Sage Journals: Discover world-class research

Abstract

The mobile wireless sensor network (MWSN) is a new style WSN with mobile sinks or sensors in the network. MWSN has advantages over static WSN in the aspect of better energy efficiency, improved coverage, and superior channel capacity. However, mobile nodes also bring some security problems. For example, it is difficult to ensure secure communications among the mobile nodes and static nodes. In this paper, we design a lightweight mobile reauthentication protocol for mobile nodes. The designed protocol provides forward secure pairwise key for the mobile node when it moves from one cluster to another. Thus, the mobile sensor node can be authenticated by the new cluster head, and the privacy of his origin area is protected. In addition, the security and performance analysis shows that our scheme meets the need of lower communication and computation overhead, while achieving security requirement for mobile sensor node application in MWSN.

1. Introduction

WSN has become more and more prospective in human life with the development of hardware and communication technologies. However, due to the static network style, there are some natural limitations of WSN, such as network connectivity and network lifetime [1–4]. Furthermore, more and more researches find that the mobility in WSN not only improves the overall network lifetime and the data capacity of the network, but also addresses delay and latency problems [5–9]. There are many researches on how to realize better energy efficiency, improve coverage, enhance target tracking, and cause superior channel capacity for MWSN. However, limited researches consider the issue caused by the mobile sensor nodes, such as credibility with low consumption overhead and secure communication in MWSN. While more and more application scenarios require mobile sensors in WSN, such as traffic detection, animal observation E-Health, and battlefield. Furthermore, some present researches begin to consider the mobile adversary [10], which brings new security problems. Therefore, we should pay attention to realize the mutual authentication between the mobile node and the cluster efficiently, generate the new pairwise key, and make sure of the security of data transmission.

The framework of MWNS is given as in Figure 1. The network considers four types of entities: (1)

base station—as usual, the base station is assumed to be absolutely secure, which has plenty bandwidth, energy, storage space, and computation capability;

(2)

cluster head—cluster head is assumed to have more storage space, energy, communication range, and computation capability than sensor node, and notice that, in general, the communication range of cluster head is also larger than the sensor node;

(3)

static sensor node—we consider static sensor nodes in our network model, since they can work for the cluster head, and relay for mobile sensor node which has smaller communication range than cluster head, and in general, we assume that it has limited storage space, energy, and communication range;

(4)

mobile sensor node—the mobility is the only difference between the mobile sensor node and the static sensor node, and the mobile sensor node roams from one cluster to another cluster and communicates with the nodes in the cluster.

Figure 1

The framework of MWSN.

Based on the framework of MWSN, some security problems are brought by the mobility of mobile sensor node. One problem is the identity authentication of mobile sensor node in the new cluster. The other is the new session key generated between mobile sensor node and new cluster to ensure security communication. Moreover, we should protect the privacy of the mobile sensor node which means attackers cannot track it. Therefore, the security requirements of MWSN are given as follows: (1)

the authentication of identity—making sure that the identities of both parties who generate the key are trusted;

(2)

data integrity—ensuring that only authorized person can modify the transmission of information;

(3)

message privacy—transferring data can only be read by authorized parties;

(4)

key freshness: ensuring that the existing session key is only used at current stage;

(5)

node Resiliency—the network recovers when some nodes are captured by malicious attackers;

(6)

privacy-preserving—since the mobile sensor node roams in WSN, attackers cannot track the mobile sensor node;

(7)

scalability—allowing revoking and joining nodes. With the expanding of the network scale, it has little effect on the storage space of nodes and communication consumption.

In this paper, we focus on the security requirements caused by mobile sensor nodes. For the mobile sensor node in MWSN, we present an efficient node reauthentication and key generation scheme for mobile sensor nodes which consumes less communication and computation overhead and protects the privacy of the mobile sensor node. The security and performance analysis shows that our re-authentication scheme for mobile sensor node cannot only efficiently realize the secure requirements for MWSN, but also suit for the limited resource WSN.

The rest of this paper is organized as follows. In Section 2, we introduce the related work. We present our protocol in Section 3. Section 4 gives the security analysis of re-authentication protocol. Section 5 gives the performance analysis and simulation. Finally, we conclude the paper in Section 6.

2. Related Work

We introduce our related work from three aspects: the lightweight authentication schemes for WSN, the research of mobile sink in MWSN, and the re-authentication schemes for the mobile sensor node in MWSN.

The demand of lightweight is mostly considered in sensor network. All nodes in sensor network are considered to be static initially. For example, Perrig et al. [11] proposed a typical authentication scheme named μTESLA (Timed Efficient Stream Loss-Tolerant Authentication) by using the one-way hash chain. The protocol publishes the authentication key $K_{mac}$ through delay to ensure that before the $K_{mac}$ is published, the attacker cannot forge the correct broadcast packets. Du et al. [12] constructed an authentication path based on the public key mechanism by using Merkle Tree to reduce the computation and communication overheads. He also proposed dividing the entire WSN network into region Merkle Tree which can reduce the height of the Tree and the hops of the authentication. Ibriq and Mahgoub [13] proposed an efficient authentication program in which BS (Base Station) acts as the role of Certificate Authentication (CA) and assigns part of its functions to CH (Cluster Head). A sink can generate a key from “Partial Key Escrow Tab” [13] in all nodes and can be elected as Cluster Head. After the data integrated, messages are exchanged among cluster heads and finally transmitted to BS. However, since the partial key escrow Tab should be stored in every node, this scheme needs additional storage space. All these authentication protocols are for static nodes without considering roaming issue.

The advantages of mobile sink in MWSN have attracted much attention. Zhang et al. [14] proposed several efficient schemes to restrict the privilege of a mobile sink without impeding its capability of carrying out authorized operations for an assigned task. To prevent the authenticator from revealing information due to mobile sink compromises, the privileges of the authenticator are restricted by adding parameters, such as the starting time and the ending time of a task, the type of a task, and ID of the mobile sink. Vieira et al. [15] proposed a bioinspired location service named Phero-Trail location service protocol. In Phero-Trail, location information is stored in a 2D upper hull of a Sensor Equipped Aquatic Swarm, and a mobile sink uses its trajectory projected to the 2D hull to maintain location information. This enables mobile sensors to efficiently locate a mobile sink. The results show that Phero-Trail performs better than existing approaches. Agrawal et al. [16] proposed a key update protocol which securely updates the session key between a pair of nodes with the help of random inputs in mobile sensor networks. The security analysis shows that the proposed protocol resists known-key, impersonation, replay, worm, and sink hole attacks, while also provides forward secrecy, key freshness, and key control.

Recently the security of mobile sensor nodes in WSN has been paid more and more attention. Han and Kim [17] proposed the re-authentication issue concerning mobile nodes moving among sink nodes. The scheme considers the sink in the home cluster as a trusted third party. It prestores authentication information in all surrounding neighbor clusters and transfers the credible information to the new sink. The communication and computation overhead of re-authentication is reduced through credible trust. Qiu et al. [18] considered a sensor node roaming within a very large and distributed wireless sensor network, such as the application of healthcare field, in which the sensor nodes are deployed in the patient's body. When a dynamic sensor node moves to new area and wants to attack a router or a cluster head in this area, it first sends a request message to the base station. After verifying validity of the request message, the base station generates the session key for mobile node and the router and sends it to the router, and then the router sends the material of session key to the mobile node to generate the session key. Qiu also improves the E-G scheme to guarantee that two sensor nodes share at least one key with probability 1 with less storage and energy overhead. The disadvantages of Han's scheme are as follows. First, it only takes the mobile node, sink node, and base station into consideration. Then, the communication overhead of the program mostly concentrates on the mobile node, so it has influence on the lifetime of the mobile node. Lastly, the re-authentication material is prestored in the neighbor clusters, which exists unnecessary communication overhead and information leak. In Qiu's scheme, the base-station is always online and provides the full utilities. The re-authentication also depends on the base station which incurs large communication overhead.

3. The Proposed Protocol

With the mobility of MWSN, the mobile sensor nodes may move from one cluster to another. If we repeat the new nodes addition process proposed in [4], the scheme will degrade to the E-G [19] scheme. Besides, some predistribution schemes need to interrupt the operation of network and implemented by man, which is unrealistic for the running wireless network. Therefore, the roaming behavior of mobile sensor nodes must consider how to get trust from the new cluster and generate the pair-wise key to achieve security communication. Considering the security requirements caused by mobile sensor nodes in the MWSN, we design our re-authentication protocol as follows. For convenience, we assume that after the static WSN generated the communication keys for the entities, the mobile sensor nodes join the network from some clusters.

As Figure 2 shows, the whole process can be divided into two phrases. Phrase 1 the mobile sensor node $M_{A}$ first registers at the base station and then joins the network from the fixed cluster ${CH}_{A}$ (home cluster). The main purpose of this phase is to make $M_{A}$ initial trustily join the network. To ensure security, this phrase is realized by offline distribution. Phrase 2 $M_{A}$ moves to another new cluster ${CH}_{B}$ (foreign cluster), and it should pass the legal identify authentication of ${CH}_{B}$ before enjoying the resource of this cluster. To realize the mutual authentication of $M_{A}$ and ${CH}_{B}$ , we can use the trusted relationship among $M_{A}$ , ${CH}_{A}$ , and ${CH}_{B}$ . We describe the further details of re-authentication as follows. The notation used throughout our scheme is shown in Table 1.

Table 1

Notation.

Notation	Description
T	Timestamp
$M_{A}$	The mobile node
$C H_{A}, C H_{B}$	The cluster head
$K_{A, B}$	The pair-wise key for A and B
${M}_{K}$	Encrypt message M by K
$MAC (k, M)$	The message authentication code of M using K
$H ()$	Hash function
$\| \|$	Message connecting
⊕	Xor

Figure 2

The re-authentication of mobile sensor node.

3.1. Predistribution for Mobile Sensor Node

Before roaming in the network, $M_{A}$ should register in the base station and get the pre-distribution material by offline. Then, $M_{A}$ joins the network from cluster head ${CH}_{A}$ . After being authenticated by ${CH}_{A}$ , $M_{A}$ has the information including a hash number $H (I)$ , a random number $R_{1}$ , and the session key $K_{{CH}_{A}, M_{A}}$ .

3.2. Mobile Sensor Node Reauthentication

After registering at the base station and trusted joining ${CH}_{A}$ , $M_{A}$ can roam in the network among clusters. As Figure 2 shows, after completing Phase 1, the mobile sensor node $M_{A}$ moves to a foreign cluster ${CH}_{B}$ , and $M_{A}$ should pass the authentication of ${CH}_{B}$ before communicating with other nodes in ${CH}_{B}$ . The implementation mutual authentication of re-authentication protocol is shown in Figure 3. (1)

When $M_{A}$ moves to the new cluster ${CH}_{B}$ , it first launches the authentication procedure to ${CH}_{B}$ . $M_{A}$ sends the message 1: $M_{A}$ , ${CH}_{A}$ , $t_{1}$ , $MA C_{1}$ to ${CH}_{B}$ , where $t_{1}$ is the timestamp, and $MA C_{1} = (K_{C H_{A}, M_{A}}, M_{A} | | t_{1} | | H (I))$ .

(2)

When ${CH}_{B}$ receives message 1 at time $t *$ , and ${CH}_{B}$ first checks whether $(t * - t_{1}) \leq Δ t$ . If the result is valid, since there is no shared information between ${CH}_{B}$ and $M_{A}$ , and ${CH}_{B}$ would send message 2: $M_{A}$ , $t_{2}$ , $t_{1}$ , $MA C_{1}$ , $MA C_{2}$ to ${CH}_{A}$ , where $MA C_{2} = (K_{{CH}_{B}, {CH}_{A}}, M_{A} | | t_{2} | | t_{1} | | MA C_{1})$ .

(3)

Upon Receiving the message 2, the home cluster ${CH}_{A}$ verifies message as Algorithm 1 and replies message 3: $t_{3}, e_{1}, MA C_{3}$ to ${CH}_{B}$ .

(4)

After receiving the message 3, ${CH}_{B}$ verifies message as Algorithm 2. Then, $C H_{B}$ sends the message 4: $t_{4}, h_{1}, MA C_{4}$ to $M_{A}$ .

(5)

Upon receiving the message 4 from ${CH}_{B}$ , the mobile node $M_{A}$ executes Algorithm 3 to get the pair-wise key

\begin{matrix} K_{{CH}_{B}, M_{A}} = H (H (I) ∥ R_{1} ∥ R_{2}) . \end{matrix}

(1)

After generating the session key,

M_{A}

verifies the correctness of

MA C_{4}

. If the validation is right, the session key is right. During the communication with

{CH}_{B}

H (I)

and

R_{1}

should be updated as

H (I^{'})

and

R_{1}^{'}

which are used as the authentication material for the further re-authentication.

{CH}_{B}

also sends these information to the base station.

Algorithm 1: The process executed by ${CH}_{A}$ to message 2.

Require: $M_{A}$ , $t_{2}$ , $t_{1}$ , $MA C_{1}$ , $MA C_{2}$ .

1: Verify $(t  * - t_{2}) \leq Δ t$ .

2: if $t_{2}$ is valid then

3: Compute $MA C_{2} * = (K_{C H_{B}, C H_{A}}, M_{A} | | t_{2} | | t_{1} | | MA C_{1})$ .

4: if $MA C_{2} * = MA C_{2}$ then

5: Compute $MA C_{1} * = (K_{C H_{A}, M_{A}}, M_{A} | | t_{1} | | H (I))$ .

6: if $MA C_{1} * = MA C_{1}$ then

7: Compute $e_{1} = {H (I), R_{1}}_{K_{C H_{A}, C H_{B}}}$ and

$MAC (K_{C H_{A}, C H_{B}}, t_{3} | | e_{1})$ .

8: end if

9: end if

10: end if

Algorithm 2: The process executed by $C H_{B}$ to message 3.

Require: $t_{3}, e_{1}, MA C_{3}$ .

1: Verify $(t  * - t_{3}) \leq Δ t$ .

2: if $t_{3}$ is valid then

3: Compute $MA C_{3} * = MAC (K_{C H_{A}, C H_{B}}, t_{3} | | e_{1})$ .

4: if $MA C_{3} * = MA C_{3}$ then

5: Extract $H (I)$ and $R_{1}$ .

6: Generate random number $R_{2}$ , compute $K_{C H_{B}, M_{A}}$

as formula (1).

7: Compute $h_{1} = H (R_{1}) \oplus R_{2}$ .

8: Compute $MA C_{4} = (K_{C H_{B}, M_{A}}, H (I) | | t_{4} | | h_{1})$ .

9: end if

10: end if

Algorithm 3: The process executed by $M_{A}$ to message 4.

Require: $t_{4}, h_{1}, MA C_{4}$ .

1: Verify $(t  * - t_{4}) \leq Δ t$ .

2: if $t_{4}$ is valid then

3: Compute $H (R_{1})$ , and $R_{2} * = h_{1} \oplus H (R_{1})$ .

4: Compute $K_{C H_{B}, M_{A}} *$ as formula (1).

5: Compute $MA C_{4} * = (K_{C H_{B}, M_{A}} *, H (I) | | t_{4} | | h_{1})$ .

6: if $MA C_{4} * = MA C_{4}$ then

7: $K_{C H_{B}, M_{A}} * = K_{C H_{B}, M_{A}}$ .

8: end if

9: end if

Figure 3

Re-authentication of mobile sensor nodes.

For convenience, the role of home cluster is acted by the foreign cluster node through which the mobile sensor nodes have completed the re-authentication process. That means that after $M_{A}$ completes re-authentication in the foreign cluster ${CH}_{B}$ , ${CH}_{B}$ is the new home cluster of mobile sensor node. When $M_{A}$ moves to another foreign cluster ${CH}_{C}, {CH}_{B}$ acts as the home cluster which responsibly completes the re-authentication between $M_{A}$ and ${CH}_{C}$ .

Taking the issue of tracking and protecting the privacy of $M_{A}$ into account, we use the pseudonyms methods [20, 21] during the communication. The whole time of $M_{A}$ in ${CH}_{B}$ is divided in accordance with the time slice $T S_{j}$ , and the length of each time slice is $Δ t$ , which means we can get C time slices. We denote $PI D_{A, j}$ as the pseudonym of $M_{A}$ in the time slice $T S_{j}$ , where $PI D_{A, j}$ is generated by two hash seeds $H (I)$ and $R_{2}$ as formula (2)

\begin{matrix} S_{1, j} = H^{j} (R_{2}), \\ S_{2, j} = H^{j} (H (I)), \\ PI D_{A, j} = H (S_{1, j} \underset{}{\oplus} S_{2, j}) . \end{matrix}

(2)

Notice that since ${CH}_{B}$ knows $R_{2}$ and $H (I)$ , so it can trace the messages sent by $M_{A}$ in its communication range. While $M_{A}$ moves to ${CH}_{C}$ at $T S_{j}$ (actually ${CH}_{C}$ only knows a mobile sensor node named $PI D_{A, j}$ joining its cluster), because $C H_{B}$ does not have the materials to generate the pseudonyms, so it cannot trace the messages sent by $M_{A}$ out of its communication range. By this way, we can protect the privacy of $M_{A}$ .

4. Protocol Security Analysis

4.1. The Protocol Satisfies Forward Security

Suppose that the attacker gets the session key $K_{C H_{C}, M_{A}}$ between the mobile sensor node $M_{A}$ and cluster node ${CH}_{C}$ . It is difficult for attackers to derive the session key used before such as $K_{{CH}_{B}, M_{A}}$ . The session key between $M_{A}$ and ${CH}_{B}$ is determined by two random numbers $R_{1}$ and $R_{2}$ . $R_{1}$ is produced in the last re-authentication cycle and is transmitted in the ciphertext. $R_{2}$ is transmitted by the XOR hash value $h_{1}$ in message 4. If the attacker wants to obtain the plaintext $R_{1}$ , he must know the session key $K_{{CH}_{A}, M_{A}}$ between $M_{A}$ and ${CH}_{A}$ . Thus, the problem is deduced into how to get the session key between $M_{A}$ and the first cluster $C H_{A}$ . $K_{C H_{A}, M_{A}}$ is sent offline, which is assumed to be secure. $R_{2}$ is gotten by the hash and XOR of the hash value of $R_{1}$ , and according to the irreversibility of hash, the problem of obtaining the plaintext of $R_{2}$ is derived to obtain the plaintext $R_{1}$ . Even if attackers get the current session key of $M_{A}$ , they cannot derive the previous session key of $M_{A}$ through the previously analysis. The protocol satisfies forward security.

4.2. Mutual Identity Authentication

In our scheme, as there is no shared information between $C H_{A}$ and $C H_{B}$ , $C H_{B}$ cannot verify the identity of $M_{A}$ , so when $C H_{B}$ receives message 1, it transfers the message to $C H_{A}$ . $C H_{A}$ helps $C H_{B}$ authenticate the identity of $M_{A}$ by computing $MA C_{1}$ through using the hidden $H (I)$ . $M_{A}$ authenticates the identity of the foreign cluster $C H_{B}$ mainly through $MA C_{4}$ which also uses the hidden $H (I)$ . If $MA C_{4}$ is right, we believe that $C H_{B}$ has the right identity. By this way, we realize mutual identity authentication.

4.3. Prevent Man-in-the-Middle Attack

From the analysis of our scheme, an attacker can track or intercept message 1 to act the mobile sensor node $M_{A}$ and continue communicating with foreign cluster head. It makes the entire protocol go on running. Finally, feedback message (message 4) is gotten to extract the session key material. However, according to the analysis of forward security, $R_{1}$ and $R_{2}$ are not sent in plaintext. In order to attack the protocol, the previous session key should be known. And the whole problem is back to the security of $K_{C H_{A}, M_{A}}$ . For man-in-the-middle attack, as mentioned in mutual identity authentication, mutual identity authentication ensures the correctness of the identity of the message sender. $MAC$ used in every message ensures the message integrity. According to the general security assumption of $MAC$ [17], attackers cannot construct a valid message to achieve communication. So the protocol can prevent man-in-the-middle attack.

4.4. Prevent Replay Attack

When the mobile node $M_{A}$ applies to join registered foreign cluster, every message of our scheme has the current timestamp ( $t_{1}$ , $t_{2}$ , $t_{3}$ , $t_{4}$ ). The message received in $Δ t$ time, to some extent, can prevent replay attack. According to the session key generated in formula (1), the generation of session key selects new random number, which ensures the freshness of session key and prevents replay attack effectively.

4.5. Protect the Privacy of the Mobile Node

Since the communication of mobile sensor node uses the pseudonyms, attackers and other entities cannot distinguish them which protects the privacy of the mobile sensor node. But to the base station and cluster heads, they can track the mobile sensor node. After the mobile node joins the foreign cluster, the cluster head sends the $H (I)$ and $R_{1}$ to the base station, which helps the base station to track and manage the mobile sensor node. However, for the cluster head (such as $C H_{C}$ ), $C H_{C}$ only knows that the pseudonyms of the mobile sensor node $M_{A}$ is in its cluster. When the mobile sensor node $M_{A}$ moves to a new cluster head (such as $C H_{D}$ ), $C H_{C}$ does not know the pseudonyms of $M_{A}$ , and it cannot track $M_{A}$ . Therefore, the privacy preserving of mobile sensor node is conditional.

5. Protocol Performance Analysis

We give the performance analysis of our scheme in this section in terms of communication pass, message size, and computation overhead. We also give the simulation of our scheme on the NS2 simulation platform and use the time delay to reflect the efficiency of our scheme.

5.1. Communication Pass

We compared the required number of communication passes with Han's [17] and Qiu's [18] schemes, since both of them propose the reauthentication protocols for mobile sensor nodes in WSN. Table 2 shows the comparison of communication passes for mobile node reauthentication, where n denotes the number of hops from $M_{A}$ to the foreign cluster head (sink), m denotes the number of neighbor cluster heads (sinks) around the home cluster head, and l denotes the number of hops from foreign cluster (sink) to the base station which is used in Qiu's scheme.

Table 2

The required number of communication passes.

	Han's scheme	Qiu's scheme	Our scheme
Node to Sink	$2 n$	n	n
Sink to Sink	m	l	2
Sink to Node	1	1	1
BS to Sink	—	1	—

Since Han's and our schemes use the relation among cluster heads to realize re-authentication for mobile sensor node, which do not need communication with the base station, in Qiu's scheme, when the node joins a new sink, it first sends the requirement message to the base station. Actually, the message is first sent to the foreign cluster head (n hops) and then to the base station via the foreign cluster (l hops) which incurs large communication overhead. The hole communication passes are ( $n + l$ ) hops.

Although the re-authentication of Han's scheme does not need communication with the base station, he pre-stores the authentication information in all surrounding neighbor cluster heads which are related with the number of neighbor cluster (m hops), while our scheme realizes the re-authentication by the tradition tripartite authentication, which results in less communication pass.

5.2. Message Size

The message size during the re-authentication process is quantified by the byte which is to show the communication overhead. We compare the message size with Han's. We use the base parameter setting of message as Han's [17] in Table 3.

Table 3

The base parameter setting of message.

Notation	Length (byte)
MAC	4
Random number	8
Identity	1
Time stamp	8
Key size	16

From Table 4, we can see that our scheme has less message size of the whole re-authentication process. Notice that Han pre-stores authentication information in all surrounding neighbor clusters and we only consider that the number of authentication material of transmission size is 36 bytes, while the actual number may be more than 36.

Table 4

The required message size for re-authentication (byte).

	Han's scheme	Qiu's scheme	Our scheme
Node to Sink	$48 n$	$15 n$	$14 n$
Sink to Sink	$\geq  36 m$	$15 l$	53
Sink to Node	86	22	20
BS to Sink	—	50	—

Total	$\geq  (48 n + 36 m + 86)$	$15 (n + l) + 72$	$14 n + 73$

During re-authentication for mobile sensor node, we reduce the message length transmission among the entities since the data transmission consumes much more energy than computation in WSN.

5.3. Computation Overhead

Computation overhead is quantified by the number of execution encryption algorithm. As Table 5 shows, the total number of encryption/decryption of our scheme is the same as Han's, both more than Qiu's scheme. Since the re-authentication of Qiu's scheme is based on the base station, our scheme and Han's scheme are based on the relation among clusters.

Table 5

The required message size for re-authentication.

	Han's	Qiu's	Ours
Encryption/decryption in total	4	2	4
Encryption/decryption in node	1	1	2
MAC generation/verification in total	8	4	8
MAC generation/verification by node	3	2	4

5.4. Protocol Simulation

We use NS2.29 to simulate our scheme and Han’ scheme [17], since both of them realize mobile node re-authentication without requiring communication with the base station. We use the transmission delay to quantify the message size, which can reveal the availability and efficiency of our scheme. The simulation uses the mesh network topology, MAC layer uses the 802.15.4 protocol written by Zheng and Lee [22] for NS2, the routing layer uses the AODV routing protocol which has the shortest hops, the transportation layer uses the UDP protocol, and the application layer transmits the CBR packet. The message size is set as Table 4. The data transmission speed is 250 KB/S, which adopts the recommended beacon mode standard setting in reference [22].

Supposing the communication radius of the mobile sensor node and the common sensor nodes within the cluster to be 20 m, the communication radius of cluster head is 100 m. The computation delay of message 1 and message 4 in mobile sensor node is 6 ms and 3 ms [23], respectively, while the computation delay of message 2 and message 3 node is 1 ms for cluster head.

To reflect the comparison of Table 4, we design two groups of simulation for our scheme and Han's scheme [17]. The number of each group simulation is 100 times.

In Figure 4, $m = 1$ , and $n = 2$ . The simulation delay of Han's scheme is 26.217 ms, while our scheme is 18.432 ms. However, the time delay of our schemes simulation is not as good as the comparison in Table 4. Since there is an addition MAC layer head for each message, the time delay of simulation is not the same as the comparison of message size in Table 4. From Figure 4, we can know the whole delay of our scheme is less than Han's. On one hand, our scheme has less message size, on the other hand, we reduce the number of messages sending. The fluctuation of the simulation in Figure 4 is caused by $n = 2$ . Since the message transmitted by the static nodes in cluster needs to consume transmission delay (when node transmits message, it will repeat calling the sending and receiving process, and seek the routing table, which leads to more delay time), and it results in the unstability of time.

Figure 4

The time delay for $m = 1$ , $n = 2$ .

In Figure 5, $m = 1$ , and $n = 5$ . The simulation delay of Han's scheme is 60.384 ms, while our scheme is 34.8608 ms. Compared with Figure 4, with the number of relay hops increasing, the advantage of our scheme is more obvious. This is due to less communication message size of mobile sensor node. Moreover, with the number of hops increasing, the instability of the simulation is more obvious.

Figure 5

The time delay for $m = 1$ , $n = 5$ .

6. Conclusion

The security problem brought by the mobile sensors in MWSN attracts more and more attention of researchers. In this paper, we propose a re-authentication protocol for the mobile node roaming among clusters. Our protocol can transfer the credibility among the clusters which can efficiently achieve the requirements of secure identity authentication and establish the forward secure pairwise key. Meanwhile, the base station can track the mobile trajectory and protect the privacy of the mobile sensor node. We also give performance analysis and simulation for our re-authentication protocol. The results and comparison show that our protocol achieves better security and has better performance on communication overhead, message size, and computation cost.

Footnotes

Acknowledgments

This work was supported by the National Natural Science Foundation of China under Grant 61202474 and the Natural Science Foundation of Jiangsu Province under Grant BK2011464.

References

Younis

M. F.

Ghumman

Eltoweissy

Location-aware combinatorial key management scheme for clustered sensor networks

IEEE Transactions on Parallel and Distributed Systems 2006 17 8 865 882

2-s2.0-33646921841

10.1109/TPDS.2006.106

Eltoweissy

Heydari

M. H.

Morales

Sudborough

I. H.

Combinatorial optimization of group key management

Journal of Network and Systems Management 2004 12 1 33 50

2-s2.0-3543105535

10.1023/B:JONS.0000015697.38671.ec

Hussain

Kausar

Masood

An efficient key distribution scheme for heterogeneous sensor networks

Proceedings of the International Wireless Communications and Mobile Computing Conference (IWCMC ′07)

August 2007

New York, NY, USA

388 392

2-s2.0-36849060191

10.1145/1280940.1281023

Xiao

Guizani

Chen

H. H.

An effective key management scheme for heterogeneous sensor networks

Ad Hoc Networks 2007 5 1 24 34

2-s2.0-33750023217

10.1016/j.adhoc.2006.05.012

Munir

S. A.

Ren

Jiao

Wang

Xie

Mobile wireless sensor network: architecture and enabling technologies for ubiquitous computing

in Proceedings of the 21st International Conference on Advanced Information Networking and Applications Workshops

2007

Niagara Falls, Canada

113 120

Rahimi

Shah

Sukhatme

G. S.

Heideman

Estrin

Studying the feasibility of energy harvesting in a mobile sensor network

Proceedings of the IEEE International Conference on Robotics and Automation (ICRA ′03)

September 2003

19 24

2-s2.0-0344876649

Yarvis

Kushalnagar

Singh

Rangarajan

Liu

Singh

Exploiting heterogeneity in sensor networks

Proceedings of the IEEE 24th Annual Joint Conference of the IEEE Computer and Communications Societies (NFOCOM ′05)

March 2005

Hillsboro, Ore, USA

878 890

2-s2.0-25844505160

Wang

Srinivasan

Chua

Using mobile relays to prolong the lifetime of wireless sensor networks

Proceedings of the 11th Annual International Conference on Mobile Computing and Networking (MobiCom ′05)

2005

New York, NY, USA

270 283

Liu

Brass

Dousse

Nain

Towsley

Mobility improves coverage of sensor networks

Proceedings of the 6th ACM International Symposium on Mobile Ad Hoc Networking and Computing (MOBIHOC ′05)

May 2005

New York, NY, USA

300 308

2-s2.0-29844435167

10.

Pietro

R. D.

Soriente

Spognardi

Tsudik

Collaborative authentication in unattended wsns

Proceedings of the 2nd ACM Conference on Wireless Network Security (WiSec ′09)

2009

New York, NY, USA

237 244

11.

Perrig

Szewczyk

Tygar

J. D.

Wen

Culler

D. E.

SPINS: security protocols for sensor networks

Wireless Networks 2002 8 5 521 534

2-s2.0-0036738266

10.1023/A:1016598314198

12.

Wang

Ning

An efficient scheme for authenticating public keys in sensor networks

Proceedings of the 6th ACM International Symposium on Mobile ad hoc Networking and Computing (MobiHoc ′05)

2005

New York, NY, USA

58 67

13.

Ibriq

Mahgoub

A hierarchical key establishment scheme for wireless sensor networks

Proceedings of the 21st International Conference on Advanced Information Networking and Applications (AINA ′07)

May 2007

NiagaraFalls, Canada

210 219

2-s2.0-34548720911

10.1109/AINA.2007.14

14.

Zhang

Song

Zhu

Cao

Least privilege and privilege deprivation: Towards tolerating mobile sink compromises in wireless sensor networks

Proceedings of the 6th ACM International Symposium on Mobile Ad Hoc Networking and Computing (MobiHoc ′05)

May 2005

New York, NY, USA

378 389

2-s2.0-29844438193

15.

Vieira

L. F. M.

Lee

Gerla

Phero-trail: a bio-inspired location service for mobile underwater sensor networks

IEEE Journal on Selected Areas in Communications 2010 28 4 553 563

10.1109/JSAC.2010.100505

16.

Agrawal

Roman

Das

M. L.

Mathuria

Lopes

A novel key update protocol in mobile sensor networks

7671

Proceedings of the 8th International Conference on Information Systems Security (ICISS ′12)

2012

194 207 Lecture Notes in Computer Science

17.

Han

Kim

Untraceable mobile node authentication in wsn

Sensors 2010 10 4410 4429

10.3390/s100504410

18.

Qiu

Zhou

Baek

Lopez

Authentication and key establishment in dynamic wireless sensor networks

Sensor 2010 10 3718 3731

19.

Eschenauer

Gligor

A key management scheme for distributed sensor networks

Proceedings of the 9th ACM Conference on Computer and Communications Security (CCS ′02)

2002

New York, NY, USA

41 47

20.

Jiang

Zhu

Wang

A conditional privacy scheme based on anonymized batch authentication in vehicular ad hoc networks

Proceedings of the Wireless Communications and Networking Conference (WCNC ′13)

2013

Shanghai, China

21.

Sun

Lin

Shen

An efficient pseudonymous authentication scheme with strong privacy preservation for vehicular communications

IEEE Transactions on Vehicular Technology 2010 59 7 3589 3603

2-s2.0-77956735087

10.1109/TVT.2010.2051468

22.

Zheng

Lee

A comprehensive performance study of IEEE 802.15.4

Proceedings of the 4th IEEE International Conference on Wireless and Mobile Computing, Networking and Communications (WIMOB ′08)

2008

580 585

23.

Meulenaer

G. D.

Gosset

Standaert

F. X.

Pereira

On the energy cost of communication and cryptography in wireless sensor networks

Sensor Network Operations 2006 chapter 4

New York, NY, USA

Wiley-IEEE Press

218 237

A Privacy-Preserving Reauthentication Scheme for Mobile Wireless Sensor Networks

Abstract

1. Introduction

2. Related Work

3. The Proposed Protocol

3.1. Predistribution for Mobile Sensor Node

3.2. Mobile Sensor Node Reauthentication

Algorithm 1: The process executed by CH A to message 2.

Algorithm 2: The process executed by C H B to message 3.

Algorithm 3: The process executed by M A to message 4.

4. Protocol Security Analysis

4.1. The Protocol Satisfies Forward Security

4.2. Mutual Identity Authentication

4.3. Prevent Man-in-the-Middle Attack

4.4. Prevent Replay Attack

4.5. Protect the Privacy of the Mobile Node

5. Protocol Performance Analysis

5.1. Communication Pass

5.2. Message Size

5.3. Computation Overhead

5.4. Protocol Simulation

6. Conclusion

Footnotes

Acknowledgments

References

Algorithm 1: The process executed by ${CH}_{A}$ to message 2.

Algorithm 2: The process executed by $C H_{B}$ to message 3.

Algorithm 3: The process executed by $M_{A}$ to message 4.