AI-Powered Blockchain - A Decentralized Secure

Multiparty Computation Protocol for IoV

1 Gunasekaran Raja, 2 Yelisetty Manaswini, 3 Gaayathri Devi Vivekanandan, 4 Harish Sampath,
5 Kapal Dev, 6 Ali Kashif Bashir
1,2,3,4 Department of Computer Technology, Anna University, Chennai, India,
5 CONNECT Centre, Trinity College Dublin, Ireland,
6 Department of Computing and Mathematics, Manchester Metropolitan University, UK,
1 [email protected], 2 [email protected], 3 [email protected], 4 [email protected],
5 [email protected], 6 [email protected]

Abstract—The rapid advancements in autonomous technolo- a need to consider decentralized management and distributed
gies have paved way for vehicular networks. In particular, storage as the future technology. This is the current progres-
Vehicular Ad-hoc Network (VANET) forms the basis of the sion of technology towards the next generation of IoV [8].
future of Intelligent Transportation System (ITS). ITS represents
the communication among vehicles by acquiring and sharing The decentralized technology demands high security for the
the data. Though congestion control is enhanced by Internet of exchange of data and communication between vehicles.
Vehicles (IoV), there are various security criteria where entire As a solution to the issue of secure sharing of information
communication can lead to many security and privacy challenges. among vehicles, blockchain can be used to solve the problem
A blockchain can be deployed to provide the IoV devices with of secure information exchange between vehicles [10]. Un-
the necessary authentication and security feature for the transfer
of data. Blockchain based IoV mechanism eliminates the single like the traditional client- server model, blockchain supports
source of failure and remains secure at base despite having strong peer-to-peer (P2P) communication. The mining nodes of the
security, the higher level layers and applications are susceptible blockchain have a ledger which holds all of the transactions
to attacks. Artificial Intelligence (AI) has the potential to over- that took place and contains chained blocks. Therefore, a se-
come several vulnerabilities of current blockchain technology. cure log is created with timestamped records that can never be
In this paper, we propose an AI-Powered Blockchain which
provides auto coding feature for the smart contracts making altered because of proof of work (PoW). Blockchain solves the
it an intelligent contract. Moreover, it speeds up the transaction problem of trusted interactions using decentralized approach
verification and optimises energy consumption. The results show and provides the necessary authentication and authorization to
that intelligent contracts provide higher security compared to IoV devices. It is also a cost effective approach as it reduces
smart contracts considering range of different scenarios. the deployment, performance overhead and operational costs
Index Terms—Blockchain, Artificial Intelligence, Smart con-
tract, Internet of Vehicles, Vehicular Network of IoV.
The smart contract is the heart of the blockchain, where
the rules regarding the interactions are presented [13]. It
I. I NTRODUCTION
digitally facilitates, verifies and enforces the performance of
Modern lifestyle has led to an extensive usage of private the contract. Using smart contracts, the need for the third
transport vehicles that saves a lot of time and gives owners the parties to validate the transaction is eliminated. All the users
required privacy. In this era, there is a need for future vehicles of the blockchain network can view the smart contract as this
to be fully autonomous, more comfortable and greener. Since enables more people to check that the code is good but may
the impact of technology on our lives keep increasing with lead to large security holes and attacks. These attacks include
time, VANET has evolved to meet the growing demand [8]. escalation of funds if proper care is not taken into consid-
A large range of VANET applications including road safety, eration to create the smart contract rules. In consequence,
car services related to manufacturing units, optimization of the proposed decentralized AI-Powered blockchain is designed
vehicular traffic and passenger infotainment can be enhanced which is basically a combination of AI and blockchain.
using wireless communication. The AI-Powered blockchain is capable of processing the
The development of Internet of Vehicles (IoV) technology stored data and intelligently enables decision making which
leads to a large number of vehicular nodes accessing the predicts the nature of the interaction and also the resources
network. This centralized system can handle large amount it utilizes. The decision outcomes are validated by the trusted
of traffic generated from vehicular interaction. As the traffic mining nodes of blockchain. Integration AI and blockchain has
load increases on the traditional centralized server, it faces benefits such as data security enhancement and improving trust
significant challenges [1]. The whole system may crash if the in the decisions made by AI [17]. In this paper, the auto coding
central server fails, leading to huge disruption. So, there is feature for the smart contract is provided by the integration

865
Authorized licensed use limited to: University of Wollongong. Downloaded on August 12,2020 at 04:44:11 UTC from IEEE Xplore. Restrictions apply.
of AI with blockchain, thus making the smart contract an contracts which is decentralized trusted shared code [13]. The
intelligent contract. These intelligent contracts make use of smart contracts enforce rules based on which interactions take
Natural Language Processing (NLP) to auto code and use place. The execution of smart contracts and their environments
the previous data to make smart decisions. The proposed AI- are also vulnerable to attacks. AI is used to overcome these
Powered Blockchain ensures that the implementation of a challenges which boosts the efficiency of the system by many
smart contract is free of bugs and secure against attacks. The folds.
potential vulnerabilities are ruled out as they are automatically AI gives the opportunity to tackle blockchain tasks in a
handled by proposed intelligent contract. very intelligent and efficient way [14]. Integrating AI and
blockchain benefits the entire system by making use of low
II. R ELATED W ORK computational power and creates diverse datasets for further
VANET is a subset of larger set of vehicular networks called efficient processing. Transportation system has been embedded
Mobile Ad-hoc networks (MANET). Vehicles communicate with intelligence to meet the requirement of easy movement
with each other supporting highly dynamic networking topol- and interaction among vehicles [15,16]. A more trusted cyber
ogy, but this communication whilst enhancing the passenger’s space is created by this powerful combination that ensures
safety and comfort is vulnerable to many attacks [1]. security by introducing many intelligent rules [17]. These rules
For dynamic security association, security protocols are are automatically generated by AI using enormous amount of
proposed, which provides high robustness and efficient key data retrieved from the blockchain.
management systems [2,3]. Since the vehicles interact amongst
themselves through internet protocols, the Internet of Things III. P ROPOSED W ORK
(IoT) network is formed. The 5G technology has emerged In this section, we propose an AI-Powered blockchain which
providing data speeds a hundred times faster than 4G and solves the problem of traditional issues in the existing system
provides a better bandwidth. It is combined with Software- to create a safe and secure VANET framework. The proposed
Defined Network (SDN) architecture for information gathering system leverages the power of AI and a robust blockchain
at global level [4,5]. network to protect assets from security breaches and attackers.
Blockchain is a distributed, decentralized value-exchange This AI-Powered Blockchain system is applicable to both
protocol which records transactions across a huge network and static and dynamic VANET environment.
prevents alteration of any record as it requires changes to sub-
sequent blocks. The entire working mechanism of blockchain A. Blockchain Based IoV Network
makes it more adaptable to IoT . There are a lot of benefits In IoV network, safe driving and better service quality is
in combining the blockchain with IoT such as efficient fraud achieved by sharing data among the vehicles in a lane. The
management and supply chain management [6]. traditional centralized management structure in IoV requires
Smart cities are developed as a result of the enhanced large data and information storage. This requirement puts
technology advancement by the blockchain. A smart city forth many challenges and real-time responses that cannot
is an architectural setup that overcomes the challenges of be efficiently dealt with the existing architecture. In addition,
urbanization by combining new technologies [7]. This builds data manipulation of personal information uploaded to the
an intelligent transportation system in the smart city which infrastructure can be a hindrance for the future development
is secure and autonomous. Security is ensured by preventing of IoV. Thus, blockchain technology makes the system more
data forgery and personal information breach by making use secure, scalable and fault tolerant.
of blockchain technology for vehicular networking [8]. Social Blockchain addresses many failures and scalability bot-
networks require highest level of privacy protection to prevent tlenecks. As the number of vehicles increase, there is an
information leak to malicious users [9]. equivalent increase in the number of interactions in the IoV
Various security services like authentication, integrity as- network. As a consequence, there is a need for a P2P net-
surance and confidentiality are automated using the blockchain work instead of the traditional client-server model which is
[10]. This resolves the challenges of the state of art techniques fulfilled by blockchain technology. Currently, the integration of
used in the current applications. In [11], a trusted connection is blockchain with IoV is in the limelight because of the trusted
established between the interacting entities using the universal characteristics of blockchain. Though the entire process is still
data object identifier platform enabling secure digital object complicated, it provides various features like decentralized
management. distributed processing, anonymity, trusted authentication and
In order to make the entire system more effective, the verification required for the IoV network to succeed.
user should be aware of the blockchain consensus algorithms In the VANET environment, Road Side Unit (RSU) is de-
being used and also the blockchain taxonomy. The technical ployed with blockchain as the controller. RSU helps in storing
challenges as well as the recent advancements in tackling the the information of the passing vehicles in the blockchain. Since
challenges should be known. A comprehensive survey and the RSU, petrol pumps, toll gates etc., are static, they act as
review on blockchain is conducted to analyze and adapt the miner nodes in blockchain. Each vehicle in the blockchain
technology based on the specific needs of the application network is connected to nearby peer vehicles as depicted in
in [12]. Blockchain interactions take place through smart Fig. 1. Data stored in toll gates, petrol pumps, RSU is also used

866
Authorized licensed use limited to: University of Wollongong. Downloaded on August 12,2020 at 04:44:11 UTC from IEEE Xplore. Restrictions apply.
 • Blockchain creates a strong base layer but the higher
end application layers have potential vulnerabilities which
may get exploited by hackers causes irreparable loss.
AI tremendously improves the deployment of blockchain
application by predicting possible system breaches and
enhances the total security of the system.
• Every node in the blockchain has a distributed ledger
that has all the decisions made by the particular node.
This ledger data becomes available to AI for analysis and
processing. The integrity of data recorded for examination
is ensured because blockchain is tamper resistant.
• Blockchain’s smart contract code can be viewed by all the
nodes taking part in the transaction. Therefore, it is vul-
nerable to attacks and can be exploited by hackers. These
attacks can be prevented using AI which can predict
the possible vulnerabilities and introduce new intelligent
rules, improving the nature of the smart contract.
• By adopting intelligent contracts, in VANET environment
when there is an increase in number of mining nodes,
compared to other existing approaches, our proposed AI-
Fig. 1. Blockchain-based IoV Powered blockchain decreases the processing time by
25%.
• Transactions can be verified faster, energy consumption
by vehicles for easier communication with the infrastructure. can be optimized and smarter smart contracts are used
All the vehicles are registered independently in the blockchain which benefits the blockchain.
network. When the vehicles register with the help of RSU,
ethereum blockchain transactions take place. A small amount 1) Ethereum setup: AI-Powered Blockchain is modelled on
of fee is charged to register them into the network. Once Ethereum, an open source platform for distributed applications.
registered, the vehicle will broadcast an event message such It is a decentralized account-based blockchain implementation.
as traffic jam, road conditions, shortest path, accident, etc. to There are two types of accounts namely Externally Owned
neighbouring vehicles. Miner nodes in the blockchain network Account (EAO) and contract account. The EAO triggers the
validate the authenticity of the message before propagating it contract account. Contract account is the byte code and it in-
to the peer vehicles. The vehicle will broadcast the message to curs the cost for its valuable computation and storage resources
the peers, if the event message is validated to be true. Thus, the of the network. Ethereum allows people to safely communicate
tamper proof nature of the blockchain stores and manages the in a peer-to-peer fashion. A runtime environment is created
event messages. Every information recorded on the blockchain by Node Package Manager (NPM). Ethereum transaction is
becomes a verified transaction, once it is validated and signed demonstrated using the Ganache simulator. The ganache is set
by the neighbouring nodes. Various cryptographic algorithms up with user accounts loaded with fake ether. Ether is the
are employed by the blockchain which helps in the generation digital currency used in the blockchain transactions.
of the hash key. Each block is sequentially ordered based 2) AI-Powered Blockchain dependencies: The ethereum
on their hash values. The broadcasted message gets updated environment is set up and various dependencies necessary to
in each vehicle’s block in the blockchain network. There implement the blockchain are installed. NPM is the basic need
are millions of vehicles and if each region forms a separate to install all other dependencies for the Blockchain-based IoV
blockchain, there will be a significant decrease in scalability environment. Secondly, the truffle framework that has a suite
issues. of tools to write smart contracts is incorporated. The client side
application is developed using this truffle framework and also
B. AI-Powered Blockchain allows the user to test and deploy the smart contracts. Smart
The current problem with the blockchain technology is, it contracts are developed using Solidity, an object-oriented,
can never create mass adoption because of its complexity. high-level programming language. The next dependency is
AI and blockchain could be a splendid combination because Ganache, a local blockchain setup, used in the creation of
of highly sensitive and large amount of data is stored in decentralized applications and useful for testing the desktop
blockchain which is to be supplied for the AI engine to make applications. To make use of the ethereum blockchain setup,
statistical and analytical decisions. Blockchain provides the Metamask extension of google chrome is enabled. It is used to
data to be processed by AI for predictive analysis and also interact with the smart contract and also runs the decentralized
improves the performance of the system on a great scale. The ethereum application in the browser by providing a secure
AI-Powered Blockchain envisages the following key points: vault.

867
Authorized licensed use limited to: University of Wollongong. Downloaded on August 12,2020 at 04:44:11 UTC from IEEE Xplore. Restrictions apply.
 3) Intelligent Contracts: Smart contract is the vital com- Algorithm 1 Classification of text
ponent of the blockchain environment but they are not smart Input: Vocabulary for the vehicular user request
enough as they may come with several vulnerabilities. The Output: Probabilities of the output classes
smart contract code consists of rules that are executed when 1: procedure T RAIN M ULTINOMIAL (C, V EH R EQ )
certain conditions are met. Since the smart contract code is 2: Voc ← VocabularyExtraction(Veh Req)
public, any malicious intruder can go through each line of the 3: N ← Veh ReqCount(Veh Req)
code patiently in search of loopholes. Once the loopholes or 4: for cl ∈ C do
the technical flaws are found out, they can be taken advantage 5: Ncl ← ClassWise Veh Reqs Count(Veh Req,cl)
of and exploited the whole network such that funds can be 6: prior[cl] ← Ncl /N
malevolently released or transferred. The attackers can also 7: textcl ← TextConactenationOfAllVeh ReqsInClass
attack once they know how the events take place in the 8: for t ∈ Voc do
VANET environment. To enhance the smart contract and make 9: Tct ← TokensCountOfTerm(textcl , t)
it intelligent, we integrate AI with blockchain with the help 10: for t ∈ Voc do P
of dependencies. 11: condprob[t][cl] ← Tct + 1/ t0 Tct0 + 1
In our proposed AI-Powered Blockchain model, the de- 12: return V oc, prior, condprob
centralized ethereum application uses an autocoded intelligent
contract which tremendously improves its efficiency. We used
Algorithm 2 Prediction of the output class
NLP to autocode the smart contract and Naive Bayes classifier
to predict the tag for the each user request as illustrated in Input: Calculated probabilities
algorithm 1. For making intelligent decisions for each request, Output: Classified text is obtained
the contracts gets influenced by historical data and prototyping 1: procedure A PPLY M ULTINOMIAL (C, VOC , PRIOR ,
concepts. Thus, for a particular application, each vehicular user CONDPROB , V EH R EQ )
needs to specify the features to be included in their respective 2: P ← T okensExtraction(V oc, V eh Req)
block in the blockchain network. The coding of the contract 3: for cl ∈ C do
is then automatically handled by AI techniques. The proposed 4: result[cl] ← logprior[cl]
system makes the smart contract intelligent by adopting the 5: for t ∈ Voc do
text classification and autocoding procedure as illustrated in 6: result[cl] ← result[cl] + logcondprob[t][cl]
the below algorithms. 7: return argmaxcl∈C result[c]
P (V |Cn )P (Cn )
P (Cn |V ) = (1)
P (V ) vehicular user input information t is appearing in the textcl .
where, V = (v1 , v2 , ..., vn ) is the feature vector, For each class, the conditional probability is calculated for all
P (Cn |V ) is the posterior probability, the tokens as mentioned in equation 3.
P (V |Cn ) is the likelihood, Tct + 1
condprob[t][cl] = P (3)
P (Cn ) is the prior probability of class, t0 Tct0 + 1

P (V ) is the prior probability of the predictor

All the features received as input from the user are not de-
pendent on each other. There exists conditional independence
among the set of features. Since Naive Bayes algorithm also
works with the naive assumption of conditional independence,
it is the most suitable for generation of the contract.
The probability of a Vehicular Request (Veh Req) belong-
ing to a class cl is calculated. The maximum likelihood
estimate is then found out to estimate the parameters. In
algorithm 1, the vocabulary for the given vehicular user request Fig. 2. Vehicular Request classification
is extracted as a set of tokens (t1 , t2 ,. . . , tn ) and used for
classification as depicted in Fig. 2. N is the total count of To predict the class of the Veh Req, tokens are first ex-
Veh Reqs. Ncl is the number of Veh Reqs belonging to a tracted as shown in algorithm 2. For each class, the result
particular class cl. The prior probability for each class cl is variable stores the log of prior probability. The log of condi-
calculated as follows: tional probability of each token is then added to the class prior
Ncl probability as a final result. The class with maximum score is
prior[cl] = (2)
N chosen as the predicted class to which the Veh Req belongs.
textcl is the text concatenated from all the Veh Reqs belong- Once the Veh Req is classified, we use the Bayesian
ing to that class cl. Tct stores the count of number of times the network to autocode the contract, making it an intelligent

868
Authorized licensed use limited to: University of Wollongong. Downloaded on August 12,2020 at 04:44:11 UTC from IEEE Xplore. Restrictions apply.
contract. Bayesian networks is used for probabilistic queries
as it represents an entire model comprising of conditional
dependencies via a directed acyclic graph. The joint proba-
bility distribution over all the variables of a Bayesian network
model, P(V), is the product of the probability distributions
over each of the nodes conditional on their parents as shown
in equation 4. To calculate P (E = e), we need to sum over
all P (V \E, E = e) as shown below:
n
Y
P (V ) = P (V i|P a(V i)) (4)
i=1
X
P (E = e) = P (V \E, E = e) (5)
V \E

where, E is the evidence variable, Fig. 3. Ethereum transaction

P a(V ) denotes the parent of node V
The parameter class id (c id) is the predicted output class IV. P ERFORMANCE A NALYSIS
from algorithm 2. c id represents the type of Veh Req and
the rules that have to be framed. The rules pertaining only to In this section, we compared the performance of our AI-
that class are involved in the generation of intelligent contract. Powered Blockchain system with the existing smart contract
Initially, the various events in VANET environment are in blockchain. The comparison is made to depict the security
arranged in topological order. For every event, initialize the with respect to various scenarios. The vulnerabilities in smart
following variables: P r0 (X\E) represents the importance contracts have led to major attacks causing a lot of loss to
function, t represents the time interval and n represents the the organizations deploying blockchain. With an advancement
number of samples. Each solution si is generated within t of technology, the attackers have also geared up and they are
seconds and appended to the set of solutions S. The score exploiting the loopholes over the platform compromising the
function calculates the score of every solution and then security in networks. AI makes the system learn through out its
makes a corresponding entry in the score array. The score lifetime and improves the results taking the network security
is calculated based on parameters like time complexity, space to peaks. The auto coding feature of smart contracts using AI
complexity, ordering of the rules, etc. We normalize the score overcomes many of the challenges faced in the current system
array for every event and the one with the maximum score is paving way to more efficient, reliable and secure system.
chosen as the final solution as illustrated in algorithm 3. Thus, For every vehicle being registered, an ethereum blockchain
the proposed AI-Powered Blockchain system makes the smart transaction takes place as shown in Fig. 3. The transaction win-
contract intelligent by adopting the autocoding procedure and dow consists of the amount being charged for that transaction.
improves the overall security in VANET environment. This amount is also known as the gas fee in the blockchain
world. Gas fee is charged only for the transactions that require
Algorithm 3 Autocoding smart contract writing into the blockchain. Amount is not deducted for the
Input: Classified text is given as input transactions that require only fetching the vehicular user data.
Output: Autocoded smart contract The smart contract and intelligent contract generated for
1: procedure AUTOCODE (cid ) fetching details of a vehicle is shown in Fig. 4. The proposed
2: Topological sorting of the events intelligent contract is generated uniquely and dynamically
3: Initialize Pr0 (X\E), t, n, and the score from the given user request. The code is constructed by clas-
4: c ← 0, S ← ∅ sifying the input using classifier algorithm. In order to select
5: for j ← 1 to n do the finest classifier algorithm, comparison is made based on
6: if j mod t == 0 then performance parameters namely F-measure, accuracy and area
7: c←c+1 under curve (AUC). Naive Bayes classifier has approximately
8: Update Pr0 (X\E) based on S 20% increase in the performance parameters compared to
k-nearest neighbors (KNN) algorithm and Decision Tree as
9: end if
shown in Fig. 5. Hence, Naive Bayes classifier is used to
10: si ← sample generated in accordance to Prc (X\E)
predict correct class to which the given Veh Req belongs, and
11: S ← S ∪ si
the contract will be autocoded accordingly.
12: Calculate the score for every solution (si )
From Fig. 6, the security for various blockchain scenarios
13: Normalize the score arrays for every event
such as money transfer, fraud deterrent, creating transparency,
14: return max(S)
syndicated lending, anti-money laundering is tested. It can be
seen that the percentage increase in security is comparatively

869
Authorized licensed use limited to: University of Wollongong. Downloaded on August 12,2020 at 04:44:11 UTC from IEEE Xplore. Restrictions apply.
 Fig. 6. Smart contract Vs Intelligent contract

R EFERENCES
[1] H.P. Dai Nguyen and R. Zoltán, ”The Current Security Challenges of
Fig. 4. Smart contract Vs Intelligent contract Vehicle Communication in the Future Transportation System,” in IEEE
16th International Symposium on Intelligent Systems and Informatics
(SISY), Subotica, pp. 0161-0166, 2018.
[2] R. Arul, G. Raja, A. O. Almagrabi, M. S. Alkatheiri, S. H. Chauhdary and
A. K. Bashir, ”A Quantum Safe Key Hierarchy and Dynamic Security
Association for LTE/SAE in 5G Scenario,” in IEEE Transactions on
Industrial Informatics, July 2019.
[3] R. Arul, G. Raja, A. K. Bashir, J. Chaudry and A. Ali, ”A Console GRID
Leveraged Authentication and Key Agreement Mechanism for LTE/SAE,”
in IEEE Transactions on Industrial Informatics, vol. 14, no. 6, pp. 2677-
2689, June 2018.
[4] L. Xie, Y. Ding, H. Yang and X. Wang, ”Blockchain-Based Secure and
Trustworthy Internet of Things in SDN-Enabled 5G-VANETs,” in IEEE
Access, vol. 7, pp. 56656-56666, 2019.
[5] G. Raja, A. Ganapathisubramaniyan, S. Anbalagan, S. B. M Bhaskaran,
K. Raja and A. K. Bashir, ”Intelligent Reward based Data Offloading in
Next Generation Vehicular Networks,” in IEEE Internet of Things Journal,
Fig. 5. Classifiers comparison DOI:10.1109/JIOT.2020.2974631, 2020
[6] K. Christidis and M. Devetsikiotis, ”Blockchains and smart contracts for
the internet of things,” IEEE Access, vol. 4, pp. 2292–2303, 2016.
[7] P. K. Sharma, S. Y. Moon, and J. H. Park, ”Block-VN: A distributed
high when intelligent contract is used in the place of smart con- blockchain based vehicular network architecture in smart city, ” in J. Inf.
Process. Syst., vol. 13, no. 1, pp. 184–195, Mar. 2017.
tract. Hence, the proposed intelligent contract of AI-Powered [8] T. Jiang, H. Fang and H. Wang, ”Blockchain-Based Internet of Vehicles:
Blockchain is highly secure than smart contracts. Distributed Network Architecture and Performance Analysis,” in IEEE
Internet of Things Journal, vol. 6, no. 3, pp. 4640-4649, Jun. 2019.
[9] R. Yu et al., ”Authentication With Block-Chain Algorithm and Text
V. C ONCLUSION Encryption Protocol in Calculation of Social Network,” in IEEE Access,
vol. 5, pp. 24944-24951, 2017.
In this paper, we investigate how the AI-powered [10] T. Salman, M. Zolanvari, A. Erbad, R. Jain, and M. Samaka, ”Security
Blockchain technology can be applied in IoV applications services using blockchains: A state of the art survey,” in IEEE Commu-
nication Surveys Tutorials, 2018.
to overcome the security challenges. Our proposed scheme [11] H. Yin, D. Guo, K. Wang, Z. Jiang, Y. Lyu and J. Xing, ”Hypercon-
creates more efficient and reliable system with the combination nected Network: A Decentralized Trusted Computing and Networking
of AI with blockchain. The auto coding feature of intelligent Paradigm,” in IEEE Network, vol.32, no. 1, pp. 112-117, Jan.-Feb. 2018.
[12] T. M. Fernández-Caramés and P. Fraga-Lamas, ”A Review on the Use
contract is implemented using Bayesian networks. This creates of Blockchain for the Internet of Things,” in IEEE Access, vol. 6, pp.
a better blockchain network and more security rules are framed 32979-33001, 2018.
through AI-powered Blockchain during the lifetime of the [13] S. Wang, Y. Yuan, X. Wang, J. Li, R. Qin, and F. Y. Wang, ”An overview
of smart contract: architecture, applications, and future trends,” in IEEE
system. The future is towards more secure intelligent systems Intelligent Vehicles Symposium (IV), pp. 108–113, 2018.
and our paper contributes to the field of vehicular networks [14] K. Salah, M. H. U. Rehman, N. Nizamuddin and A.Al-Fuqaha,
using AI. ”Blockchain for AI: Review and Open Research Challenges, ” in IEEE
Access, vol. 7, pp. 10127-10149, 2019.
[15] S. Saharan, S. Bawa and N. Kumar, ”Dynamic pricing techniques for
ACKNOWLEDGEMENT Intelligent Transportation System in smart cities: A systematic review,”
in Computer Communications, vol. 150, pp. 603-625, 2020
This Publication is an outcome of the R&D work undertaken [16] A. Miglani and N. Kumar, ”Deep learning models for traffic flow
prediction in autonomous vehicles: A review, solutions, and challenges,”
in the project under the Visvesvaraya PhD Scheme of Ministry in Vehicular Communications, vol. 20, 2019.
of Electronics Information Technology, Government of India, [17] K. Wang, J. Dong, Y. Wang and H. Yin, ”Securing Data With Blockchain
being implemented by Digital India Corporation (formerly and AI,” in IEEE Access, vol. 7, pp. 77981-77989, 2019.
Media Lab Asia).

870
Authorized licensed use limited to: University of Wollongong. Downloaded on August 12,2020 at 04:44:11 UTC from IEEE Xplore. Restrictions apply.