Accelerated broadcast authentication with signature amortization for wsns

IJRET: International Journal of Research in Engineering and Technology eISSN: 2319-1163 | pISSN: 2321-7308
_______________________________________________________________________________________
Volume: 03 Special Issue: 15 | Dec-2014 | IWCPS-2014, Available @ http://www.ijret.org 72
ACCELERATED BROADCAST AUTHENTICATION WITH
SIGNATURE AMORTIZATION FOR WSNS
Minnu Meria Mathew1
, Anjitha Mary2
1
Student, Computer Science & Engineering, ASIET, Kerala, India
2
Assistant Professor, Computer Science & Engineering, ASIET, Kerala, India
Abstract
Asymmetric Key Cryptography is widely used in broadcasting areas for authentication. But it is considered to be expensive to
wireless sensor networks. This proposed system is a novel broadcast authentication scheme based on PKC with signature
amortization. This scheme uses single Signature for authenticating a group of broadcast messages. As a result, the overhead is
spread over that group of broadcast messages. Moreover, this scheme gives high security and low overhead also. But signature
verification in ECDSA slower than signature generation. So, broadcast authentication with ECDSA has also suffered large energy
consumption and lengthy verification delay. To reduce, this system uses cooperation among sensor nodes, which helps to
accelerate the signature verification. During Signature verification, sensor nodes which have high energy allowed to leave the
intermediary results of the signature verification process to their neighbors for accelerating the same. Simulation results show
that the overhead of message authentication and the delay of verification of authenticated messages is reduced
significantly.
Keywords: Wireless sensor networks; Broadcast authentication; Cooperative Communication; signature
amortization.
--------------------------------------------------------------------***----------------------------------------------------------------------
1. INTRODUCTION
Wireless Sensor Network composed of one or more base
station and number of sensor nodes that are scattered to
monitor environmental and physical conditions like
pressure, temperature, humidity. Base stations are the nodes
that connects users and sensor nodes. A sensor node has
limited power, storage, communication and computation.
WSN is often deployed in hostile situations which is lack of
protection of the environment, and extremely vulnerable to
attack like any other conventional network. For example, an
adversary eavesdropping radio channel, can inject, intercept,
modify, and destruct data packets. But the limited resource
characteristics and unique application features of it needs
some extra security requirements with the typical network
requirements.
In WSN, Broadcast authentication is an important security
concern, as it supports authenticated broadcasting of
messages. Generally, Point-to-point authentication is by
symmetric key cryptographic techniques. In this, the secret
key with sender and receiver is same by which the
cryptographic message authentication code over each
message is computed. In sensor networks, broadcast
authentication is difficult. The symmetric approaches for
point-to-point authentication are not secured enough for
broadcast authentication, where sensor nodes are not
mutually trusted. In symmetric scenario, compromised
receiver node can forge messages from the sender node. So
broadcast authentication for WSNs encounters many
challenges. Due to the constraints of sensor nodes, the
broadcast authentication schemes by conventional Public
Key Cryptography (PKC) are too expensive for WSNs due
to this reason.
Perrig [9] propose µTESLA as a broadcast authentication
scheme for WSNs. µTESLA exploits symmetric scenario to
authenticate broadcasted messages. It satisfies limitations of
WSNs. S it is efficient. But there is chances for node
compromise attacks because of delayed disclosure of secret
keys. To boost the performance of µTESLA, number of
variants[10] [11] [12] are proposed. But this group of
µTESLA have some shortcomings such as maintenance of
time synchronization and distribution of the initial
parameter. These are achieved by unicast transmission. So it
shows heavy overhead. Another shortcoming is delayed
authentication.
For the public-key based authentication technique, each
message is transmitted along with the digital signature of the
message which produced using the sender’s private key.
Every intermediate forwarder and the last receiver can
authenticate by using public key of sender. One of the recent
developed technique under public-key cryptographic
schemes is elliptic curve cryptography (ECC). It is
considered to be more beneficial in terms of memory usage,
message complexity, and security resilience.. RSA [16] is a
public key cryptography based signature algorithm that
widely used for authentication purposes today. ECC [14]
also offers equivalent security as compared with RSA at
much smaller key sizes. The smaller keys shows savings in
memory, bandwidth and computational power usage since
constraints of WSNs. So now-a-days ECC become more

_______________________________________________________________________________________
attractive for constrained wireless devices. Some other
broadcast authentication schemes based on PKC are also
existing which are proposed by Ren [16] and Liu[18]. Some
works are done for finding the techniques to reduce the
overhead of PKc based schemes. The main technique for
this is signature amortization proposed by Gennaro[17]
which is to sign efficiently digital streams.
This paper proposes a novel broadcast authentication
scheme based on PKC with signature amortization for
WSNs. This scheme uses only single signature by ECDSA
to authenticate a group of broadcast messages. So, the
overhead is amortized over that group of broadcast
messages. However, this system shows low overhead and
high security. Security of this system is strong as typical
broadcast authentication schemes by PKC. No time
synchronization is required for this scheme and also it can
achieve immediate authentication. Thus, shortcomings of
µTESLA can be overcome. But ECDSA shows signature
verification delay since its signature verification is much
slower than signature generation. To overcome this, we
propose a scheme with cooperative communication. For this,
the network forms clusters with cluster head, which have
high energy in that group. Each cluster head releases
intermediate results of computation to the neighbours of that
cluster.
2. PROPOSED BROADCAST
AUTHENTICATION SCHEME
The proposed scheme uses ECDSA for authentication,
which is based on Public Key Cryptographic technique. It
also uses a signature amortization technique for reducing the
overhead and allows the network to function efficiently.
This scheme exploit one signature to authenticate a group of
messages. The one and only signature is used to authenticate
the authenticator in first Extended Block (EBlock0 ). The
authenticator in EBlock0 is used to authenticate next
extended Block EBlock1 that has a group of b broadcast
messages and one authenticator. The authenticator in
EBlock1, is used to authenticate EBlock2 that has
authenticator and another group of b broadcast messages.
This process till EBlockk. From this brief, this is clear that
the all messages authenticated by single signature.
WSN applications require low power, less memory space
and bandwidth. ECC is best for this application because of
its speed and security. But ECDSA needs an addition point
and two multi scalar for verifying signature in WSNs, that
because decrease signature verification speed, in the result,
it has also incurred problems such as long verification delay
and high energy consumption. To reduce these problems,
here use cooperation among sensor nodes for accelerating
the verification of a single signature. In this, system allow
some sensor nodes that are selected by clustering to release
the intermediary multiplication results in signature
verification stage to their neighbours during the verification
process. The overall idea of proposed system shown in
Fig.1. The message authentication stage is performed by
three steps:
Step 1 generating extended blocks step
Step 2 broadcasting extended blocks step
Step 3 verifying extended blocks step.
Fig-.1. Overall System
2.1 Extended Block generation
All broadcast messages in M is divided into p blocks
Block1,...,Blockp. Every block Blockk, 1 ≤ k ≤ p, contains b
messages. These blocks together form a vector B =
[Blockp]Ti=1,...,p that shown in equation (1), here n = pb.
(1)
The p messages in each row of B are concatenated into a
long string. Here considering Blockk, so concatenation is
denoted by CON(Blockk).
CON(Blockk) = m(k-1)b+1||...||mkb, 1 ≤ k ≤ p (2)
Then, CON(Blockk) is padded with authenticators in each
block indicated as PAD(CON(Bk)) which shown in
equations (3) where dk+1 is authenticator. Block Blockk and
digest dk+1 comprise an extended block EBlockk.
PAD(CON(Blockk)) =CON(Blockk)||dk+1, 1≤k≤p (3)
Algorithm
1. Splitting-up broadcast messages into p blocks Block1…
Blockp
2. Initialize dk+1 and k=1…p
3. Perform the following steps
3.1 Concatenate messages in Blockk to generate
CON(Blockk)

_______________________________________________________________________________________
3.2 Pad CON (Blockk) with digest dk+1 to generate Pad
(CON (Blockk) )
3.3 Compute the digest of result in step 3.2 using a collision
resistant hash function H
3.4 Let EBlockk = [ Blockk dk+1]
3.5 Decrement k
4. Repeat step 3 till i greater than or equal to 1
5. Sign the digest with senders private key PRs to generate
EBlock0=d1║ E(PRs,d1)
6. Let EBlock= [EBlocki] i=0..p
2.2 Extended Block Broadcasting
EBlockk’s authentication is based on EBlockk−1. Thus,
EBlockk−1 should reach before EBlockk and so on. So this is
by the sequential broadcast with reliability. A receiving
node that receives a message mj, then it first confirms that
mj is in current extended block EBlockk. Then secondly it
also confirms that the previous extended block EBlockk−1
has received and authenticated. If EBlockk-1 has not
authenticated yet, then wait for short period for completion
of its authentication. After completing authentication of
block EBlockk, the message mj will broadcast. The receiver
node accepts message mj and broadcasts an
acknowledgement that indicates all missed messages in that
block. Hence the overhead for acknowledgement of each
message is reduced. That is, here one acknowledgement is
used to specify the missed messages in one extended block.
But it may leads to large acknowledgement size.
2.3 Accelerated Verification of Extended Block
The signature verification process is accelerated by releasing
few intermediate computation results in the WSN by the
sensor nodes. This WSN authentication scheme performs
better compared to other authentication schemes. Nodes that
communicate through cooperative communication transmit
data packets with each other. In WSNs, sensor node are
distributed spatially and the nature of communication is
generally broadcast through wireless medium. Hence,
cooperative communications in WSNs can enhance the
performance, especially network reliability.
2.3.1 Clustering
For implementing co-operative communication, it is needed
to arrange the nodes into clusters. Each cluster has a cluster
head, which has high energy in that cluster. Clustering
means grouping the sensor nodes into clusters. The cluster
formation is done by two steps where the the first step is
group the nodes geographically and second step is Cluster-
Head selection form the cluster-Member nodes. The Cluster-
Head which is the higher level in clusters periodically
transmit here it is intermediate results to the members of the
cluster. Thus, the signature verification becomes easier and
faster for each node. All the time, the Cluster-Head nodes
send data to member nodes. Thus it always spend large
amount of energy than member nodes. In some cases, the
energy of Cluster Head is completely used. So it cannot be
act as Head for long. As a solution for this energy starvation
of some nodes, it is need to periodically re-elect a node
which has highest energy in that cluster as new Cluster-
Head. The fig 3 shows the flow diagram of CH selection.
Fig- 2 Cluster model
Fig -3 Flow diagram of CH selection
2.3.2 Accelerated Verification
In WSNs, the broadcast authentication schemes based on
PKC techniques provides strong security and immediate
message authentication than the same based on symmetric-
key approaches. But the schemes used public-key
Cryptographic techniques, verification of signature is much
slower than the schemes that based on symmetric-key
cryptographic techniques. Due to this, when large number of
message are there for authenticated broadcast, a number of
messages might wait for signature verifications.
By using the cooperation among sensor nodes, this issue of
slower signature verification of ECDSA can be reduced.
This is by randomly release the intermediary computation
results (iR) of signature verification to their neighbours.
Cluster-Head releases the results in this system. Then many
sensor nodes which are the cluster-Members use the
received intermediary results to verify signature.
Generally every sensor nodes independently perform the
same verification procedure during the broadcast
authentication. Every node needs to calculate two scalar
multiplications denoted iR1 and iR2. It leads to high energy
consumption. Thus, some sensor nodes that have high
energy which selected as Cluster-Heads allow to release
some intermediary results to their neighbous. So the member
nodes in a cluster does not need to calculate the two scalar
multiplications. Only one value is calculatedindependently
and other got from Cluster Head. Hence the overall
signature verification time can be reduced noticeably. And
also, the overall energy consumption of network will be
decreased drastically. This is shown in Figure 4.
When a Cluster-Head releases its intermediary
multiplication results (iR), all its neighbouring nodes which
are Cluster-Members accelerate the signature verification by

_______________________________________________________________________________________
just computing one elliptic curve point addition and one
scalar multiplication. When the sensor nodes utilize two
scalar multiplications as intermediary computation values
for verification, there is a chance for receiving fake from an
attacker. To prevent this attack, the sensor nodes are
permitted to use at most only one intermediate value (iR1 or
iR2) from the neighbouring nodes.
Fig-4 Enhanced scheme for ECDSA signature Verification
3. PERFORMANCE ANALYSIS
Accelerated Authentication scheme for WSN is
implemented using JiST simulator (Java in Simulation
Time). This JiST simulator is high performance discrete
event simulation engine which runs on Java Virtual
Machine. Its capabilities are similar to NS2 and Glomosin,
but JiST is able to simulate much larger network. During
experiments, performance metrics such as Throughput,
Signature Verification delay, Routing Overhead,
Communication and Computation Overhead and Energy
Consumption are considered.
Total consumed energy: Overall energy consumed in the
network by all nodes.
Te=Σ Ce, where Te is total consumed energy, Ce is overall
energy by all nodes
End-to End delay: Time taken to transfer a packet from
source to destination node.
End-to-end delay = (Σ recv pkt – sent pkt / Σ recv pkt)*100,
where recv pkt is no. Of received packets, sent pkt is no. Of
sent packets
Overhead: Overall consumption for completing a process.
It calculated as a ratio of the correct packets to the total
received packet.
Overhead = Σ RTR pkt / Σ recv pkt
Table 1 shows the parameter used to simulate the proposed
system. The Results of the simulation are shown in the
Table 2 and Table 3, which shows the Energy Consumption,
Average Verification Delay, Communication Overhead and
Computation Overhead with the different no. Of nodes by
setting up the no of clusters as 4 and different number of
clusters or cluster heads for a sensor network 100 sensor
nodes respectively.
Table-1 Simulation Parameters
Sl.
No.
Parameter Value
1 Number of nodes 150 (varied)
2 Simulation Area 1000m ×1000m
3 Traffic Type CBR
4 MAC Type 802.11 MAC
Layer
5 Packet Size 512 bytes
6 Node Initial Energy 100J
7 Transmission Power 0.007
8 Reception Power 0.007
Table-2 Simulation Results (here, no of clusters=4)
No of
Node
s
Energy
Consumpt
ion (J)
Communi
cation
Overhead
Comput
ation
Overhea
d
Avg
Verificati
on Delay
20 11 230 338 15
30 11 250 354 18
40 13 300 374 20
60 14 348 413 27
80 16 412 445 34
100 17 468 489 42
130 21 490 538 42
Table-3 Simulation Results (here, no of nodes = 100)
No of
Cluste
rs
Energy
Consumpt
ion (J)
Communica
tion
Overhead
Computa
tion
Overhea
d
Avg
Verifica
tion
Delay
2 23 380 514 53
3 25 412 503 48
4 17 468 489 42
5 19 547 498 37
6 22 554 493 33
7 21 587 482 35
8 20 598 479 38

_______________________________________________________________________________________
In this work, the performance analysis is carried out in a
sensor network that simulated using JiST by varying
parameters i.e. number of nodes, while keeping other
parameters constant and also with different number of
clusters. Four PKC based authentication schemes, i.e. RSA,
ECDSA, ECDSA with Signature Amortization and
Accelerated ECDSA Authentication with Signature
Amortization are considered in the comparison. Figure 5, 6,
7 and 8 shows the simulation graphs for different number of
nodes verses Energy Consumption, Communication
Overhead and Computation Overhead, Average Verification
Delay respectively.
0
10
20
30
40
50
60
70
80
90
Energyconsuption(J)
Number of Nodes
Energy Consumption (J)
RSA
ECDS
A
ECDS
A with
SA
ECDS
A with
SA and
CC
Fig-5 Comparison on the basis of Energy Consumption
0
100
200
300
400
500
600
700
CommunicationOverhead
Number of Nodes
Communication Overhead
RSA
ECDSA
ECDSA
with SA
ECDSA
with SA
and CC
Fig-6 Comparison on the basis of Communication Overhead
0
100
200
300
400
500
600
700
ComputationOverhead
Number of Nodes
Computation Overhead
RSA
ECDSA
ECDSA
with SA
ECDSA
with SA
and CC
Fig-7 Comparison on the basis of ComputationOverhead
Fig-8 Comparison on the basis of Avg Verification Delay
4. CONCLUSION
PKC-based broadcast authentication schemes specially
signature-based schemes for WSNs have more attention in
recent years since it shows strong security and immediate
message authentication. As compared with others,
verification of public-key cryptographic signatures
consumes higher amount of energy and this process is also
slow. The proposed scheme is an efficient broadcast
authentication, which accelerating the verification process of
PKC based signatures in WSNs by exploiting the
cooperative communication between sensor nodes. It
reduces overall energy consumption. From the simulation
Results, the overhead of this system is less than the typical
signature-based schemes. Besides this, the proposed system
uses less energy than others. By accelerating the verification
of signature, this scheme reduces the signature verification
delay also.
REFERENCES
[1]. Akyildiz , W. Su, Y. Sankarasubramaniam, and E.
Cayirci, “A survey on sensor networks,” IEEE Commun.
Mag., vol. 40, no. 8, pp. 102–114, 2002.
[2]. S Patil, Dr Vijaya Kumar, S. Sinha snd R. jamil “A
Survey on Authentication Techniques for WSN” IJAER
Vol. 7,no.11,2012
[3]. Z Li and Guang Gong “A Survey on Security in
Wireless Sensor Networks”
[4]. T Kavitha and D Sridharam “Security Vulnerabilities in
Wireless Sensor Networks: A Survey” JIAS, 031-044, 2010
[5]. Y.Wang, G.Attebury and B. Ramamurthy “A Survey of
Security Issues in Wireless Sensor Networks” , IEEE ,Vol.8,
No.2, 2006
[6]. M Luk, A Perrig and B Whillock “Seven cardinal
properties of sensor network broadcast authentication”
ACM , pp. 147-156, 2006
[7]. G. Sharma,S. Bala and Anil K V “ Security Framework
for Wireless Sensor Networks-Review” ScienceDirect, 978-
987, 2012
[8]. A. Perrig, R. Canetti, J. Tygar, and D. Song, “Efficient
authentication and signing of multicast streams over lossy
channels,” in Proc. 2000 IEEE Symposium on Security and
Privacy, pp. 56–73.
[9]. A. Perrig, R. Szewczyk, J. D. Tygar, V. Wen, and D. E.
Culler, “SPINS:security protocols for sensor networks,”
Wireless Networks, vol. 8, no.5,pp. 521–534, 2002.

_______________________________________________________________________________________
[10]. D. Liu and P. Ning, “Multilevel μTESLA: broadcast
authentication for distributed sensor networks,” ACM Trans.
Embeded Computing Syst.,vol. 3, no. 4, pp. 800–836, 2004.
[11]. P. Ning, A. Liu, and W. Du, “Mitigating DoS attacks
against broadcast authentication in wireless sensor
networks,” ACM Trans. Sensor Networking, vol. 4, no. 1,
pp. 1–35, 2008.
[12]. T. Kwon and J. Hong, “Secure and efficient broadcast
authentication in wireless sensor networks,” IEEE Trans.
Computers, vol. 59, no. 8, pp. 1120–1133, 2010.
[13]. F. Amin, A.H.Jahangir and H.Rasifard “ Analysis of
Public Key Crptography for Wireless Sensor Networks
Security” World Academy of Science, 2008
[14]. Asha Rani and M Sinha “Elliptic Curve
Cryptography(ECC) for security in Wireless Sensor
Networks” , IJERT
[15]. G S Quirino, ARL, Ribeiro and E D Moreno “
Asymmetric Encryption in Wireless Sensor Networks”
INTECH , 2012
[16]. K. Ren, W. Lou, K. Zeng, and P. Moran, “On
broadcast authentication in wireless sensor networks,” IEEE
Trans. Wireless Commun., vol. 6, no. 11, pp. 4136–4144,
2007.
[17]. J. M. Park, E. Chong, and H. Siegel, “Efficient
multicast packet authentication using signature
amortization,” in Proc. 2002 IEEE Symposium on Security
and Privacy, pp. 227–240.
[18]. An Liu and Peng Ning “TinyECC: A Configurable
Library for Elliptic Curve Cryptography in Wireless Sensor
Networks” IEEE, 2008
[19]. Y Liu , J Li and M Guizani “PKC Based Broadcast
Authentication using Signature Amortization for WSNs”
IEEE TRANSACTIONS ON WIRELESS
COMMUNICATIONS, VOL. 11, NO. 6, JUNE 2012

Accelerated broadcast authentication with signature amortization for wsns

More Related Content

What's hot (13)

Viewers also liked (20)

Similar to Accelerated broadcast authentication with signature amortization for wsns (20)

More from eSAT Publishing House (20)

Recently uploaded (20)

Accelerated broadcast authentication with signature amortization for wsns