Secure and distributed data discovery and dissemination in wireless sensor networks

Secure and Distributed Data Discovery and
Dissemination in Wireless Sensor Networks
Daojing He, Member, IEEE, Sammy Chan, Member, IEEE, Mohsen Guizani, Fellow, IEEE,
Haomiao Yang, Member, IEEE, and Boyang Zhou
Abstract—A data discovery and dissemination protocol for wireless sensor networks (WSNs) is responsible for updating
configuration parameters of, and distributing management commands to, the sensor nodes. All existing data discovery and
dissemination protocols suffer from two drawbacks. First, they are based on the centralized approach; only the base station can
distribute data items. Such an approach is not suitable for emergent multi-owner-multi-user WSNs. Second, those protocols were
not designed with security in mind and hence adversaries can easily launch attacks to harm the network. This paper proposes the
first secure and distributed data discovery and dissemination protocol named DiDrip. It allows the network owners to authorize
multiple network users with different privileges to simultaneously and directly disseminate data items to the sensor nodes.
Moreover, as demonstrated by our theoretical analysis, it addresses a number of possible security vulnerabilities that we have
identified. Extensive security analysis show DiDrip is provably secure. We also implement DiDrip in an experimental network of
resource-limited sensor nodes to show its high efficiency in practice.
Index Terms—Distributed data discovery and dissemination, security, wireless sensor networks, efficiency
Ç
1 INTRODUCTION
AFTER a wireless sensor network (WSN) is deployed,
there is usually a need to update buggy/old small pro-
grams or parameters stored in the sensor nodes. This can be
achieved by the so-called data discovery and dissemination pro-
tocol, which facilitates a source to inject small programs,
commands, queries, and configuration parameters to sensor
nodes. Note that it is different from the code dissemination
protocols (also referred to as data dissemination or reprog-
ramming protocols) [1], [2], which distribute large binaries
to reprogram the whole network of sensors. For example,
efficiently disseminating a binary file of tens of kilobytes
requires a code dissemination protocol while disseminating
several 2-byte configuration parameters requires data dis-
covery and dissemination protocol. Considering the sensor
nodes could be distributed in a harsh environment,
remotely disseminating such small data to the sensor nodes
through the wireless channel is a more preferred and practi-
cal approach than manual intervention.
In the literature, several data discovery and dissemina-
tion protocols [3], [4], [5], [6] have been proposed for WSNs.
Among them, DHV [3], DIP [5] and Drip [4] are regarded as
the state-of-the-art protocols and have been included in the
TinyOS distributions. All proposed protocols assume that
the operating environment of the WSN is trustworthy and
has no adversary. However, in reality, adversaries exist and
impose threats to the normal operation of WSNs [7], [8].
This issue has only been addressed recently by [7] which
identifies the security vulnerabilities of Drip and proposes
an effective solution.
More importantly, all existing data discovery and dis-
semination protocols [3], [4], [5], [6], [7] employ the cen-
tralized approach in which, as shown in the top sub-figure
in Fig. 1, data items can only be disseminated by the base
station. Unfortunately, this approach suffers from the sin-
gle point of failure as dissemination is impossible when
the base station is not functioning or when the connection
between the base station and a node is broken. In addition,
the centralized approach is inefficient, non-scalable, and
vulnerable to security attacks that can be launched any-
where along the communication path [2]. Even worse,
some WSNs do not have any base station at all. For exam-
ple, for a WSN monitoring human trafficking in a
country’s border or a WSN deployed in a remote area to
monitor illicit crop cultivation, a base station becomes an
attractive target to be attacked. For such networks, data
dissemination is better to be carried out by authorized net-
work users in a distributed manner.
Additionally, distributed data discovery and dissemina-
tion is an increasingly relevant matter in WSNs, especially
in the emergent context of shared sensor networks, where
sensing/communication infrastructures from multiple own-
ers will be shared by applications from multiple users. For
example, large scale sensor networks are built in recent
D. He is with the School of Computer Science and Engineering, South China
University of Technology, Guangzhou 510006, P.R. China, and also with
the College of Computer Science and Technology, Zhejiang University,
Hangzhou 310027, P.R. China. E-mail: hedaojinghit@gmail.com.
S. Chan is with the Department of Electronic Engineering, City University
of Hong Kong, Hong Kong SAR, P. R. China.
E-mail: eeschan@cityu.edu.hk.
M. Guizani is with Qatar University, Qatar. E-mail: mguizani@ieee.org.
H. Yang is with the School of Computer Science and Engineering, Univer-
sity of Electronic Science and Technology of China, P. R. China.
E-mail: haomyang@uestc.edu.cn.
B. Zhou is with the College of Computer Science, Zhejiang University,
P. R. China. E-mail: zby@zju.edu.cn.
Manuscript received 31 Dec. 2013; revised 29 Mar. 2014; accepted 31 Mar.
2014. Date of publication 10 Apr. 2014; date of current version 6 Mar. 2015.
recommended for acceptance by V. B. Misic.
For information on obtaining reprints of this article, please send e-mail to:
reprints@ieee.org, and reference the Digital Object Identifier below.
Digital Object Identifier no. 10.1109/TPDS.2014.2316830
IEEE TRANSACTIONS ON PARALLEL AND DISTRIBUTED SYSTEMS, VOL. 26, NO. 4, APRIL 2015 1129
1045-9219 ß 2014 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.
See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
For More Details Contact G.Venkat Rao
PVR TECHNOLOGIES 8143271457

projects such as Geoss [9], NOPP [10] and ORION [11].
These networks are owned by multiple owners and used by
various authorized third-party users. Moreover, it is
expected that network owners and different users may have
different privileges of dissemination. In this context, distrib-
uted operation by networks owners and users with different
privileges will be a crucial issue, for which efficient solu-
tions are still missing.
Motivated by the above observations, this paper has the
following main contributions:
1) The need of distributed data discovery and dissemi-
nation protocols is not completely new, but previous
work did not address this need. We study the func-
tional requirements of such protocols, and set their
design objectives. Also, we identify the security vul-
nerabilities in previously proposed protocols.
2) Based on the design objectives, we propose DiDrip.
It is the first distributed data discovery and dissem-
ination protocol, which allows network owners and
authorized users to disseminate data items into
WSNs without relying on the base station. More-
over, our extensive analysis demonstrates that
DiDrip satisfies the security requirements of the
protocols of its kind. In particular, we apply the
provable security technique to formally prove
the authenticity and integrity of the disseminated
data items in DiDrip.
3) We demonstrate the efficiency of DiDrip in practice
by implementing it in an experimental WSN with
resource-limited sensor nodes. This is also the first
implementation of a secure and distributed data dis-
covery and dissemination protocol.
The rest of this paper is structured as follows. In Sec-
tion 2, we first survey the existing data discovery and
dissemination protocols, and then discuss their security
weaknesses. Section 3 describes the requirements for a
secure and distributed extension of such protocols. Sec-
tion 4 presents the network, trust and adversary models.
Section 5 describes DiDrip in details. Section 6 provides
theoretical analysis of the security properties of DiDrip.
Section 7 describes the implementation and experimental
results of DiDrip via real sensor platforms. Finally,
Section 8 concludes this paper.
2 SECURITY VULNERABILITIES IN DATA
DISCOVERY AND DISSEMINATION
2.1 Review of Existing Protocols
The underlying algorithm of both DIP and Drip is Trickle
[12]. Initially, Trickle requires each node to periodically
broadcast a summary of its stored data. When a node has
received an older summary, it sends an update to that
source. Once all nodes have consistent data, the broadcast
interval is increased exponentially to save energy. However,
if a node receives a new summary, it will broadcast this
more quickly. In other words, Trickle can disseminate
newly injected data very quickly. Among the existing proto-
cols, Drip is the simplest one and it runs an independent
instance of Trickle for each data item.
In practice, each data item is identified by a unique key
and its freshness is indicated by a version number. For
example, for Drip, DIP and DHV, each data item is repre-
sented by a 3-tuple key; version; data , where key is used
to uniquely identify a data item, version indicates the fresh-
ness of the data item (the larger the version, the fresher the
data), and data is the actual disseminated data (e.g., com-
mand, query or parameter).
2.2 Security Vulnerabilities
An adversary can first place some intruder nodes in the net-
work and then use them to alter the data being disseminated
or forge a data item. This may result in some important
parameters being erased or the entire network being
rebooted with wrong data. For example, consider a new data
item (key, version, data) being disseminated. When an
intruder node receives this new data item, it can broadcast a
malicious data item (key, versionÃ
, dataÃ
), where versionÃ

version. If dataÃ
is set to 0, the parameter identified by key will
be erased from all sensor nodes. Alternatively, if dataÃ
is dif-
ferent from data, all sensor nodes will update the parameter
according to this forged data item. Note that the above
attacks can also be launched if an adversary compromises
some nodes and has access to their key materials.
In addition, since nodes executing Trickle are required to
forward all new data items that they receive, an adversary
can launch denial-of-service (DoS) attacks to sensor nodes
by injecting a large amount of bogus data items. As a result,
the processing and energy resources of nodes are expended
Fig. 1. System overview of centralized and distributed data discovery and dissemination approaches.
1130 IEEE TRANSACTIONS ON PARALLEL AND DISTRIBUTED SYSTEMS, VOL. 26, NO. 4, APRIL 2015

to process and forward these bogus data items, rather than
on the intended functions. Any data discovery and dissemi-
nation protocol based on Trickle or its variants is vulnerable
to such a DoS attack.
3 REQUIREMENTS AND DESIGN CONSIDERATION
A secure and distributed data discovery and dissemination
protocol should satisfy the following requirements:
1) Distributed. Multiple authorized users should be
allowed to simultaneously disseminate data items
into the WSN without relying on the base station.
2) Supporting different user privileges. To provide flexibil-
ity, each user may be assigned a certain privilege
level by the network owner. For example, a user can
only disseminate data items to a set of sensor nodes
with specific identities and/or in a specific localized
area. Another example is that a user just has the priv-
ilege to disseminate data items identified by some
specific keys.
3) Authenticity and integrity of data items. A sensor node
only accepts data items disseminated by authorized
users. Also, a sensor should be able to ensure that
received data items have not been modified during
the dissemination process.
4) User accountability. User accountability must be pro-
vided since bad user behaviors and insider attacks
should be audited and pinpointed. That is, a sender
should not be able to deny the distribution of a data
item. At the same time, an adversary cannot imper-
sonate any legitimate user even if it has compro-
mised the network owner or the other legitimate
users. In many applications, accountability is desir-
able as it enables collection of users’ activities. For
example, from the dissemination record in sensor
nodes, the network owner can find out who dissemi-
nates most data. This requires the sensor nodes to be
able to associate each disseminated data with the
corresponding user’s identity.
5) Node compromise tolerance. The protocol should be
resilient to node compromise attack no matter how
many nodes have been compromised, as long as the
subset of non-compromised nodes can still form a
connected graph with the trusted source.
6) User collusion tolerance. Even if an adversary has com-
promised some users, a benign node should not
grant the adversary any privilege level beyond that
of the compromised users.
7) DoS attacks resistance. The functions of the WSN
should not be disrupted by DoS attacks.
8) Freshness. A node should be able to differentiate
whether an incoming data item is the newest
version.
9) Low energy overhead. Most sensor nodes have limited
resources. Thus, it is very important that the security
functions incur low energy overhead, which can be
decomposed to communication and computation
overhead.
10) Scalability. The protocol should be efficient even for
large-scale WSNs with thousands of sensors and
large user population.
11) Dynamic participation. New sensor nodes and users
can be dynamically added to the network.
In order to ensure security, each step of the existing data
discovery and dissemination protocol runs should be identi-
fied and then protected. In other words, although code dis-
semination protocols may share the same security
requirements as listed above, their security solutions need
to be designed in accordance with their characteristics. Con-
sidering the well known open-source code dissemination
protocol Deluge [1] as an example. Deluge uses an epidemic
protocol based on a page-by-page dissemination strategy
for efficient advertisement of metadata. A code image is
divided into fixed-size pages, and each page is further split
into same-size packets. Due to such a way of decomposing
code images into packets, our proposed protocol is not
applicable for securing Deluge.
The primary challenge of providing security functions in
WSNs is the limited capabilities of sensor nodes in terms of
computation, energy and storage. For example, to provide
authentication function to disseminated data, a commonly
used solution is digital signature. That is, users digitally
sign each packet individually and nodes need to verify the
signature before processing it. However, such an asymmet-
ric mechanism incurs significant computational and com-
munication overhead and is not applicable to sensor nodes.
To address this problem, TESLA and its various extensions
have been proposed [13], [14], which are based on the
delayed disclosure of authentication keys, i.e., the key used
to authenticate a message is disclosed in the next message.
Unfortunately, due to the authentication delay, these mech-
anisms are vulnerable to a flooding attack which causes
each sensor node to buffer all forged data items until the
disclosed key is received.
Another possible approach to authentication is by sym-
metric key cryptography. However, this approach is vulner-
able to node compromise attack because once a node is
compromised, the globally shared secret keys are revealed.
Here we choose digital signatures over other forms for
update packet authentication. That is, the network owner
assigns to each network user a public/private key pair that
allows the user to digitally sign data items and thus authen-
ticates himself/herself to the sensor nodes. We propose two
hybrid approaches to reduce the computation and commu-
nication cost. These methods combine digital signature with
efficient data Merkle hash tree and data hash chain, respec-
tively. The main idea is that signature generation and verifi-
cation are carried out over multiple packets instead of
individual packet. In this way, the computation cost per
packet is significantly reduced. Since elliptic curve cryptog-
raphy (ECC) is computational and communication efficient
compared with the traditional public key cryptography,
DiDrip is based on ECC.
To prevent the network owner from impersonating
users, user certificates are issued by a certificate authority of
a public key infrastructure (PKI), e.g., local police office.
4 NETWORK, TRUST AND THREAT MODELS
4.1 Network Model
As shown in the bottom subfigure in Fig. 1, a general
WSN comprises a large number of sensor nodes. It is
HE ET AL.: SECURE AND DISTRIBUTED DATA DISCOVERY AND DISSEMINATION IN WIRELESS SENSOR NETWORKS 1131

administrated by the owner and accessible by many users.
The sensor nodes are usually resource constrained with
respect to memory space, computation capability, band-
width, and power supply. Thus, a sensor node can only
perform a limited number of public key cryptographic
operations during the lifetime of its battery. The network
users use some mobile devices to disseminate data items
into the network. The network owner is responsible for gen-
erating keying materials. It can be offline and is assumed to
be uncompromisable.
4.2 Trust Model
Networks users are assigned dissemination privileges by
the trusted authority in a PKI on behalf of the network
owner. However, the network owner may, for various rea-
sons, impersonate network users to disseminate data items.
4.3 Threat Model
The adversary considered in this paper is assumed to be
computationally resourceful and can launch a wide range
of attacks, which can be classified as external or insider
attacks. In external attacks, the adversary has no control of
any sensor node in the network. Instead, it would eaves-
drop for sensitive information, inject forged messages,
launch replay attack, wormhole attacks, DoS attacks and
impersonate valid sensor nodes. The communication chan-
nel may also be jammed by the adversary, but this can only
last for a certain period of time after which the adversary
will be detected and removed.
By compromising either network users or sensor nodes,
the adversary can launch insider attacks to the network. The
compromised entities are regarded as insiders because they
are members of the network until they are identified. The
adversary controls these entities to attack the network in arbi-
trary ways. For instance, they could be instructed to dissemi-
nate false or harmful data, launch attacks such as Sybil attacks
or DoS attacks, and be non-cooperative with other nodes.
5 DIDRIP
Referring to the lower sub-figure in Fig. 1, DiDrip consists of
four phases, system initialization, user joining, packet pre-
processing and packet verification. For our basic protocol,
in system initialization phase, the network owner creates its
public and private keys, and then loads the public parame-
ters on each node before the network deployment. In the
user joining phase, a user gets the dissemination privilege
through registering to the network owner. In packet pre-
processing phase, if a user enters the network and wants to
disseminate some data items, he/she will need to construct
the data dissemination packets and then send them to the
nodes. In the packet verification phase, a node verifies each
received packet. If the result is positive, it updates the data
according to the received packet. In the following, each
phase is described in detail. The notations used in the
description are listed in Table 1. The information processing
flow of DiDrip is illustrated in Fig. 2.
5.1 System Initialization Phase
In this phase, an ECC is set up. The network owner carries
out the following steps to derive a private key x and some
public parameters fy; Q; p; q; hð:Þg. It selects an elliptic curve
E over GFðpÞ, where p is a big prime number. Here Q
denotes the base point of E while q is also a big prime number
and represents the order of Q. It then selects the private key
x 2 GFðqÞ and computes the public key y ¼ xQ. After that,
the public parameters are preloaded in each node of the net-
work. We consider 160-bit ECC as an example. In this case, y
and Q are both 320 bits long while p and q are 160 bits long.
5.2 User Joining Phase
This phase is invoked when a user with the identity UIDj,
say Uj, hopes to obtain privilege level. User Uj chooses the
private key SKj 2 GFðqÞ and computes the public key
PKj ¼ SKjÁQ. Here the length of UIDj is set to 2 bytes, in
this case, it can support 65,536 users. Similarly, assume that
160-bit ECC is used, PKj and SKj are 320 bits and 160 bits
long, respectively. Then user Uj sends a 3-tuple UIDj;
Prij; PKj to the network owner, where Prij denotes the
dissemination privilege of user Uj. Upon receiving this mes-
sage, the network owner generates the certificate Certj. A
form of the certificate consists of the following contents:
Certj ¼ fUIDj; PKj; Prij; SIGxfhðUIDjkPKjkPrijÞg, where
the length of Prij is set to 6 bytes, thus the length of Certj is
88 bytes.
TABLE 1
Notations
Fig. 2. Information processing flow in DiDrip.

5.3 Packet Pre-Processing Phase
Assume that a user, say Uj, enters the WSN and wants to
disseminate n data items: di ¼ fkeyi; versioni; dataig, i ¼ 1;
2; . . .; n. For the construction of the packets of the respective
data, we have two methods, i.e., data hash chain and the
Merkle hash tree [15].
For data hash chain approach, a packet, say Pi is com-
posed of packet header, di, and the hash value of packet Piþ1
(i.e., Hiþ1 ¼ hðPiþ1Þ) which is used to verify the next packet,
where i ¼ 1; . . .; n À 1. Here each cryptographic hash Hi is
calculated over the full packet Pi, not just the data portion di,
thereby establishing a chain of hashes. After that, user Uj
uses his/her private key SKj to run an ECDSA sign operation
to sign the hash value of the first data packet hðP1Þ and then
creates an advertisement packet P0, which consists of packet
header, user certificate Certj, hðP1Þ and the signature
SIGSKj fhðP1Þg. Similarly, the network owner assigns a pre-
defined key to identify this advertisement packet.
With the method of Merkle hash tree, user Uj builds a
Merkle hash tree from the n data items in the following way.
All the data items are treated as the leaves of the tree. A new
set of internal nodes at the upper level is formed; each inter-
nal node is computed as the hash value of the concatenation
of two child nodes. This process is continued until the root
node Hroot is formed, resulting in a Merkle hash tree with
depth D ¼ log2ðnÞ. Before disseminating the n data items,
user Uj signs the root node with his/her private key SKj and
then transmits the advertisement packet P0 comprising user
certificate Certj, Hroot and SIGSKj fHrootg. Subsequently, user
Uj disseminates each data item along with the appropriate
internal nodes for verification purpose. Note that as
described above, user certificate Certj contains user identity
information UIDj and dissemination privilege Prij. Before
the network deployment, the network owner assigns a pre-
defined key to identify this advertisement packet.
5.4 Packet Verification Phase
When a sensor node, say Sj, receives a packet either from an
authorized user or from its one-hop neighbours, it first
checks the packet’s key field
1) If this is an advertisement packet (P0 fCertj; hðP1Þ;
SIGSKj fhðP1Þgg for the data hash chain method while
P0 ¼ fCertj; root; SIGSKj frootgg for the Merkle hash
tree method), node Sj first pays attention to the legal-
ity of the dissemination privilege Prij. For example,
node Sj needs to check whether the identity of itself is
included in the node identity set of Prij. If the result
is positive, node Sj uses the public key y of the net-
work owner to run an ECDSA verify operation to
authenticate the certificate. If the certificate Certj is
valid, node Sj authenticates the signature. If yes, for
the data hash chain method (respectively, the Merkle
hash tree method), node Sj stores UIDj; H1
(respectively, UIDj; root ) included in the adver-
tisement packet; otherwise, node Sj simply discards
the packet.
2) Otherwise, it is a data packet Pi, where i ¼ 1; 2; . . . ;
n. Node Sj executes the following procedure:
For the data hash chain method, node Sj checks the
authenticity and integrity of Pi by comparing the hash
value of Pi with Hi which has been received in the
same round and verified. If the result is positive and
the version number is new, node Sj then updates the
data identified by the key stored in Pi and replaces its
stored round, Hi by round, Hiþ1 (here Hiþ1 is
included in packet Pi); otherwise, Pi is discarded.
For Merkle hash tree method, node Sj checks the
authenticity and integrity of Pi through the already
verified root node received in the same round. If the
result is positive and the version number is new, node
Sj then updates the data identified by the key stored
in Pi; otherwise, Pi is discarded.
Remark: To prevent the network owner from impersonat-
ing users, system initialization and issue of user certifi-
cates can be carried out by the certificate authority of a
PKI rather than the network owner.
Comparing the two methods, the data hash chain
method incurs less communication overhead than the Mer-
kle hash tree method. In the data hash chain method, only
one hash value of a packet is included in each packet. On
the contrary, in the Merkle hash tree method, D (the tree
depth) hash values are included in each packet. However, a
limitation of the data hash tree method is that it just works
well in networks with in-sequence packet delivery. Such a
limitation does not exist in the Merkle hash tree method
since it allows each packet to be immediately authenticated
upon its arrival at a node. Therefore, the choice of each
method depends on this characteristic of the WSNs.
5.5 Enhancements
We can enhance the efficiency and security of DiDrip by
adding additional mechanisms. Readers are referred to the
Appendix of this paper for details.
6 SECURITY ANALYSIS OF DIDRIP
In the following, we will analyze the security of DiDrip to
verify that the security requirements mentioned in Section 3
are satisfied.
Distributed. As described in Section 5.2, in order to pass
the signature verification of sensor nodes, each user has to
submit his/her private key and dissemination privilege to
the network owner for registration. In addition, as described
above, authorized users are able to carry out dissemination
in a distributed manner.
Supporting different user privileges. Activities of network
users can be restricted by setting user privilege Prij, which
is contained in the user certificate. Since each user certificate
is generated based on Prij, it will not pass the signature ver-
ification at sensor nodes if Prij is modified. Thus, only the
network owner can modify Prij and then updates the certif-
icate accordingly.
Authenticity and integrity of data items. With the Merkle
hash tree method (rsp. the data hash chain method), an
authorized user signs the root of the Merkle hash tree (rsp.
the hash value of the first data packet hðP1Þ) with his/her pri-
vate key. Using the network owner’s public key, each sensor
node can authenticate the user certificate and obtains the
user’s public key. Then, using the user’s public key, each
node can authenticate the root of the Merkle hash tree (rsp.

the hash value of the first data packet hðP1Þ). Subsequently,
each node can authenticate other data packets based on the
Merkle hash tree (rsp. the data hash chain). With the assump-
tion that the network owner cannot be compromised, it is
guaranteed that any forged or modified data items can be
easily detected by the authentication process.
User accountability. Users’ identities and their dissemina-
tion activities are exposed to sensor nodes. Thus, sensor
nodes can report such records to the network owner period-
ically. Since each user certificate is generated according to
the user identity, except the network owner, no one can
modify the user identity contained in the user certificate
which passes the authentication. Therefore, users cannot
repudiate their activities.
Node compromise and user collusion tolerance. As described
above, for basic protocol, only the public parameters are
preloaded in each node. Even for the improved protocol,
the public-key/dissemination-privilege pair of each net-
work user is loaded into the nodes. Therefore, no matter
how many sensor nodes are compromised, the adversary
just obtains the public parameters and the public-key/dis-
semination-privilege pair of each user. Clearly, the adver-
sary cannot launch any attack by compromising sensor
nodes. As described in Section 5, even if some users collude,
a benign node will not grant any dissemination privilege
that is beyond those of colluding users.
Resistance to DoS attacks. There are DoS attacks against basic
DiDrip by exploiting: (1) authentication delays, (2) the expen-
sive signature verifications, and (3) the Trickle algorithm.
First, with the use of Merkle hash tree or data hash chain,
each node can efficiently authenticate a data packet by a
few hash operations. Second, using the message specific
puzzle approach, each node can efficiently verify a puzzle
solution to filter a fake signature message and to forward a
data packet using Trickle without waiting for signature ver-
ification. Therefore, all the above DoS attacks are defended.
DiDrip can successfully defeat all three types of DoS
attacks even if there are compromised network users and
sensor nodes. Indeed, without the private key and the unre-
leased puzzle keys of the network users, even an inside
attacker cannot forge any signature/data packets.
Ensurance of freshness. If the privilege of a user allows him/
her to disseminate data items to his/her own set of nodes, the
version number in each item can ensure the freshness of
DiDrip. On the other hand, if a node receives data items from
multiple users, the version number can be replaced by a time-
stamp to indicate the freshness of a data item. More specifi-
cally, a timestamp is attached into the root of the Merkle hash
tree (or the hash value of the first data packet).
Scalability. Different from centralized approaches, an
authorized user can enter the network and then dissemi-
nates data items into the targeted sensor nodes. Moreover,
as to be demonstrated by our experiments in a testbed with
24 TelosB motes in the next section, the security functions in
our protocol have low impact on propagation delay. Note
that the increase in propagation delay is dominated by the
signature verification time incurred at the one-hop neigh-
boring nodes of the authorized user. Thus, the proposed
protocol is efficient even in a large-scale WSN with thou-
sands of sensor nodes. Also, as shown in Section 5.2, our
protocol can support a large number of users.
Also, as described above, DiDrip can achieve dynamic par-
ticipation. Moreover, in the next section, our implementation
results will demonstrate DiDrip has low energy overhead.
In the following, we give the formal proof of the authen-
ticity and integrity of the disseminated data items in DiDrip
based on the three assumptions below:
Assumption 1: There exist pseudo-random functions
which are polynomially indistinguishable from truly ran-
dom functions.
Assumption 2: There exist target collision-resistance
(TCR) hash functions [16], where if for all probabilistic-poly-
nomial-time (PPT) adversaries, say A, A have negligible
probability in winning the following game: A first choose a
message m, and then A are given a random function hð:Þ. To
win, A must output m
0
6¼ m such that hðm
0
Þ ¼ hðmÞ. Note
that in our scheme, the TCR hash function can be imple-
mented by the common hash functions, such as SHA-1.
Assumption 3: ECC signature is existentially unforge-
able under adaptive chosen-message attacks. Note that
in our scheme, ECC signature can use the standard
ECDSA of 160 bits.
Theorem 1. DiDrip achieves the authenticity and integrity of data
items, assuming the indistinguishability between pseudo-ran-
domness and true randomness, and assuming that hð:Þ is a
TCR hash function and ECC signature is existentially unforge-
able under adaptive chosen-message attacks.
Proof. Our theorem follows from Theorem A.1 in [16], and
thus here we only give a proof sketch briefly. To begin
with, we assume ECC signature generation and verifica-
tion guarantee authenticity and integrity of signed mes-
sages, and every receiver has obtained an authentic copy
of the legitimate sender’s public key. And we also assume
that hð:Þ is a TCR hash function. Therefore, here the secu-
rity of DiDrip is proven based on the indistinguishability
between pseudo-randomness and true randomness.
First, there exists a PPT adversary A, which can defeat
authenticity of data items in DiDrip. This means that A
controls the communication links and manages, with
non-negligible probability, to deliver a message m to a
receiver R, such that the sender S has not sent m but R
accepts m as authentic and coming from S. Then there is
a PPT adversary B that uses A to break the indistinguish-
ability between pseudo-randomness and true random-
ness with non-negligible advantage. That is, B gets
access to h (as an oracle) and can tell with non-negligible
probability if h is a pseudorandom function (PRFð:Þ) or
if h is a totally random function.
To this end, B can query on inputs x of its choice and be
answered with hðxÞ. Hence first B simulates for A a net-
work with a sender S and a receiver R. Then B works by
running A in the way similar to that in [17]. Namely, B
chooses a number l 2 f1; . . .; ng at random, where n is the
total number of packets to be sent in the data dissemina-
tion. Note that B hopes that A will forge the lth packet Pl.
B grants access to the oracle h, which is either PRFð:Þ
or an ideal random function. B can adaptively query an
arbitrarily chosen x to the oracle and get the output
which is either PRFðxÞ or a random value uniformly
selected from f0; 1gÃ
. After performing polynomially
many queries, B finally makes the decision of whether or

not the oracle is PRFð:Þ or the ideal random function. As
a result, B wins the game if the decision is correct.
We argue that B succeeds with non-negligible proba-
bility sketchily. If h is a truly random function then A has
only negligible probability to successfully forge the
packet Pl in the data items. Therefore, if h is random then
B makes the wrong decision only with negligible proba-
bility. On the other hand, we have assumed that if the
authentication is done using PRFð:Þ then A forges some
packet with non-negligible probability . It follows that if
h is PRFð:Þ then B makes the right decision with proba-
bility at least =l (which is also non-negligible).
In addition, if the adversary A is able to cause a
receiver R to accept a forged packet Pl, it implies the
adversary A is able to find a collision on hðP
0
l Þ ¼ hðPlÞ in
the packet PlÀ1. However, according to Assumption 2,
hð:Þ is a TCR hash function. Moreover, due to the unfor-
geability of ECC signature (Assumption 3), it is impossi-
ble that A hands R a forged initial packet from S (For
data hash chain approach, A forges the signature
SIGSKj ðP1Þ; For Merkle hash tree method, A forges the
signature SIGSKj ðrootÞ. Therefore, the above contradic-
tions means also that the authenticity and integrity of
data items in DiDrip. tu
7 IMPLEMENTATION AND PERFORMANCE
EVALUATION
We evaluate DiDrip by implementing all components on an
experimental test-bed. Also, we choose Drip for perfor-
mance comparison.
7.1 Implementation and Experimental Setup
We have written programs that execute the functions of the
network owner, user and sensor node. The network owner
and user side programs are C programs using OpenSSL [18]
and running on laptop PCs (with 2 GB RAM) under Ubuntu
11.04 environment with different computational power.
Also, the sensor node side programs are written in nesC
and run on resource-limited motes (MicaZ and TelosB). The
MicaZ mote features an 8-bit 8-MHz Atmel microcontroller
with 4-kB RAM, 128-kB ROM, and 512 kB of flash memory.
Also, the TelosB mote has an 8-MHz CPU, 10-kB RAM, 48-
kB ROM, 1MB of flash memory, and an 802.15.4/ZigBee
radio. Our motes run TinyOS 2.x. Additionally, SHA-1 is
used, and the key sizes of ECC are set to 128 bits, 160 and
192 bits, respectively. Throughout this paper, unless
otherwise stated, all experiments on PCs (respectively, sen-
sor nodes) were repeated 100,000 times (respectively,
1,000 times) for each measurement in order to obtain accu-
rate average results.
To implement DiDrip with the data hash chain method
(rsp. the Merkle hash tree method), the following function-
alities are added to the user side program of Drip: con-
struction of data hash chain (rsp. Merkle hash tree) of a
round of dissemination data, generation of the signature
packet and all data packets. For obtaining version number
of each data item, the DisseminatorC and DisseminatorP
modules in the Drip nesC library has been modified to
provide an interface called DisseminatorVersion. More-
over, the proposed hash tree method is implemented with-
out and with using the message specific puzzle approach
presented in Appendix, resulting in two implementations
of DiDrip; DiDrip1 and DiDrip2. In DiDrip1, when a node
receives a signature/data packet with a new version num-
ber, it authenticates the packet before broadcasting it to its
next-hop neighbours. On the other hand, in DiDrip2, a
node only checks the puzzle solution in the packet before
broadcasting the packets. We summarize the pros and
cons of all related protocols in Table 2.
Based on the design of DiDrip, we implement the veri-
fication function for signature and data packets based on
the ECDSA verify function and SHA-1 hash function of
TinyECC 2.0 library [19] and add them to the Drip nesC
library. Also, in our experiment, when a network user
(i.e., a laptop computer) disseminates data items, it first
sends them to the serial port of a specific sensor node in
the network which is referred to as repeater. Then, the
repeater carries out the dissemination on behalf of the
user using DiDrip.
Similar to [7], we use a circuit to accurately measure the
power consumption of various cryptographic operations
executed in a mote. The Tektronix TDS 3034C digital oscillo-
scope accurately measures the voltage Vr across the resistor.
Denoting the battery voltage as Vb (which is 3 volts in our
experiments), the voltage across the mote Vm is then Vb À Vr.
Once Vr is measured, the current through the circuit I can
be obtained by using Ohm’s law. The power consumed by
the mote is then VmI. By also measuring the execution time
of the cryptographic operation, we can obtain the energy
consumption of the operation by multiplying the power
and execution time.
7.2 Evaluation Results
The following metrics are used to evaluate DiDrip; memory
overhead, execution time of cryptographic operations and
propagation delay, and energy overhead. The memory
overhead measures the required data space in the imple-
mentation. The propagation delay is defined as the time
from construction of a data hash chain until the parameters
on all sensor nodes corresponding to a round of dissemi-
nated data items are updated.
TABLE 2
The Pros and Cons of All Related Protocols

Table 3 shows the execution times of some important
operations in DiDrip. For example, the execution times for
the system initialization phase and signing a random
20-byte message (i.e., the output of SHA-1 function) are
1.608 and 0.6348 ms on a 1.8-GHz Laptop PC, respectively.
Thus, if SHA-1 is used, generating a user certificate or sign-
ing a message takes 0.6348 ms on a 1.8-GHz Laptop PC.
Fig. 3 shows the execution times of SHA-1 hash function
(extracted from TinyECC 2.0 [19]) on MicaZ and TelosB
motes. The inputs to the hash function are randomly gener-
ated numbers with length varying from 24 to 156 bytes in
increments of 6 bytes. Note that, in our protocol, the hash
function is applied to an entire packet. There are several rea-
sons that, possibly, a packet contains a few tens of bytes.
First, the advertisement packet has additional information
such as certificate and signature. Second, with the Merkle
hash tree method, each packet contains the disseminated
data item along with the related internal nodes of the tree
for verification purpose. Third, although several bytes is a
typical size of a data item, sometimes a disseminated data
item may be a bit larger. Moreover, for sensors with IEEE
802.15.4 compliant radios, the maximum payload size is 102
bytes for each packet. Therefore, we have chosen a wider
range of input size to SHA-1 to provide readers a more com-
plete picture of the performance. We perform the same
experiment 10,000 times and take an average over them. For
example, the execution times on a MicaZ mote for inputs of
54 bytes, 114 bytes, and 156 bytes are 9.6788, 18.947, and
28.0515 ms, respectively. Also, the execution times on a
TelosB mote for inputs of 54, 114, and 156 bytes are 5.7263,
10.7529, and 15.629 ms, respectively.
To measure the execution time of public key cryptogra-
phy, as shown in Table 41
, we have implemented the ECC
verification operation (with a random 20-byte number as
the output) of TinyECC 2.0 library [19] on MicaZ and TelosB
motes. For example, it is measured that the signature verifi-
cation times are 2.436 and 3.955 seconds, which are 252 and
691 times longer than SHA-1 hash operation with a 54-byte
random number as input on MicaZ and TelosB motes,
respectively. It can be seen packet authentication based on
the Merkle hash tree (or data hash chain) is much more effi-
cient. Therefore, it is confirmed that DiDrip is suitable for
sensor nodes with limited resources.
Next, we compare the energy consumption of SHA-1 hash
function and ECC verification under the condition that the
radio of the mote is turned off. When a MicaZ mote is used in
the circuit, Vr ¼ 138 mV, I ¼ 6.7779 mA, Vm ¼ 2.8620 V,
P ¼ 19.3983 mW. When a TelosB mote is used, Vr ¼ 38 mV,
I ¼ 1.8664 mA, Vm ¼ 2.9620 V, P ¼ 5.5283 mW. With the
execution time obtained from Fig. 3, the energy consumption
on the motes due to the SHA-1 operation can be determined.
For example, the energy consumption of SHA-1 operation
with a random 54-byte number as input on MicaZ and
TelosB motes are 0.18775 and 0.03166 mJ, respectively. Also,
the energy consumption of ECC signature verification
operation on MicaZ and TelosB motes are 2835.2555 and
1316.1777 mJ, respectively.
Next, the impact of security functions on the propagation
delay is investigated in an experimental network as shown
in Fig. 4. The network has 24 TelosB nodes arranged in a
4 Â 6 grid. The distance between each node is about 35 cm,
TABLE 3
Running Time for Each Phase of the Basic Protocol of DiDrip (Except the Sensor Node Verification Phase)
Fig. 3. The execution times of SHA-1 hash function on MicaZ and TelosB
motes.
TABLE 4
Running Time for ECC Signature Verification
Fig. 4. The 4 Â 6 grid network of TelosB motes for measuring propaga-
tion delay.
1. Note that ECC-160 is faster than ECC-128, because the column
width of ECC-160 is set to 5 for hybrid multiplication optimization
while that of ECC-128 is set to 4.

and the transmission power is configured to be the lowest
level so that only one-hop neighbours are covered in the
transmission range. The repeater is acted by the node locat-
ing at the vertex of the grid.
In the experiments, the packet delivery rate from the net-
work user is 5 packets/s. The lengths of round and data
fields in a data item are set to 4 bits and 2 bytes, respec-
tively. A hash function with 8-byte truncated output is used
to construct data hash chains. An ECC-160 signature is 40
bytes long. Each experiment is repeated 20 times to obtain
an average measurement. Figs. 5 and 6 plot the average
propagation delays of Drip, DiDrip1, and DiDrip2 when the
data hash chain and Merkle hash tree methods are
employed, respectively. It can be seen that the propagation
delay almost increases linearly with the number of data
items per round for all three protocols. Moreover, the secu-
rity functions in DiDrip2 have low impact on propagation
delay. For these five experiments of the data hash chain
method, DiDrip2 is just 3.448, 4.158, 3.222, 3.919 and 2.855 s
more than that of Drip, respectively. Note that the increase
in propagation delay is dominated by the signature verifica-
tion time incurred at the one-hop neighboring nodes of the
base station. This is because each node carries out signature
verification only after forwarding data packets (with valid
puzzle solutions).
Table 5 shows the memory (ROM and RAM) usage of
DiDrip2 (with the data hash chain method) on MicaZ and
TelosB motes for the case of four data items per round. The
code size of Drip and a set of verification functions from
TinyECC (secp128r1, secp160r1 and secp192r1, which are
implementations based on various elliptic curves according
to the Standards for Efficient Cryptography Group) are
included for comparison. For example, the size of DiDrip
implementation corresponds to 26.18 and 56.82 percent of the
RAM and ROM capacities of TelosB, respectively. Clearly, the
ROM and RAM consumption of DiDrip is more than that of
Drip because of the extra security functions. Moreover, it can
be seen that majority of the increased ROM is due to TinyECC.
8 CONCLUSION AND FUTURE WORK
In this paper, we have identified the security vulnerabil-
ities in data discovery and dissemination when used in
WSNs, which have not been addressed in previous
research. Also, none of those approaches support distrib-
uted operation. Therefore, in this paper, a secure and dis-
tributed data discovery and dissemination protocol
named DiDrip has been proposed. Besides analyzing the
security of DiDrip, this paper has also reported the evalu-
ation results of DiDrip in an experimental network of
resource-limited sensor nodes, which shows that DiDrip
is feasible in practice. We have also given a formal proof
of the authenticity and integrity of the disseminated data
items in DiDrip. Also, due to the open nature of wireless
channels, messages can be easily intercepted. Thus, in the
future work, we will consider how to ensure data confi-
dentiality in the design of secure and distributed data
discovery and dissemination protocols.
APPENDIX
FURTHER IMPROVEMENT OF DIDRIP SECURITY AND
EFFICIENCY
By the basic DiDrip protocol, we can achieve secure and dis-
tributed data discovery and dissemination. To further
enhance the protocol, here we propose two modifications to
Fig. 5. Propagation delay comparison of three protocols when the data
hash chain method is employed.
Fig. 6. Propagation delay comparison of three protocols when the Merkle
hash tree method is employed.
TABLE 5
Code Sizes (Bytes) on MicaZ and TelosB Motes

improve the efficiency and security of DiDrip. For brevity,
only those parts of the basic protocol that require changes
will be presented.
Avoiding the Generation, Transmission and
Verification of Certificates
There are some efficiency problems caused by the genera-
tion, transmission, and verification of certificates. First, it is
not efficient in communication, as the certificate has to be
transmitted along with the advertisement packet across
every hop as the message propagates in the WSN. A large
per-message overhead will result in more energy consump-
tion on each sensor node. Second, to authenticate each adver-
tisement packet, it always takes two expensive signature
verification operations because the certificate should always
be authenticated first. To address these challenges, a feasible
approach is that before the network deployment, the public-
key/dissemination-privilege pair of each network user is
loaded into the sensor nodes by the network owner. Once a
new user joins the network after the network deployment,
the network owner can notify the sensor nodes of the user’s
public key/dissemination privilege through using the pri-
vate key of itself. The detailed description is as follows.
User Joining Phase
According to the basic protocol of DiDrip, user Uj generates
its public and private keys and sends a 3-tuple
UIDj; Prij; PKj to the network owner. When the net-
work owner receives the 3-tuple, it no longer generates the
certificate Certj. Instead, it signs the 3-tuple with its private
key and sends it to the sensor nodes. Finally, each node
stores the 3-tuple.
Packet Pre-Processing Phase
The user certificate Certj stored in packet P0 is replaced by
UIDj.
Packet Verification Phase
If this is an advertisement packet, according to the received
identity UIDj, node Sj first picks up the dissemination priv-
ilege Prij from its storage and then pays attention to the
legality of Prij. If the result is positive, node Sj uses the pub-
lic key PKj from its storage to run an ECDSA verify opera-
tion to authenticate the signature; otherwise, node Sj
simply discards the packet. Note that node Sj does not need
to authenticate the certificate.
As described above, the public-key/dissemination-privi-
lege pair UIDj; Prij; PKj of each network user is just
2 þ 6 þ 40 ¼ 48 bytes. Therefore, assuming the protocol
supports 500 network users, the code size is about 23 KB.
We consider the resource-limited sensor nodes such as
TelosB motes as examples. The 1-MB Flash memory is
enough for storing these public parameters.
Message Specific Puzzle Approach for Resistance to
DoS attacks
DiDrip uses a digital signature to bootstrap the authentica-
tion of a round of data discovery and dissemination. This
authentication is vulnerable to DoS attacks. That is, an
adversary may flood a lot of illegal signature message (i.e.,
advertisement messages in this paper) to the sensor nodes to
exhaust their resources and render them less capable of proc-
essing the legitimate signature messages. Such an attack can
be defended by applying the message specific puzzle
approach [2]. This approach requires each signature message
to contain a puzzle solution. When a node receives a signa-
ture message, it first checks that the puzzle solution is correct
before verifying the signature. There are two characteristics
of the puzzles. First, the puzzles are difficult to be solved but
their solutions are easy to be verified. Second, there is a tight
time limit to solve a puzzle. This discourages adversaries to
launch the DoS attack even if they are computationally pow-
erful. More details about this approach can be found in [2].
Another advantage of applying the message specific puz-
zle is to reduce the dissemination delay, which is the time for
a disseminated packet to reach all nodes in a WSN. Recall
that in step 1.a) of the packet verification phase, when a node
receives the signature packet, it first carries out the signature
verification before using the Trickle algorithm to broadcast
the signature packet. This means that the dissemination
delay depends on the signature verification time tsv. On the
other hand, when the message specific puzzle approach is
applied, a node can just verify the validity of puzzle solution
before broadcasting the signature packet. Then, the dissemi-
nation delay only depends on the puzzle solution verifica-
tion time tpv. Since tpv ( tsv, the dissemination delay is
significantly reduced. Moreover, the reduction in dissemina-
tion delay is proportional to the network size. This is demon-
strated by the experiments presented in Section 7.2.
ACKNOWLEDGMENTS
This research is supported by a strategic research grant from
City University of Hong Kong [Project No. 7004054], the
Fundamental Research Funds for the Central Universities,
and the Specialized Research Fund for the Doctoral Pro-
gram of Higher Education. D. He is the corresponding
author of this article.
REFERENCES
[1] J. W. Hui and D. Culler, “The dynamic behavior of a data dissemi-
nation protocol for network programming at scale,” in Proc. 2nd
Int. Conf. Embedded Netw. Sensor Syst., 2004, pp. 81–94.
[2] D. He, C. Chen, S. Chan, and J. Bu, “DiCode: DoS-resistant and
distributed code dissemination in wireless sensor networks,”
IEEE Trans. Wireless Commun., vol. 11, no. 5, pp. 1946–1956, May
2012.
[3] T. Dang, N. Bulusu, W. Feng, and S. Park, “DHV: A code consistency
maintenance protocol for multi-hop wireless sensor networks,” in
Proc. 6th Eur. Conf. Wireless Sensor Netw., 2009, pp. 327–342.
[4] G. Tolle and D. Culler, “Design of an application-cooperative
management system for wireless sensor networks,” in Proc. Eur.
Conf. Wireless Sensor Netw., 2005, pp. 121–132.
[5] K. Lin and P. Levis, “Data discovery and dissemination with
DIP,” in Proc. ACM/IEEE Int. Conf. Inf. Process. Sensor Netw., 2008,
pp. 433–444.
[6] M. Ceriotti, G. P. Picco, A. L. Murphy, S. Guna, M. Corra, M.
Pozzi, D. Zonta, and P. Zanon, “Monitoring heritage buildings
with wireless sensor networks: The Torre Aquila deployment,” in
Proc. IEEE Int. Conf. Inf. Process. Sensor Netw., 2009, pp. 277–288.
[7] D. He, S. Chan, S. Tang, and M. Guizani, “Secure data discovery
and dissemination based on hash tree for wireless sensor
networks,” IEEE Trans. Wireless Commun., vol. 12, no. 9, pp. 4638–
4646, Sep. 2013.

[8] M. Rahman, N. Nasser, and T. Taleb, “Pairing-based secure tim-
ing synchronization for heterogeneous sensor networks,” in Proc.
IEEE Global Telecommun. Conf., 2008, pp. 1–5.
[9] Geoss. [Online]. Available: http://www.epa.gov/geoss/
[10] NOPP. [Online]. Available: http://www.nopp.org/
[11] ORION. [Online]. Available: http://www.joiscience.org/ocean
observing/advisors
[12] P. Levis, N. Patel, D. Culler, and S. Shenker, “Trickle: A self-regu-
lating algorithm for code maintenance and propagation in wire-
less sensor networks,” in Proc. 1st Conf. Symp. Netw. Syst. Design
Implementation, 2004, pp. 15–28.
[13] A. Perrig, R. Canetti, D. Song, and J. Tygar, “Efficient and secure
source authentication for multicast,” in Proc. Netw. Distrib. Syst.
Security Symp., 2001, pp. 35–46.
[14] Y. Chen, I. Lin, C. Lei, and Y. Liao, “Broadcast authentication in
sensor networks using compressed bloom filters,” in Proc. 4th
IEEE Int. Conf. Distrib. Comput. Sensor Syst., 2008, pp. 99–111.
[15] R. Merkle, “Protocols for public key cryptosystems,” in Proc. IEEE
Security Privacy, 1980, pp. 122–134.
[16] M. Bellare and P. Rogaway, “Collision-resistant hashing: Towards
making UOWHFs practical,” in Proc. Adv. Cryptology, 1997,
pp. 56–73.
[17] A. Perrig, R. Canetti, J. Tygar, and D. Song, “Efficient authentica-
tion and signing of multicast streams over lossy channels,” in
Proc. IEEE Security Privacy, 2000, pp. 56–73.
[18] OpenSSL. [Online]. Available: http://www.openssl.org
[19] A. Liu and P. Ning, “TinyECC: A configurable library for elliptic
curve cryptography in wireless sensor networks,” in Proc. ACM/
IEEE Inf. Process. Sensor Netw., 2008, pp. 245–256.
Daojing He (S’07-M’13) received the BEng
and MEng degrees from the Harbin Institute of
Technology, China, and the PhD degree from
Zhejiang University, China, all in computer
science in 2007, 2009, and 2012, respectively.
He is with the School of Computer Science and
Engineering, South China University of Technol-
ogy, P.R. China, and also with the College of
Computer Science and Technology, Zhejiang
University, P.R. China. His research interests
include network and systems security. He is an
associate editor or on the editorial board of some international journals
such as IEEE Communications Magazine, Springer Journal of Wireless
Networks, Wiley’s Wireless Communications and Mobile Computing
Journal, Journal of Communications and Networks, Wiley’s Security and
Communication Networks Journal, and KSII Transactions on Internet
and Information Systems. He is a member of the IEEE.
Sammy Chan (S’87-M’89) received the BE and
MEngSc degrees in electrical engineering from
the University of Melbourne, Australia, in 1988 and
1990, respectively, and the PhD degree in com-
munication engineering from the Royal Melbourne
Institute of Technology, Australia, in 1995. From
1989 to 1994, he was with Telecom Australia
Research Laboratories, first as a research engi-
neer, and between 1992 and 1994 as a senior
research engineer and project leader. Since
December 1994, he has been with the Department
of Electronic Engineering, City University of Hong Kong, where he is cur-
rently an associate professor. He is a member of the IEEE.
Mohsen Guizani (S’85-M’89-SM’99-F’09)
received the BS (with distinction) and MS
degrees in electrical engineering, the MS and
PhD degrees in computer engineering in 1984,
1986, 1987, and 1990, respectively, from Syra-
cuse University, Syracuse, New York. He is cur-
rently a professor and the associate vice
president for Graduate Studies at Qatar Univer-
sity, Qatar. His research interests include com-
puter networks, wireless communications and
mobile computing, and optical networking. He
currently serves on the editorial boards of six technical journals and
the founder and EIC of “Wireless Communications and Mobile
Computing” Journal published by John Wiley (http://www.interscience.
wiley.com/jpages/1530-8669/). He is a fellow of the IEEE and a senior
member of ACM.
Haomiao Yang (M’12) received the MS and PhD
degrees in computer applied technology from the
University of Electronic Science and Technology
of China (UESTC) in 2004 and 2008, respec-
tively. From 2012 to 2013, he worked as a post-
doctoral fellow at Kyungil University, Republic of
Korea. Currently, he is an associate professor at
the School of Computer Science and Engineer-
ing, UESTC, China. His research interests
include cryptography, cloud security, and big
data security. He is a member of the IEEE.
Boyang Zhou is currently working toward the
PhD degree from the College of Computer Sci-
ence at Zhejiang University. His research areas
include software-defined networking, future
internet architecture and flexible reconfigurable
networks.
For more information on this or any other computing topic,
please visit our Digital Library at www.computer.org/publications/dlib.

Secure and distributed data discovery and dissemination in wireless sensor networks

More Related Content

What's hot (16)

Similar to Secure and distributed data discovery and dissemination in wireless sensor networks (20)

More from Pvrtechnologies Nellore (20)

Recently uploaded (20)

Secure and distributed data discovery and dissemination in wireless sensor networks