1 Introduction

Data aggregation is an important paradigm in wireless sensor networks (WSNs) thatcombines the data arriving from different nodes in the network. The data aggregation in such networks aims to reduce redundancy and minimizes the total transmissions to save the energy in the network. Hence, the communication at different levels are reduced with data aggregation and this reduces the consumption of power in the network. Reducing the consumption of energy and the data redundancy is always an issues in WSN with data aggregation protocol [1].

Such data collection and storage in WSNs raises the issues related to user privacy. The collected data consumed from the sensor nodes are monitored at regular intervals. This can lead to privacy threat of the user inside the VANET architecture and it is limited to their personal behavior, message transmitted, vehicle physical information and its source/destination regions. The main problem is due to the activity of the 3rd party service providers, who manages the data collected inside the VANET [2].

Since, many researchers’ aims to reduce the communication cost by scraping off the entire security. This leads to devastating results due to serious attacks and security breaches in VANETs by the attackers. This leads to consideration of both security and efficiency while designing the security protocol. The consideration of vehicle density, mobility and constrained paths can lead to improved security in VANETs with improved efficiency. Further, encryption is seen as interesting method, which preserves the privacy of user data, however, the secret encryption keys of the user is revealed to some third parties for processing. Hence, the reliability of information leads to matter of insecurity.

The researches on VANETs has improved its fundamentals by the addition of security parameters in its deployed architecture. This is revealed in many conventional works. However, the addition of security parameters has reduced the scalability of the VANETs, since overhead is a reasonable criticism. The most related architecture that serves for VANET security is the PKI-supported asymmetric authentication with anonymity. Here, each message is signed to be authenticated by the receiver. Though, the security is improved with increased efficiency in asymmetric algorithms, however, the algorithms like ECC (Elliptic Curve Cryptography) suffers from overhead [3]. This leads to a question, whether the security in VANET is reliable or not and this paper attempts to improve the security in VANET using improved Fully Homomorphic Encryption (FHE).

Due to such privacy concern, Homomorphic Encryption (HE) is an encryption model that makes the user to preserve his data with good confidentiality. Also, it helps to improve the user privacy over processing of information by the untrusted third parties. Hence, the HE is been considered as a substantial approach to maintain the privacy and confidentiality in many real world applications. In these applications, the HE uses new approaches torun on encrypted input without the information of primary data. This guarantees the privacy of user from the third untrusted parties [4].

In this work, a data aggregation module is proposed to improve the security with reduced computational overhead in MANETs. This paper forms a secure data aggregation protocol using Fully Homomorphic Encryption (FHE). The main aim is to improve the security features of users in VANETs with reduced overhead. Instead of signature encryption algorithms, the proposed method concentrates on FHE to maintain the user privacy. This accounts for increased security for the vehicles inside the VANET and overhead is reduced in such network with reduced computation. The used in different vehicles are not shared between them, however, pseudonym is shared in terms of its speed and location is shared. The proposed method hence adopts full homomorphic encryption that realizes three important properties that include: confidentiality of nodes, message integrity, and the computationalcorrectness. We have introduced the FHE scheme in VANETs to improve the security based on dynamic nodes.

The contributions of the paper is summarized as follows:

  • This method avoids the leakage of distance estimation by the other nodes in the network and avoid nodes being attacked by the malicious users. Hence, a secure data aggregation protocol called Fully Homomorphic Encryption is used to improve the user privacy in VANETs.

  • The proposed system manages the encrypted data and provides fully flexible access control on the cipher text with two level decryption and achieves well the re-encryption with homomorphism. Thus the authorized users can only access the information in more secured way.

  • Pseudonym invisible re-encryption (including initial and final encryption) is used in Fully Homomorphic Encryption to improve the encryption, distanceverification and node location.It is also used to protect the privacy of nodes in VANETs.

The outline of the paper is mentioned as follows: The section 2 studies the conventional researches used in data aggregation. The section 3 provides the FHE and section 4 provides FHE to improve the data aggregation in WSNs. Section 5 evaluates the proposed method with various metrics and discusses its effectiveness with other methods. Section 6 concludes the paper with future work.

2 Related works

Data aggregation techniques is been used widely to reduce the consumption of energy in WSNs. Data aggregation in wireless sensor networks uses both cluster and non-cluster based approaches and that include:

The cluster data aggregation is generally used to reduce the energy consumption in the network by splitting the entire network to form a set of nodes. This forms a stable network architecture. The similarity based data aggregation function involves the use of clustering method that pairs the objects based on similarity level. This method specifically used to reduce the data latency in the network. The mobility based technique changes its architecture in clusters based on the mobility of the nodes in the network. The dynamicity in such network concentrates on loss in packets associated with its mobility. The distance based technique involves the use of distance estimation approaches to find the shortest path between the source and sink. This method concentrates mainly on collision avoidance and energy consumption.

The non-cluster data aggregation technique fits with static or dynamic WSNs, where the messages are transmitted either directly or indirectly in a randomly deployed architecture. This includes mobile sink and relay based technique. The mobile sinktechniques uses a non-stationary node collector to collect the data from nodes on-the-fly. This concentrates on the network connectivity with enhanced network coverage. The relay basedtechnique uses factors like distance and energy to transmit the packets in the form of relays. This method helps in reducing the computational complexity, since, it avoids the widespread of packets in the network [5].The above techniques in data aggregation suitably works on improving the efficiency of the network, however, security concerns are not addressed well in the above techniques.

The homomorphic encryption technique is used widely in WSNs, however, method relating to VANETs is quite low. Only, Song et al. (2015) [6] proposed FHE with multiparty security scheme to avoid the distance based estimation of the vehicle in the VANETs. Other methods relating to FHE in VANET is unavailable. However, there are several HE encryption schemes:

Gentry’s fully homomorphic scheme [7] uses probabilistic decryption algorithm and can be adopted over alow multiplicative degree algebraic circuit. This leads to a bit complexity of Õ(λ3.5) per binary addition/multiplication gate with λ as security parameter. Coron et al. (2011) [8] proposed FHE over integers and reduced the public key size to Õ(λ7) and later Coron et al. (2012) [9] reduced the public key size to Õ(λ5).Gentry’s fully homomorphic scheme with bootstrap functionality is proposed by Gentry &Halevi (2011) [10] to reduce the asymptotic complexity from Õ(n2.5) to Õ(n1.5) candidates over n dimensions. Gentry et al. (2012) [11] used FHE to reduce the overload and decrypted the message with size Õ(λ) and depth polylog(λ). Chen and Nguyen, (2012) [12] proposed simple FHE and provided challenges over integers to reduce the complexity. This algorithm is used for practical purpose and similar to this aspect, Fan & Vercauteren (2012) [13] used Brakerski’s FHE with Learning With Errors (LWE) problem to the ring-LWE setting, which is used for practical setting. Gentry et al. (2012) [14] used pure FHE with better bootstrapping technique that reduces the 1GB file to 24 MB. Fau et al. (2013) [15] executed FHE on BGV-style cryptosystems with classical integer manipulation operators like arithmetic, logical, bitshift, comparison, etc. The FHE scheme without Gentry’s bootstrapping procedure is proposed by Brakerski et al. (2014) [16] reduced the computational complexity of Õ(λL3) without quasi-polynomial factors for non-boot strapping and uses quasi-polynomial factors for bootstrapping with computational complexity of Õ(λ2). Brakerski &Vaikuntanathan, (2014) [17], proposed Leveled FHE scheme based on short vector problems with a computational complexity of k · polylog(k) + log|DB| bits. Boneh et al. (2014) [18] proposed Fully Key-Homomorphic Encryption using the public-key homomorphism property. Wang et al. (2015) [19] used FHE with Fast Fourier Transform with reduced time for encryption, decryption and re-encryption. Recent technique on FHE include threshold FHE with monotone boolean formula and threshold access structure [20], CCA1-secure FHE scheme with two multi-key identity FHE: LWE and sub-exponential in-distinguishability obfuscation [21], FHE with sorting over encrypted data without deciphering [22] and algebraic FFE with multivariate polynomials [23].

3 Design goals

The basic scheme is to summarize the security in messages sent over VANET network. To make an application to be safety, simplicity and robustness is an important tool to attain it. However, overhead generated during message transmission, leaves room for improving the architecture. It is in fact that there exist a tradeoff between the security and efficiency in the network, where either of the one is improved at the expense of other. The design goals are chosen such that the tradeoff between the security and efficiency is avoided by exploiting the VANET properties with finite solutions. The main focus of the paper is to keep the security and efficiency at equal levels using fully homomorphic message encryption.

FHE aims at providing supportive computation over the data encrypted and ensure better data confidentiality [24]. The main design goals include: mutual authentication, conditional privacy-preservation, data confidentiality and the computational correctness.

Mutual authentication

A reliable trusted communication is needed to be established between the users based on mutual authentication. In VANETs, if the vehicles need to communicate with other vehicles, necessary evidences should be sent as identities while authenticating exist between the users. If the vehicles are verified, the verified user has the ability to communicate with other trusted vehicles. This can reduce the risk of a vehicle being deceived by the other interrupts in the VANET network.

Conditional privacy-preservation

The conditional privacy-preservation is regarded as a single entity that combines traceability and anonymity of the vehicle. Since, the vehicle in the VANET network possess close associated with its identity, distance, location, etc. Hence, it is necessary for the network to prevent these secret data being revealed with the other vehicles in the network. These security requirements are achieved by the usage of pseudonyms than its real identity, while communicating with other vehicles. Additionally, the source of malicious behavior is traced to avoid the false information being spread in the network regarding the traffic information services.

Data confidentiality

In VANET, since the communication established between the vehicles is broadcast in nature, the overhead by the interrupters can occur. Hence, encryption of the message from the vehicles is required and this can avoid the leakage of real information from the vehicle.

Computational correctness

In VANETs, a privacy preservation distancemechanism has to be used for forwarding the packets using selfish vehicle nodes during relaying operation. Also, this should motivate fairly the active participants in the network and suppress the activity of the malicious nodes using rewarding/punishment mechanism. Further, the fairness is ensured by estimating the correctness of the values of reputation or credit. The vehicle node, which is active and plays a cooperative role with other nodes in transferring the message, increases the computational efficiency. Also, the detection and removal of malicious nodes should be done effectively.

4 Security protocol design

The VANET has several entities that include: infrastructure nodes (in), vehicle unit (vu) and road side unit (rsu). The in has authorities, service providers and vehicle manufacturers, vuand rsuhas encryption device, communication or on-board device (obu), sensors and storage device.

Here, the vehicles are connected to the static nodes or rsu for exchanging the information between them or with the in. Using this entity, two different context, namely adhoc and infrastructure is identified. The entries that manages the traffic and provides external support lie in infrastructure context. The other entity includes service providers, vehicle manufacturers also belong to this context. Based on the trust with third parties, the entities may be considered as fully reliable for other operations. To avoid sporadic communication between the vehicles in adhoc environments, obu is used. It also has set of sensor made inbuilt in the vehicle for measuring the changes in its target location and monitoring the own status. The processing of sensor output is done by computational devices and the data is stored in storage unit inside the vehicle.

The other computational tools include: VANET service or data service (ds) that provide interconnection of nodes in the network for storing the data and other computations. Access Server (as) to compute the secure data between the vehicles in the VANETs. Data Encrypts (de) for collecting the pseudonym and encrypting it and storing it inas. The de is used for further computation and analysis. Users or vehicle (v) process the encrypted data for further analysis (Fig. 1).

Fig. 1
figure 1

VANET architecture model

Full size image

The main aim of the proposed VANET model is to provide security to the messages which is been transmitted with/without infrastructure deployment. The information includes: speed and current location.

In the proposed method, to improve the privacy preservation in the encryption and decryption process, a self-generated pseudonym is allowed to generate during the process of authentication. Since, the regional trusted authorities broadcasts periodically the public key through rsu during the generation of pseudonym. This is generally done when the vehicle is intended to use the pseudonym generation or generate a new one or update it. Thus, the self-generated pseudonym of a vehicle is given by,

$$ Pn= Time\left|\left|p(id)\right|\right| hr\Big\Vert rsu $$

where, p(id) is the encrypted value of real vehicle’s identification. This is encrypted using a current public key and hr is the vehicle code on its home location.

In the proposed method, the rsu broadcast the information in a periodical manner and the operations of road safety unit is a tamper-proof, which is performed in a trust based manner. The proposed method is intended to operate in an adaptive way, when the vehicle is authenticating itself as a new node or updating itself as a present pseudonym.

  1. Step 1:

    Vehicle-to-vehicle authentication, where the roadside unit broadcast the information about it.

  2. Step 2:

    The vehicle within the range of home location authenticates with the roadside unit.

  3. Step 3:

    The roadside unit broadcasts the self-generated pseudonym.

  4. Step 4:

    The authentication between vehicle to vehicle is carried out within its transmission range.

  5. Step 5:

    Vehicle is allowed to authenticate with the server or it is called as cross vehicle to vehicle authentication.

  6. Step 6:

    The server queries the road side unit for authentication of vehicles.

  7. Step 7:

    The reply from roadside unit is sent to server and then the vehicle is authenticated inside the network (Table 1).

Table 1 Preliminary Notations
Full size table

5 Preliminaries and notations

5.1 Additive homomorphic encryption

The additive homomorphic encryption follows Paillier cryptosystem [23] that has encrypted the data with similar keys (p), which is presented as: [mi]p(i = 1,2,...,N). Hence, the additive homomorphic encryption follows:

$$ {D}_s\left({\Pi}_{i=1}^N{\left[{m}_i\right]}_p\right)=\sum \limits_{i=1}^N{m}_i $$
(1)

where,

Dsis the homomorphic decryption and ‘s’ is thesecret key.

5.1.1 Encryption and decryption

Consider, two large prime numbers r and t, and then n is the mul(r,t). Take as the cyclic group element with modulo in terms of quadratic residuesn2 and the element has two elements of maximal order g and h. Here, h can be computed using gxmodn2.

where, x∈ [1,λ(n2)] and λ() is the Euler function with co-prime factor x with order possess higher probability. Hence, there exist a maximal order for value h.

5.1.2 Key generation

Here, n, g and h is the public parameters that has a random secret value, x.the value of h is gx mod n2 and x∈ [1,order()].

5.1.3 Encryption (En)

The generation of cipher text from the message m∈\( {\mathrm{Z}}_n^{\ast } \)and r is given as:

$$ {\left[m\right]}_h=\left(T,T\right)=\left\{{h}^r\left(1+{m}^{\ast }n\right),{g}^r\right\}\ {modn}^2 $$
(2)

5.1.4 Decryption (Dn)

If the values of m and x is known, then the message is decrypted as:

$$ m=L\left(\frac{T}{{\left({T}^{\hbox{'}}\right)}^x}\operatorname{mod}{n}^2\right) $$
(3)

where, \( L(u)=\frac{u-1}{n} \)

6 FHE scheme with re-encryption

The proposed method uses pseudonym-invisible re-encryption to attain FHE in vanets. The section discusses the algorithm used in proposed method.

6.1 Key generation

The key generation to secure the message uses two parameter: security (k) and prime numbers of large values, which is two in number: p, q. and hence L(p) = k = L(q).To exhibit the secure prime property, two primes are used: p′, q′ and this satisfies the condition: p = 1 + 2p′ and q = 1 + 2q′. Further, n = pq is computed and a generator (g) is used that possess maximal order. The key pair is thus generated with ds and as:

$$ \left({s}_{ds}=a,{p}_{ds}={g}^a\right)\ and\ \left({s}_{as}=a,{p}_{as}={g}^b\right) $$
(4)

The Diffie-Hellman key is then negotiated using,

$$ P={p}_{ds}^{sk_{as}}={p}_{as}^{sk_{ds}}={g}^{a\ast b}\operatorname{mod}{n}^2 $$
(5)

The data is encrypted by making P as public over all vehicles and it is published using as to all the vehicles within the coverage area. The key pair is thus generated by the user, (si,pi), which is modified as (ki,\( {g}^{k_i} \)) and the public parameters include:{g,n,P}.

The En and Dn algorithm is applied with the key pairs: (si,pi) = (ki,\( {g}^{k_i} \)) and this is been supported by the 2-level decryption over encrypted data.

6.2 Two key encryption (2-En)

The cipher text is processing by encrypting the original message with key from the server P, rather than acquiring the key from pi. The message mi∈Ζnfrom a vehicle uses random number r for encrypting the message using P, where r ∈ [1, 0.25n]. The message [mi] is chosen instead if mi, this makes the decryption to be done under the proper coordination between ds and as. The generation of cipher text occurs by:

$$ \left[{m}_i\right]={\left[{m}_i\right]}_P=\left\{T,{T}_i^{\hbox{'}}\right\} $$
(6)

Where,Ti is calculated as: (1 + min)Prmodn2 and the value of \( {T}_i^{\hbox{'}} \)is calculated as: grmodn2.

6.3 Two key decryption (2-Dn)

Initially, the decryption over [mi] is carried out by ds and [mi] is converted to another cipher text for further decryption, the initial decryption is carried out as:

$$ {\displaystyle \begin{array}{l}{\left[{m}_i\right]}_{p_{ac}}=\left\{{T}_i^{(1)},{T}_i^{\hbox{'}(1)}\right\}\\ {}\kern2.00em =\left\{{T}_i,{\left({T}_i^{\hbox{'}}\right)}^{s_{ds}}\right\}\\ {}\begin{array}{cccc}& & & \begin{array}{cc}& =\end{array}\end{array}\left\{\left(1+{m}_in\right){P}^r{g}^{ra}\right\}\operatorname{mod}{n}^2\\ {}\begin{array}{cccc}& & & \begin{array}{cc}& =\end{array}\end{array}\left\{{\left(1+{m}_in\right)}_{p_{ac}^{ar}}{P}^r{g}^{ra}\right\}\operatorname{mod}{n}^2\end{array}} $$
(7)

Upon receiving the encrypted data from Eq.(7), here, direct decryption over ac with secret key is given as:

$$ {\displaystyle \begin{array}{l}{T}_i^{\hbox{'}(2)}={\left({T}_i^{\hbox{'}(1)}\right)}^{s_{ac}}\\ {}\kern1.00em ={g}^{rab}\kern0.5em \\ {}\begin{array}{cccc}& & ={P}^r\operatorname{mod}{n}^2& \end{array}\end{array}} $$
(8)
$$ {m}_i=L\left(\frac{T_i^{(1)}}{T_i^{\hbox{'}(2)}}\operatorname{mod}{n}^2\right) $$
(9)

The pseudonym invisible re-encryption is proposed using re-encryption method, which is different from the scheme discussed above. Here, the encrypted message under public key is converted to cipher text and u with key pairs \( \left({s}_j,{p}_j\right)=\left({k}_j,{g}^{k_j}\operatorname{mod}{n}^2\right) \)acquire mi from [mi].

6.4 Initial re-encryption (IREn)

The direct decryption is prevented by ac and the ds initiates the initial phase re-encryption, which is given as:

  • Select cid

  • Then compute \( {h}_1=H\left({\left({p}_j\right)}^{s_{ds}}\left|\left| cid\right.\right.\right) \)

  • \( {\left[{m}_i\right]}^{+}=\left\{\overline{T},{\overline{T}}^{\prime}\right\}=\left\{{T}_i,{\left({T}_i^{\hbox{'}}\right)}^{s_{ds}}{g}^{h_1}\right\} \)

6.5 Final re-encryption (FREn)

Once the data packets are received [mi]+, the ac decrypts the received [mi]+:

  • (Compute) \( {h}_2=H\left({\left({p}_j\right)}^{s_{ac}}\left|\left| cid\right.\right.\right) \)

  • \( {\left[{m}_i\right]}_{p_j}=\left\{\overline{T},{\overline{T}}^{\prime}\right\}=\left\{\overset{\frown }{T},{\left({\overset{\frown }{T}}^{\prime}\right)}^{s_{ac}}{g}^{h_2}\right\} \)

6.6 Decryption (RDn):

The u is used to decrypt the encrypted message\( {\left[{m}_i\right]}_{p_j} \)and using cid \( {h}_1^{\hbox{'}} \)and\( {h}_2^{\hbox{'}} \)is computed as:

$$ {\displaystyle \begin{array}{l}{h}_1^{\hbox{'}}=H\left(\left({\left({p}_{ds}\right)}^{s_j}\left|\left| cid\right.\right.\right)\right)\\ {}\kern0.5em =H\left(\left({g}^{a\times {s}_j}\left|\left| cid\right.\right.\right)\right)\\ {}\begin{array}{cc}& ={h}_1\end{array}\end{array}} $$
$$ {\displaystyle \begin{array}{l}{h}_2^{\hbox{'}}=H\left(\left({\left({p}_{ac}\right)}^{s_j}\left|\left| cid\right.\right.\right)\right)\\ {}\kern0.5em =H\left(\left({g}^{b\times {s}_{k_j}}\left|\left| cid\right.\right.\right)\right)\\ {}\begin{array}{cc}& ={h}_2\end{array}\end{array}} $$

The results are achieved based on interaction between ac and ds through addition and removal of masks from the cipher text. The proposed FHE avoids such interaction. This avoids the impersonation attack and avoids collision of data between the servers with u.

7 Evaluation and discussions

The performance of proposed method for secure authentication of message during aggregation is measured. Here, the computational overhead is a prime concern associated with generation of key to secure the message transmitted between the vehicles. Initially, the influence of the secret key and then the influence of n is estimated.

In this experiment, the operation time of the proposed method is measured in terms of En time, Dn time, 2-En, 2-Dn, IREn, IREn and RDn. The Enc and Dec defines the single layer encryption, 2-Enc and 2-Dec are two layered encryption, IREn and RDn represents the four exponentiations and two hash computations, and FREn represents one hash and three exponentiations computation.

7.1 Influence of secret key

Here, the influence of secret key length is estimated for the proposed re-encryption scheme in FHE method. The secret key bit length is set in the interval between [100,800] in the addition of 100. The performance of proposed method is shown in Fig. 2. It is seen that use of secret key provides nil influence on encryption and re-encryption. The other algorithms has to do a exponentiations over the secret key, since, the computation time increases with increasing size of secret key. It is known that the size of exponentiations affects the efficiency of proposed one. Thus if the random number length is fixed during encryption and re-encryption, the impacts of other related parameters on cost can be reduced. It is found that the increased length of secret key has caused a prominent increase in the computational cost, since the increased key accompanies more of computation to encrypt or decrypt the original message. The increase in key length is subjected to increase with the density of the network, where privacy might tend to reduce when the key size is smaller for large dense network. The increase in key length increases automatically the overhead of computation and it reluctantly reduces the faster delivery of encrypted messages. Hence, the key size has to be maintained at a promisable rate to avoid overhead and complexity (Table 2).

Fig. 2
figure 2

Influence of Secret Key

Full size image
Table 2 Influence of Secret Key
Full size table

7.2 Influence of ‘n’ length

This is adopted using various size of module between [256–1048] in the multiples of 2. The results are shown in Fig. 3. The impact of length of n has influenced well the efficiency of VANETs. It is seen that the computation time of the proposed method increases with increasing n length, on the other hand, the increasing modular size provides higher guarantees on security. It is seen that the computation time of the proposed algorithm has reduced 10 ms less for 1024/2048 bits. This can be generalized as the decryption is found efficient than the other related operations, which is appropriate for the u with very scarce resource. It could be seen that this algorithm fits well with the operation for VANET security, where u plays a major role with limited resource. Additionally, it is seen that the correctness of the proposed method is proved in terms of better implementation and analysis. From the results, the initial decryption in 2-Dn has higher computation cost due to four exponents, however, 2-Dn has reduced time than IREn and IREn. This is due to the presence of efficient 160 bit exponentiation than IREn and IREn. Also, 2-Dn has one exponentiations and it is found more efficient than the others (Table 3).

Fig. 3
figure 3

Influence of length n

Full size image
Table 3 Influence of length n
Full size table

7.3 Input data length

The proposed system is tested for its flexibility for different length based on input data mi and this is set with different value of range [50,500] bits in multiples of 2. It is seen from the Fig. 4 that the computation time of the algorithm varies less w.r.t original input data. This could be inferred that the proposed method has the ability to deal with various messages from the vehicle with higher privacy protection.Since, the method does not employs any compression method to compress the larger message, the overhead of the system tend to increase. Also, the computational efficiency of the system is largely reduced, when the overhead of system increases. Additionally, loosing of message tend to happen when the sent message for encryption is of higher quantifiable rate (Table 4).

Fig. 4
figure 4

Influence of input data length

Full size image
Table 4 Influence of input data length
Full size table

The other factors apart from the length of the message or n bits or secret key is the speed of the vehicle, mobility, delay, total number of broadcasted message and vehicle density.

8 Conclusions

This paper proposes a privacy preservation scheme using Full Harmonic Encryption (FHE) in VANETs to preserve the message during transmission with higher end security. Hence, amodified FHE with re-encryption scheme is utilized to improve the quality of encryption/decryption of the messages with high privacy. Taking into consideration, the higher mobility of vehicles inside the VANETs, the re-encryption scheme is designed specifically to reduce the computation time. Also, mutual authentication is provided with the use of as and ds in the VANET architecture. A special consideration is given to re-encryption in homomorphic encryption, which provides dual authorized entry of vehicles inside the MANETs. Also, the security requirement is guaranteed with this dual authentication scheme with higher privacy.