES
approach. The blocks might be chained in an acyclic
way, each block has only one transaction and each [1] Gao, J. et al. A blockchain-sdn enabled internet of vehicles environment
for fog computing and 5g networks. vol. 7, 4278–4291 (2020). Doi:
new transaction requires PoW validation of two 10.1109/JIOT.2019.2956241.
previous transactions. After validation, a transaction [2] XiaoDong Zhang, R. L. & Cui, B. A security architecture of vanet
gains a weight calculated based on the amount of based on blockchain and mobile edge computing. (in Proc. 1st IEEE
Int. Conf. Hot Inf.-Centric Netw 258–259 (2018).
necessary work to validate that transaction, and this [3] H, R. et al. Tanglecv: Decentralized technique for secure message shar-
weight is its trustworthiness. The tangleCV Policy ing in connected vehicles. ACM (2019). Doi: 10.1145/3309171.3309177.
states that a transaction that has a weight higher than [4] Fu, Y. et al. Blockchain-enabled security in electric vehicles cloud
and edge computing. IEEE Network 32, 78–83 (2018). Doi:
a threshold might self-validate, otherwise it has to 10.1109/MNET.2018.1700344.
do a PKI authentication to validate. In the tangle [5] Kang, J. et al. Blockchain for secure and efficient data sharing in
architecture, this policy has many advantages such vehicular edge computing and networks. vol. 6, 4660–4670 (2019).
Doi: 10.1109/JIOT.2018.2875542.
as reducing bandwidth congestion and removing [6] Guo, S. et al. Trust access authentication in vehicular network based
delays. But there was no security test done in [3]. on blockchain. vol. 16, 18–30 (2019).
Also, it’s difficult to compare this policy with other [7] Tan, H. & Chung, I. Secure authentication and key management with
blockchain in vanets. IEEE Access, vol. 8, pp. 2482–2498 (2019). Doi:
consensus algorithms, since the architecture is a 10.1109/ACCESS.2019.2962387.
little different. [8] Yahiatene, Y. et al. A blockchain-based framework to se-
8) DM-CDS - The Distributed miners connected dom- cure vehicular socialnetworks. Trans. Emerging Telecommun.
Technol, vol. 30, no. 8, [Online]. Available: https:// onlineli-
inating set algorithm is based on another algorithm, brary.wiley.com/doi/abs/10.1002/ett.3650 (2019).
DSP-CDS or Distributed single-phase connected [9] Qiu, C. et al. Blockchain-based distributed software-defined vehicular
dominating set. DSP-CDS is used to decide if a networks via deep q-learning. in Proc. ACM DIVANet, Montreal, QC,
Canada, Nov. 2018, pp. 8–14 (2018).
given node should enter a set of dominating nodes, [10] Yuchuan Fu, F. R. Y. et al. Vehicular blockchain-based collective
it adapts well to changes in the network topology. learning for connected and autonomous vehicles. IEEE (2019). Doi:
DM-CDS uses a miner score function that considers 10.1109/MNET.001.1900310.
[11] Wang, Y. et al. Challenges and solutions in autonomous driving:
several parameters, such as a trust model, degree of A blockchain approach. IEEE Network 34, 218–226 (2020). Doi:
connectivity, average link quality, to decide which 10.1109/MNET.001.1900504.
node may validate the transactions. This algorithm [12] Dai, Y. et al. Permissioned blockchain and deep reinforcement learn-
ing for content caching in vehicular edge computing and networks.
is a little similar to PoS where instead of stake, 19243175 (2019). Doi: 10.1109/WCSP.2019.8928099.
the main factor that decides is the miner score. [13] V. De Maio, B. U., R. & Brandic, I. Energy and profit-aware proof-
And since the miner score is variable with time, of-stake offloading in blockchain-based vanets. ACM 177–186 (2019).
Doi: 10.1145/3344341.3368797.
the miners would variate more than in PoS.
IV. C ONCLUSION
The Internet of Vehicles (IoV) will play an important role
in future smart cities, providing road safety, convenience and
several other benefits. Therefore, it is necessary to search for