Content uploaded by Nadeem Javaid
Author content
All content in this area was uploaded by Nadeem Javaid on Oct 14, 2019
Content may be subject to copyright.
Blockchain based Vehicular Trust Management and
Less Dense Area Optimization
Shahid Abbas, Nadeem Javaid∗
COMSATS University Islamabad, Islamabad 44000, Pakistan
∗Correspondence: nadeemjavaidqau@gmail.com
Abstract—In the vehicular ad hoc network, the variation in the
network is increasing rapidly due to convergence of cities into a
smart city. With the rapid increase of vehicles, many problems
arise in the network such as cost, scalability, security and trust
management. To address these problems, we proposed the trust
management system through blockchain based authentication of
the vehicles by assigning unique identification. If still malicious
vehicles exist in the network, trust is achieved by calculating
the data that is sent from the vehicles so that network does not
get stuck due to malicious data. In addition, we also proposed
the crowdsensing concept using vehicles’ sensors data and we
minimize the cost of deploying the sensors and also optimize
the density of area with respect to chance minimum occurrence
of vehicles. We performed these solutions on the etheriam
environment. vehicle to vehicle and vehicle to infrastructure
communication got secured from malicious vehicles and malicious
data. In crowdsensing, cost is reduced as compared to the wireless
sensor network.
Index Terms—Blockchain, Crowdsensing, Intelligent Vehicles,
Repeater
I. INTRODUCTION
From previous few decades, vehicular ad-hoc network and
vehicular automation are the very hot topics in research. The
concept of smart city carries the Vehicular Network (VN) on
the top, which is totally about the communication of vehicles,
i.e., sharing of required data with Vehicle to Vehicle (V2V) and
Vehicle to Controller (V2C), to generate the fully cooperative
environment using automation and intelligent VN. There are
many concepts proposed by different researchers in which a
centralized vehicular network is more prominent. However, the
word centralized is usually not suitable for mobile devices [1].
We can use centralized concept in a static environment while
vehicles are mobile devices so, a centralized network faces
many problems in communicating with their nodes that change
their position instantly [2]. However, this concept works for
limited defined areas, which is not much useful. Usually, cities
have to cover a large area that should be smart. So, there are
limited implementations of a VN that work with a centralized
approach because of above mentioned issues.
In our proposed idea, we used the concept of blockchain
in which every block of the chain is interconnected with
each other in chronological order using hashing [3]. Each
block contains the hash of its previous block. So, if someone
maliciously wants to change the ledger of the nodes, the hash
of every block will change automatically. However, blockchain
is much secure and reliable due to the decentralized approach.
The core benefit of using blockchain is decentralization. How-
ever, the security and integrity of data are also achieved [4].
Moreover, Peer to Peer(P2P) transaction on occurrence of any
event is also be achieved. Nodes trusted each other without
knowing, they share resources by paying cryptocurrency.
Vehicles must be intelligent concerning to the environment
and network. Intelligent means vehicles must be capable of
deciding the active environment. A decision must be accu-
rate using different sensors and availability of travel parallel
vehicles’ data that [5]. In centralized approach, it is difficult
for the environment to exchange the data between vehicles
because every vehicle shares the data with a central body,
a decentralized approach is necessary to be involved so that
other vehicles get able to fetch and share the data [1]. However,
this phenomenon is time consuming, costly and insecure. Our
proposed idea about vehicular communication is using the
blockchain technology in which every node of the network
works as the block of the chain that interacts with the remain-
ing network directly. Now vehicles communicate directly with
each other to exchange the required data in real time, sharing
vehicles will take incentive via digital currency. Therefore,
vehicles decide in real-time and behave like intelligent vehicles
[6].
Trust management of vehicles is the phenomenon in which
all vehicles should have to authenticate first to communicate
with the rest of the network. If still any malicious vehicles
exist In the network and send malicious data toward the
blockchain so, this becomes difficult for the vehicles to decide
whether data is correct or incorrect. To overcome this issue,
we proposed an idea in which we decide about the vehicles
who send correct data. We may check the correctness of the
data by counting and make groups of the data through defining
the threshold and Send correct data to the other vehicles that
require [7].
Crowdsensing is the process in any requester that requests
the data from a large number of people or devices in the form
of input [8]. This fetched data might be paid or unpaid depends
upon the agreement. This crowdsensing is usually used by
the Wireless Sensor Network (WSN) [1]. While WSN costly
due to deployment of a very big amount of sensors to sense
the different types of data. We have an automated network
of vehicles that are not used for any other purposes, due to
blockchain characteristics, we can use vehiles’ sensors data for
different purposes or sell the data to different companies that
have required. Using crowdsensing technique, we request for
the data of VN because in smart vehicles there are approx-
imately all types of sensors. When some companies require
the relevant data, the company will generate a request through
blockchain, requested data is forwarded to the requester and
take advantage in sense of digital currency.
A. Our Contribution
In this paper, we proposed the model in which secure
provisioning of data and trust management of vehicles is under
consideration.
•We proposed the blockchain based framework to provide
the decentralized system in which every vehicle can
communicate to the other vehicle directly.
•Based on the proposed system in which we implement
the hashing technique to overcome the privacy issue.
•According to the proposed system model, we implement
the system in which is two types of data are counted that
are true or false. Vehicles that are existing in the least
group will be blocked.
•We introduce the concept of crowdsourcing in VN based
on blockchain. According to our knowledge crowdsourc-
ing in the VN based on blockchain is not proposed by
anybody.
•Finally, we conduct the simulation of our proposed sys-
tem model to present the validation of our system model.
II. RE LATE D WOR K
A. Vehicular Communication
In [11], there are ad hoc network for Dedicated Short
Range Communication (DSRC), wireless network and cellular
network those do not pay secure data communication among
the vehicles. There are some security protocols offered by
cellular networks; however, these protocols are not up to date.
Still, some companies are trying to develop these protocols
up to a more efficient level. Secondly, trust point of vehicles
is compromised because there are more chances of communi-
cated vehicles are exist outside to the area and these might be
malicious. To overcome this issue, we introduced blockchain,
use Proof of Work (PoW) mechanism for addition of new
vehicle, and introduce central trust point for vehicles from
which all vehicles are authenticated [12],[13]. Central trust
point gives an ID to the vehicle from which other vehicles
make trust and communicate safely. Vehicles that provide
some data get an incentive in terms of digital currency. It is
possible that one device has few sensors; however, this vehicle
required the data of relevant sensor, this data fetched from
other vehicles and pay incentive for this data.
In [6], Vehicles are able to communicate with another
vehicle directly there is an extra node that is backup of CN in
case of unavailability. Authors proposed an idea of distributed
architecture, which is capable of adding and maintaining many
devices in different areas. Scalability of the system is main
purpose of the article in which introduced three types of nodes
that overcome the connectivity issues in different divided
areas. CNs are RSUs these are fixed in some areas to cover
the specific area. While MNs are the type of vehicles that act
as the moving node of the chain while their priority level is
less. ONs are the vehicles that usually required connectivity
option, firstly ONs try to connect with nearby CN otherwise
connect with MN and communicate to the network. In future,
authors could work for the financing through blockchain and
introduce wearable technology that may convert the traditional
vehicles into the smart vehicles. In addition, we deals with area
optimization by introducing small cells.
In [4], some malicious vehicles are trying to attack the
centralized system and there are chances to be hijacked due
to single point of failure. The bottleneck attack is unable to
handle by a centralized system. Authors proposed blockchain
to the connectivity of vehicles in a distributed manner. Nodes
of blockchain are interconnected with each other by means
of hashes, and there is a public ledger that contains all
transactions of the network. To overcome these types of
different problems we introduce many blockchains to divide
the data and these chains interact with each other for collective
system’s work. This proposed model does not discuss the
sharing of traffic data among the vehicles. Authors do not
address the reliability of network in IoV so further will work
on these issues.
In [7], authors proposed that it is much difficult for a vehicle
to evaluate authentication of neighboring unfamiliar vehicles.
Besides, this issue is more sewer when neighboring vehicles
are malfunctioned or attackers. In centralized Data Manage-
ment System (SDMS), all vehicular data and ratings are routed
toward the centralized clouds because in some scenario data
is required to vehicle in a very short time interval. However,
this quality of service could not be achieved in all relevant
scenarios. Due to the high mobility of vehicles, network
topology will more dynamic and is relatively difficult to face
the entire vehicles in a given threshold. Authors introduced a
new Decentralized Trust Management Scheme (DTMS) that is
based on blockchain technology. In this scheme, RSUs are able
to connect in chain manner and exchange their information
relative to vehicular trust. It is now more consistent and
reliable because it is impossible to be hacked and able to
bane malicious nodes’ entry. There is a mechanism in which
applied PoW and Proof of Stack (PoS) mechanism to compete
with the other RSU to make the new node validation. In this
scenario, vehicles are communicated with other vehicles as
well, send information about traffic, etc., to the RSU to check
and validate, after this make useful information to part of
blockchain.
B. Network Communication
In [7], authors addressed the problem that there is huge
chunk of data for transmission in Device to Device (D2D)
based cellular network, but it is not able to transmit data. In
addition, it cannot support the huge number of D2D mobile
devices, so few mobile nodes have to disconnect for some
time. The main critical issue is, determining the access of
the users in D2D network for the authenticity of Channel
State Information (CSI). Typically with huge amount of CSI
value user are allotted a huge volume of bandwidth. Authors
addressed the issue that authentication of peers for the data
transmission as CSI is necessary because it transmits the big
data over the network so it should not be malicious. It also
consumes the whole bandwidth of the link.
In [13], it is necessary to maintain distributed system for
portable devices because centralized system does not allow
mobility. In addition, many other issues arise such as scala-
bility, security, privacy and consumption of bandwidth toward
the single server. Authors divided model into two parts using
blockchain. Core network part contains MN, which has high
computation power and large capacity of data storage. Where
these MN are responsible for adding the node by solving a
mathematical puzzle using PoW. These nodes are embedded
with a Software Defined Network (SDN) that provide high
agility, security and reduce the management cost. While,
core network part also contains edge nodes, which have less
computation power and storage capacity. It locally controls
the public network. If edge node fails then only local network
will down. Etherem environment is used for the smart contract
while the Argon2 hashing technique is used for encryption. In
future, it will improve the performance of the edge nodes by
enabling cache on the node to obtain the data easily.
In [14], authors proposed that blockchain stores the data of
all the nodes. So, data get increases linearly with respect to
number of nodes and time. Above mention, dynamics leads to
the following problem: 1) very short storage room at each node
in the blockchain. 2) unaffordable size of system which may
require high bandwidth for the new node to be synchronized.
This problem is called bloating problem. Authors solved
this bloating problem by using Network-Coding Distributed-
Storage (NC-DS).
In [9], authors proposed that in wireless sensor network
sensors data is fetched by some organization. However, data
retrieved by WSN is more costly due to deployment cost
in WSN, sensors are usually not portable. So the concept
of crowdsensing through mobile devices is introduced but
main issue in this approach is privacy protection and lack of
trust relationship. Due to above mention factor, many mobile
devices do not take part in sensing. To address this issue of
privacy we mainly focus on the location information of the
mobile node because mobile users are very conscious about the
location information and ID. Therefore, in proposed solution
authors encode every part of the users’ information and
combine them by the polarity of a common method to make
the full form of encoded users. An attribute of the blockchain
also plays positive role in this issue. Transaction between
nodes is encrypted and data is transparent to everyone due
to blockchain. Using blockchain, we can effectively protect
the users’ personals [15]. Android Studio and Eclipse are
the experimental platforms in which different simulations are
performed. In results the compare the majority of gender to
evaluate that which gender is more interested in proposed
idea of privacy protection. While in proposed model men
are more attracted and conscious about their privacy. Due to
incentive effect of the system. In algorithm time complexity
using Markel tree in blockchain is O(n) that is minimum
as compared to another method. In future work, authors
will expand the scope of proposed system in which they
will capable of gaining more sensors data by using different
attributes of a user.
III. PROB LE M STATEM EN T
In WSN, sensors’ data is fetched by some organizations
[9]. However, data retrieved by WSN is more costly due to
deployment cost in WSN, sensors are usually not portable. So,
the concept of crowdsensing through mobile devices is intro-
duced but the main issue in this approach is privacy protection
and lack of trust relationship. Due to the above mention factor,
many mobile devices do not take part in sensing. In [4], privacy
has very significance in the VN. Privacy requires protection
and trustful communication between vehicles.
In [7], when an event occurs on any location, there might
be malicious or faulty vehicles that send false data to the
controller and it will cause for even more risky. Therefore,
it is more considerable problem that is addressed.
Crowdsensing is the mechanism for the request based data
retrieval from the peer nodes of any network. The peer nodes
might be sensors in any sensor network [9]. The data can be
paid or unpaid as well. In [10], authors had not discussed the
crowdsourcing while it will be more useful and cost-effective
for wireless sensor networks. The cost will be minimized
greatly due to already deployed sensors in the vehicles.
In [6], authors proposed the mechanism of VN based on
blockchain in which concept of Minor Nodes (MNs), Ordinary
nodes (ONs) and Controller Nodes (CNs) is introduced. MNs
are directly connected with CN that is Road Side Unit (RSU)
and ON has an option, if it is in the range of RSU, get
connected. If no RSU is in a range of ONs then connect with
MN. Limitation of this system is if one ON exists in a less
dense area where neither CN nor MN.
IV. PROPOSED SYSTEM MODEL
A. Trustful Vehicular Communication
In this proposed model central trust point for vehicles from
which all vehicles are authenticated as shown in Fig. 1. This
trust point gives an ID to the vehicle from which other vehicles
make trust and communicate safely. For example, one vehicle
wants to communicate with other vehicles to evaluate the
data. Vehicles that provide some data to neighboring vehicles
requires the incentive in term of digital currency. It may be
possible that one device has less sensors or damaged sensors
while, it has required the data for some decision making. It
requested the data from other vehicles that are moving parallel.
Firstly, every vehicle gets registration with the blockchain and
gets Identification (ID) from the blockchain. Secondly, the
Vehicle authenticated by the CN using ID which is assigned
by the CN on registration stage.
B. Trust Management of Vehicles
In [4], authors proposed the idea about trust management of
vehicles in the blockchain network. Because there is always
Fig. 1: Trustful Vehicular Communication
some vehicles that misuse the information or provide false
information to the network as shown in Fig. 2. We introduce
the mechanism from which blockchain able to know the
credibility of information and notify the relevant vehicle for
three times. If the vehicles still send the false data then it
gets banned by the blockchain. Many vehicles send false
information to the network, due to their malicious behavior.
This problem also needs to address, that how can blockchain
evaluates which vehicle sending the false data. Let us say, that
data is requested from area A and in that area several vehicles
are moving when data is requested then each vehicle solves
the query and sends back the data to the controller. Now there
is ambiguity between the data collected from the vehicles that
which group of vehicles sending correct data. We count the
number of data with the same values and decided which group
of vehicles provides correct data.
Pn=Pi+Pj(1)
Where (Pi)is the number of vehicles sends the right data that
is considered in positive value and (Pj)contains the negative
value put total entries in (Pn). Now add these two values if
the calculated value is in the form of positive then it means
data is accurate and (Pj)will be accepted, while, if it is in
negative then (Pi)will be accepted and their data will be now
considered being accurate.
C. Optimization of Less Dense Area
VN maintains the network in which all vehicles must be
connected to the network. Some vehicles connect with CN,
other those are unable to be in the range of CN connect with
MN. MNs are high priority vehicles with large capacity of
computational power and storage, the vehicles always be in
the range of controller if provide connection to ONs. When
ONs want to connect with network with the help of MN then
MN acts like a bridge between CN and ON as shown in
Fig. 3. Moreover, if MN has the relevant data that is requested
by the ON, MN sends this data to the ON in response of
transaction [6]. However, there is a limitation in this concept
Fig. 2: Trust Management of Vehicles
that MN does not cover the wide area properly, so most of
the area covered by many CNs. While in the same area, there
are several regions with a less dense population of limited
vehicles. If we place CN to this region, it will be very costly
to deploy for such areas. Therefore, we place repeater with low
computation power and high signal strength. This repeater acts
like an MN. However, it is not mobile and covers the specified
area. If any vehicle wants to connect with the network from
less dense areas, it will entertain through repeater. Repeater
just passes the information from ON to the network. In any
city, there are many less dense areas, so we have covered these
all areas through multiple repeaters. Repeaters are cheap and
acting like nano and pico cells as used in a cellular network.
These are also for the region where base transceiver station
coverage is unable to reach.
D. Crowdsensing
Vehicles are enriched with the many sensors and actuators
for making the efficient VN that is able to decide condition of
the traffic, environment, and roads. Let us assume a company
A is working on weather forecasting but has constrained
resources to evaluate the actual and real time data. In a
conventional approach, they have to place sensors in different
areas for real time results for which company has to pay a high
cost. While, by applying our proposed idea of crowdsensing
through VN, they have to pay only for the data, no need
to deploy their own sensors network. Vehicles have many
sensors and they continuously send these sensors’ data to the
network. Crowdsensing companies are also part of blockchain
and generated a request to the CN for the specific regions’
data. CN already contains the data of that region or places real
Fig. 3: Optimization of Less Dense Area
time query to the vehicles of the concern region to provide the
relevant data.
V. DISCUSSION OF RE SU LTS
In this section we discussed Results of our scheme and
compare our proposed scheme with benchmark scheme. Ad-
ditionally, we have discussed the performance parameters of
proposed model and existing models in this section.
In the section, we represent the comparison of our proposed
model with benchmark models [6], [7], [9] in Table. 1.
A. Less Dense Area
We proposed an idea of less dense area optimization, where
we had covered the area where minimum vehicles exist. We
have introduced the concept of repeater, which is able to
connect the vehicle that is traveling in less dense areas. In
experiments, we had connected ten repeaters with a single
Controller node. In Fig. 1, four ONs can connect with the
repeater. However, we had install repeaters and every repeater
connects with CN. If a CN has the capacity to provide
services to eighteen ONs, with the first four repeaters, four
ONs connect with each while fifth repeater has only two
ONs’ capacity. After this, no ON was allowed to connect
because CN had reached its maximum capacity as shown
in Fig. 4. In Fig. 2, we had covered the less dense area
through multiple repeaters. If one repeater had covered the
2km then obviously this increase will go up linearly. However,
issue in this concept will cover the very low area relative to
an area covered by the controller. We have cover the area
efficiently while, system becomes more complex because there
are several small connection point and it is more complex to
install and troubleshoot. So, this considers the trade of the
system.
Fig. 4: Capacity of Repeaters
Fig. 5: Area Covered By Vehicle
B. Authenticity of Vehicular Data
In this section, we had solved the problem of ambiguity
(weather data is false or true) of the sensor’s data. Let us
say that if we have ten vehicles those send the data of a
sensor. Some of them sending the false value and some sending
the true value. Then the situation is quite ambiguous for the
other nodes, who require the data. To solve this issue we
had grouped up the vehicles on the bases of their data and
calculated the percentage of their sensors’ data. If one group
got 51% strength then this group is assumed to be true in
Fig. 6 it is shown. We did the experiments multiple time and
each of time results come accurately.
C. Crowdsensing of Vehicle
In Fig. 7, we had reduced the cost of sensors’ data by
the crowdsensing. Because there is an extra cost to deploy
Wireless Sensor Network and get data from already deployed
network also costly because this is also deployed for the
business by any company. We had used this concept in VN
so that we have not paid the high cost because vehicles used
these sensors for their use.
VI. CONCLUSION
In this paper, we proposed a scenario on the VN using
blockchain for some basic purposes that cannot be achieved
TABLE I: Comparison with Benchmark Scheme
Ref No. Trust Full Com-
munication
Trust Manage-
ment of Data
Crowd Sensing Area Optimiza-
tion
[6] XXXX
[7] X X X X
[9] XXXX
Proposed Model X X X X
Fig. 6: Authenticity of Vehicles Data
Fig. 7: Crowdsensing of vehicle
by using centralized technology such as trust management,
scalability and crowdsensing are the main three aspects that are
discussed in the paper. We used blockchain to make our net-
work purely decentralized. Decentralization is the key aspect
of the network that makes able the vehicle for crowdsensing
the data through a network and will be very cost effective
for the user. We proposed the concept for optimization of low
dense areas in which we discussed the area which has less
traffic but this is part of our smart city. We put repeater here
for availability to connect with blockchain instead of CN.
In future, we will work about proper transaction mechanism
and make trust management more convenient for the network
by doing a rating of the vehicles.
REFERENCES
[1] Ahmed, Syed Hassan, et al. ”Research on Efficient Data Forwarding in
Vehicular Networks.” Mobile Information Systems 2019 (2019).
[2] Jiang, Tigang, Hua Fang, and Honggang Wang. ”Blockchain-based
internet of vehicles: Distributed network architecture and performance
analysis.” IEEE Internet of Things Journal (2018), pp 1-10.
[3] S. Nakamoto, “Bitcoin: A Peer-to-Peer Electronic cash system,” Tech.
Rep., 2008.
[4] Jiang, Tigang, Hua Fang, and Honggang Wang. ”Blockchain-based
internet of vehicles: Distributed network architecture and performance
analysis.” IEEE Internet of Things Journal (2018).
[5] Ren, Yongjun, et al. ”Incentive mechanism of data storage based on
blockchain for wireless sensor networks.” Mobile Information Systems
2018 (2018).
[6] Sharma, Pradip Kumar, Seo Yeon Moon, and Jong Hyuk Park. ”Block-
VN: A distributed blockchain based vehicular network architecture in
smart City.” JIPS 13.1 (2017): 184-195.
[7] Yang, Zhe, et al. ”Blockchain-based decentralized trust management in
vehicular networks.” IEEE Internet of Things Journal 6.2 (2018): 1495-
1505.
[8] ”https://en.wikipedia.org/wiki/Crowdsensing” last visited 13 July, 2019.
[9] Jia, Bing, et al. ”A Blockchain-Based Location Privacy Protection In-
centive Mechanism in Crowd Sensing Networks.” Sensors 18.11 (2018):
3894.
[10] Xu, Yang, et al. ”Towards secure network computing services for
lightweight clients using blockchain.” Wireless Communications and
Mobile Computing 2018 (2018).
[11] Xie, Lixia, et al. ”Blockchain-Based Secure and Trustworthy Internet of
Things in SDN-Enabled 5G-VANETs.” IEEE Access 7 (2019): 56656-
56666.
[12] Lin, Di, and Yu Tang. ”Blockchain Consensus Based User Access
Strategies in D2D Networks for Data-Intensive Applications.” IEEE
Access 6 (2018): 72683-72690.
[13] Sharma, Pradip Kumar, and Jong Hyuk Park. ”Blockchain based hybrid
network architecture for the smart city.” Future Generation Computer
Systems 86 (2018): 650-655.
[14] Dai, Mingjun, et al. ”A low storage room requirement framework for
distributed ledger in blockchain.” IEEE Access 6 (2018): 22970-22975.
[15] Singh, Madhusudan, and Shiho Kim. ”Branch based blockchain technol-
ogy in intelligent vehicle.” Computer Networks 145 (2018): 219-231.
