With the ever-growing increase in the number of vehicles on road today, the security of the driver and the passengers has become more crucial and challenging than ever before in history. A secure and authenticated message transmission amongst the vehicles is the desired solution everyone is aiming at. The advancements in digital communication have given a new horizon to Vehicle-Vehicle communication. Nevertheless, authentication and privacy is still an open issue to be addressed in order to have a robust VANET. This survey compiles the contribution by the researchers for a Secure and Authenticate formulation of VANET. Towards the end, we enlist some open issues which can make a base for future research work.
VANET seems to be a promising solution to have secure transportation but it comes with a lot of challenges. There can be multiple types of attacks on various layers of networks which can be categorised in either Active or Passive ones. Most of the attacks are found at Network Layer which is the most crucial to keep the network secure[3], in this survey the authors discusses all the possible attacks in VANET. To say it in a crux form, VANET needs to be secured for privacy, authenticity and shall ensure secure message dissemination [5]against all the mentioned attacks [3].
There are many researchers who have come up with some potential protocols to provide a secure and authenticate VANET which are discussed and compared at length in this survey. The protocols have been clumped according to their core technique and have been elaborated according to their year of proposal. The main focus of this survey is to compile the significant contributions in VANET in the last decade despite of the technology so that it can facilitate the group of researchers who are working in the same direction.
This work is categorized in three major segments. In the first segment, the related surveys have been articulated for their contribution. In the second segment, we categorise the protocols according to their techniques and compare their progress and results. In the last segment we lay down the learning from all the existing protocols and list the potential concerns for future researches.
The research progress in V2V communication has been reviewed by many authors. This section summaries some of the significant researches being done in the last decade.
[5] discusses comprehensively about various privacy preserving and authentication techniques used in Vehicular Ad-Hoc Network. The survey compared the work done in the field of security & authentication based on ID-Based, RSU based Key Authentication, Cryptography based bilinear pairing and smart card based protocols. It also shed light upon the protocols that addressed the overhead in the PKI based methods. Not only that, the work also compared the protocols that use message aggregation & cooperative message authentication including batch verification. The survey concluded that three major areas need the robust solutions namely a) authentication of the messaged as well as the vehicles. b) Secure message dissemination and c) user privacy. [6] unfolds how the progress made in the V2V communication has not only given a modern approach to the Transportation but has given hackers the access to the life of the people travelling in the vehicle. Manifesting the point authors have looked upon all major attacks that have been made in the recent past in sensing, communication and control layers. The authors also presented an extensive study on all prospective attacks in V2X communications giving a contrast of their detection probability, properties and ease of attack. Towards the end paper unfolds various Machine Learning & Blockchain based solutions to deal the aforementioned attacks. [3] categorized the attacks on V2V communication on the basis of the Network Layer, being Active or Passive attack and type of attack. Along with the listing and elaborating the attacks authors talked about the proposed solutions which are given by the researchers for each of the attacks. Moreover, it is concluded that a protocol is required which can deal with all types of attacks and provide a mechanism where a trust system is build amongst the people in the Vehicular Network. [7] grouped the attacks in five security services namely Availability, Confidentiality, Authenticity, Integrity and Non-Repudiation and listed the attacks under them and analyzed the authentication schemes for each one of them. Authors also studied various Network & Mobility Simulators. [8] presented the discussion on various algorithms using Machine Learning as a major contributor in safe and authenticated message transmission in V2V. Authors also discussed that there is a need of an algorithm which not only protect from DoS attacks but covers the damage from other attacks as well.
[9] present survey on the researches being done for wireless communication in Vehicles related to their applications in real life, their corresponding protocols and multiple security issues. Authors have analyzed all the researches for their strengths and weaknesses for future work perspectives. [10] gave a contrast of various security issues over the VANET and listed the proposed solutions to handle those issues. The analysis also mentions some motivation for the future work in the same direction. [11] focused how urbanization is leading to increase traffic in the cities and consequently the need of secure data transmission is highlighted. Moreover the authors logged a range of researches talking about inter transmission o f the packets in multiple models of communication. Routing scalability and reliability can be assured using clustering says [12]. The authors made a taxonomy of the clustering techniques used for selection of cluster head, affiliation and overall management. Not only the benefits but the shortcomings of the algorithms have also been analyzed. The requirements of modern cars and the security issues and proposed solutions are listed by [13]. Whereas, the importance of Encryption and authentication is highlighted by [14] in their survey where the a comparative study is presented for the existing algorithm for VANET. Working on VANET requires the know-how of various Simulation tools which is studied and analyzed by [15] in their paper. They also gave contrast between various attacks and mechanisms used to deal with them. [16] Emphasizes on the authentication for VANET and therefore did a survey on various schemes available for ensuring the authentication. [17]not only emphasized upon the existing security threats but explores on the possible attacks in future which provides a motivation to the researchers to explore and work on their possible solutions. Countermeasures for priori and prosteriori are discussed and analyzed by [18] and a sketch had been proposed for the same alongside the limitations and tensions in VANET then were elaborated and compared.
This section elaborates various researches being done over a decade for secure Vehicle to Vehicle communication. We have categorized the algorithms in specific categories whereas some of the algorithms are common and fall in more than one category.
There are five different categorisation of VANET routing algorithm on the basis of the area and/or application. They are, Geo cast routing protocol, Cluster based routing protocol, Broadcast routing protocol, Topology based routing protocol and Position based routing protocol. [19] Over the years the researches have proposed various models choosing a specific or a hybrid approach to have an efficient and secure model of VANET. Out of all these categories, clustering based algorithms have been explored the most and given some significant contribution in VANET. In this section, various clustering routing algorithms have been compared and analyzed.
[20] proposed a location based routing protocol with cluster based flooding for VANET. The performance of this algorithm has considered a positional routing protocol called LORA_CBF and also with two non-positional routing algorithms AODV[21] and DSR [22] . The results showed that the performance of positional based is better than the non-positional ones. [23] came up with a clustering based algorithm with delay tolerance. It worked on route discovery for data dissemination following the clustering techniques. It performed better in terms of cost, overheads , stability of the clusters, packet loss ration and end-to-end delay when compared with [24]&[25].
In [23] clusters are formed on the basis of the average speed of the vehicles whereas [26] grouped the vehicles with long travel time and less deviation in speed is selected as Cluster Head and/or Improved cluster head. Due to the same [26] proved showed lower end-to-end delay as compared to CBLR[27].
[28]shared an improved Genetic algorithm-based route optimization technique (IGAROT) which used KMeans Clustering to remove road anomalies. When the algorithm was compared to GA, IGAROT showed 42.0% , 75.7% and 4.24% better results in high, medium and low car density environment respectively.
It is ideal for the dynamically changing topologies, large networks & high mobility.
It costs less than [24] & [25] due to the minimum overhead Prompt notifications of the agencies involved in road maintenance.
Efficient and distributed Improved cluster head 3.2. Fog Computing Based Algorithms for Security
[29]Discussed about various perspectives of data transmission in Vehicles. The paper analyzed various available models for data transmission like 3G and 4G cellular networks, RSUs, Mobile Cloud Computing etc and also analyzed their shortcomings in terms of expense and availability. They proposed to utilize the idle parked and moving vehicles as both computational and communication vehicles. The huge potential of VFC is taken forward to make V2V communication better. [30]Has worked upon reducing the hand off time and increasing the throughput for the vehicular data transmission in Fog computing. The research proposes a solution called as CVFH – A cross Layered and Neighbouring vehicle aided fast handoff wherein they focussed upon two major strategies to achieve their target of reducing the hand off and increasing the throughput during the high speed of Vehicles and High rate of packet transfer. CVFH suggests that in order to decrease hand off delay the Vehicle can choose the Access Point used by the neighbouring Vehicle. Secondly CVFH suggests that the Vehicle shall carry on with the existing Access Point by the time a new connection is established. The simulations have shown CVFH to be a better performer as compared to IEEE 802.11 [31]Proposed a data dissemination model to address the issues involved in Fog Computing while transferring High Volume and non urgency based data transmission. The research also focuses on the fact that the fog model becomes expensive and how to utilize the fog Based Computing efficiently. Therefore they proposed Delay Tolerable Network(DTN) approach . In this approach they assign different role to Cloud and Fog Severs. Here, the Cloud is given the job to manage data flows & content queries where as the Fog Servers are dedicated to disseminate data using Delay Tolerant Network. The research results showed that their model depicts a higher delivery success ratio and less delay to ensure a better system to disseminate important data in Fog Computing. [32]Initiated to work on a model that creates a layer in between the Vehicle and the server to speed up the data transfer. Just like when the client send a request to the server and cache memory maintains the information for better processing the next time. They have proposed a model called SIVNFC which prevents DoS Attacks across the network. [33]
Have done an analysis to check the data transmission on 28GHz. In order to overcome the problems to high level of interference and High Packet Error they have done a deep interference analysis to derive the right expressions of Packet Error Probability and Ergodic Capacity. The same has been simulated extensively to be proven to next revolutionary step in V2V Communication.
[34] Introduced an intermediary fog layer to improve latency in Vehicle to Vehicle communication. The algorithm works on the efficiency in terms of improving the delay in the messages. To ensure the authentication it used Long Term & Short Term Pseudonym/Keys. [35]integrated hybrid optimization techniques for authentication of the network. It integrated Cuckoo Search Algorithm, Firefly Neural Network, Firefly Algorithm and Key Distribution Establishment in the protocol to ensure the network is prevented from DoS and SNI Attacks. The results showed improved throughput and less jitter. [36]extended the research and introduced the implementation of Artificial Bee Colony (ABC) and Genetic Algorithms to detect real time DoS attacks in IEEE 802.11p. Algorithm is also capable to classify between genuine and attacked vehicles with the help of Feed forward back propagation neural network (FFBPNN). The algorithm proved to give 92% accuracy and reduced the jitter by 72%.
To ensure the safe communication in VANET, trusted authority (TA) comes into the picture. Self
authentication amongst the RSU and Vehicles can diminish the burden of TA. [37] proposed a fog based model
for anonymous authentication to ensure real time communication and reduced pressure on TA using pseudonym
based tracking mechanism to update and track. The algorithm gave promising results in terms of privacy
protection & authentication. Where[37] talked about anonymous authentication [38] proposed an algorithm for
mutual authentication and generates a secret key for secure communication between two vehicles. It ensures
privacy protection and claims practicality for the implementation redu
introduce the concept of Fog Computing and used to the concept of LTP/LTK & STP/STK A Fog integrated VANET Scheme compared with Firefly and Cuckoo VFC with Hybrid Optimization Algorithm like Cuckoo, Firefly & ABC Fog Computing & d pseudonym-based batch anonymous authentication Authenticate Key Agreement without bilinear pairing
key transfer is reduced due to the Fog Node.
Provides real time detection of HDSA in IEEE802.11p
The need of VANET came into the picture when the number of vehicles increased due to urbanization and affordability of vehicles. Initially horns, hand waves and other mediums of traffic controls were sufficient and gradually the traffic police came into the picture. Now, with the advancement of technology & digitization the call for VANET has tremendously increased so as to improvise the transportation system and ensure smooth and safe travel. However, technology always comes with a threat of security & authentication. In order to authenticate the sender of the information various protocols have been introduced. The protocols can be categorized into two heads, individual authentication & batch verification. With the increased number of vehicles, one-to-one authentication becomes impossible & inefficient and therefore the concept of Batch Scheme of authentication proved itself to be more effective [39]. [39] have integrated Bilinear pairing to authenticate the signatures in batch and performance proved the algorithm to be performing better than [40] and is also capable to handle replay and non-repudiation. However, it still lacks when it comes to identifying illegal signatures. To overcome the shortcomings of [39], [41] came up with another algorithm after analyzing the work of [41] and arriving at a conclusion that it lacks the capability to handle impersonation attack. The algorithm is based upon the four step process Key Generation & pre-distribution, pseudo identity generation, message signing and message verification. Bilinear pairing has been recognized as one of the most complex operation when we talk of cryptography in modern times [42]. To provide a better an simpler mechanism a set of researchers have come up with CPAA based schemes without integrating Bilinear pairing. [42] Proposed a scheme which supports privacy protection and mutual authentication at the same time. The proposed scheme could be used for V2I and V2V both. Whereas [43] continued the work on CPAA and integrated One way Hash function and Elliptic Curve Cryptography provide a scheme with lesser cost of computation and communication. To reduce the cost even more [44] suggested to opt for aggregate signature instead of individual signature verification. They integrated Elliptic Curve Cryptosystem. The research results show that the signature verification time has improved from 94.3% to 92.7%.
[39] Batch Scheme based on Bilinear
Compared with [40] and performed better in terms of efficiency
2018 [44]
Conditional Privacy-Preserving Authentication (CPPA) without bilinear pairing DiffieHellman is used to ensure security & Discrete Logarithmic encryption scheme is used for Cryptography.
ID Based Signature Scheme without bilinear paring IBS-CPAA added to Elliptic Curve Cryptography & general one way Hash Function Certificate less aggregate signature without pairings
In the last decade Smart Card Based authentication has been explored more than ever before. They have an incredible capacity to hoard enormous amount of data with the help o f embedded circuit chip and memory chip. Smart cards offer a strong security base and have been used in many authentication algorithms, encryption techniques and asymmetric key services. [46]
Researchers have explored Smart Cards authentication capabilities for VANET as well. [47]’s PAAVE generates on-the-fly anonymous keys amongst RSU and the vehicles. Their results show improved efficiency, Communication Overhead and Computational time. [48] continued the work to provide an even more efficient way which not only diminishes the cost but also withstands attacks like offline password guessing attack, smart card loss attack, impersonation attack and some more. [49] proposed Anonymous & light weight scheme for the authentication of User and the message at not only authentication stage but also at password change and data transmission phases as well. The protocol can deal with offline password guessing attack and impersonation attack and also works on Packet loss ratio & latency. All these protocols used Smart Card for anonymous authentication whereas [50] came up with an authentication protocol which uses real identity of the drivers. It uses Spanish eICard and ensured 60-70% correctly obtained identities.
PAAVEAnonymous authentication using Smart Cards Efficient Authentication Protocol Ensures anonymity by dynamically generated login ids Anonymous & light weight scheme for the authentication of User and the message.
Diffie-Helman protocol; smart cards; hash functions; centralized trusted authority for loading smart-cards with login id and password Obtains the real identities of the drivers through t he Spanish eICards
anonymous keys on the fly for vehicle s and RSUs. Vehicle has to store one cryptographic key Can resist offline password guessing attack, smart card loss attack, impersonation attack, DOS Attack.
Provide anonymity preservation Can deal with as offline password guessing attack and impersonation attack
results shown on the real map during the simulation
2010 [47] 2014 [48] 2017 [49] 2016 [50]
Researchers have been working intensely on improving the security mechanism in V2V communication and have proposed various techniques as discussed in this survey. By comparing the work accomplished till date some open issues have been discovered which can be used as a potential input for the future researchers. Firstly, all the cluster based algorithms have been proved to be better than other techniques but the researchers must keep on comparing their work within the domain with other cluster based algorithms. Also, it is assumed that the real time environment has the RSUs and other infrastructure in place which is not the reality. The design complexities while implementation should not be ignored. Apart from Firefly and Cuckoo, spline and polynomial fit could be used for better comparison and improved results. There are very less impressions of researches using IoT in this domain, so that’s another prospering field. Implementation of Elliptic Curve cryptography in future to stimulate the identification process to work on location based mechanism is yet another unexplored area. These are few areas where the future researches can be directed and help in achieving the real time solutions.
In this survey we discussed about various algorithms focusing on Secure and Authenticated algorithm under four categories namely Clustering Based, Fog Based, ID Based and Smart Card Based. Towards the end we shared some uncovered topics for future work with a hope that these issues are addressed by the researchers and a potential solution is proposed.