A REFERENCE BASED, TREE STRUCTURED TIME SYNCHRONIZATION APPROACH AND ITS ANALYSIS IN WSN

International Journal of Ad hoc, Sensor & Ubiquitous Computing (IJASUC) Vol.2, No.1, March 2011
DOI : 10.5121/ijasuc.2011.2103 20
A REFERENCE BASED, TREE STRUCTURED TIME
SYNCHRONIZATION APPROACH AND ITS ANALYSIS
IN WSN
Surendra Rahamatkar1
and Dr. Ajay Agarwal2
1
Dept. of Computer Sc., Nagpur Institute of Technology, Nagpur, India
rahamatkar_s@rediffmail.com,
2
Dept. of MCA, Krishna Inst. of Engg. & Technology, Ghaziabad, India
ajay.aagar@gmail.com
ABSTRACT
Time synchronization for wireless sensor networks (WSNs) has been studied in recent years as a
fundamental and significant research issue. Many applications based on these WSNs assume local clocks
at each sensor node that need to be synchronized to a common notion of time. Time synchronization in a
WSN is critical for accurate time stamping of events and fine-tuned coordination among the sensor nodes
to reduce power consumption. This paper proposes a bidirectional, reference based, tree structured time
synchronization service for WSNs along with network evaluation phase. This offers a push mechanism for
(i) accurate and (ii) low overhead for global time synchronization. Analysis study of proposed approach
shows that it is lightweight as the number of required broadcasting messages is constant in one
broadcasting domain.
KEYWORDS
Ad Hoc Tree Structure; Clock synchronization; Wireless sensor networks, Hierarchical sensor network.
1. INTRODUCTION
Wireless sensor networks (WSNs) can be applied to a wide range of applications in domains as
diverse as medical, industrial, military, environmental, scientific, and home networks [5]. Since
the sensors in a WSN operate independently, their local clocks may not be synchronized with
one another. This can cause difficulties when trying to integrate and interpret information
sensed at different nodes. For instance, if a moving car is detected at two different times along a
road, before we can even tell in what direction the car is going, the detection times have to be
compared meaningfully. In addition, we must be able to transform the two time readings into a
common frame of reference before estimating the speed of the vehicle. Estimating time
differences across nodes accurately is also important in node localization. For example, many
localization algorithms use ranging technologies to estimate inter-nodes distances; in these
technologies, synchronization is needed for time-of-flight measurements that are then
transformed into distances by multiplying with the medium propagation speed for the type of
signal used such as radio frequency or ultrasonic. There are additional examples where
cooperative sensing requires the nodes involved to agree on a common time frame such as
configuring a beam-forming array and setting a TDMA (Time Division Multiple Access) radio
schedule [6]. These situations mandate the necessity of one common notion of time in WSNs.
Therefore, currently there is a huge research interest towards developing efficient clock
synchronization protocols to provide a common notion of time.
Time synchronization of WSNs is crucial to maintain data consistency, coordination, and
perform other fundamental operations. Further, synchronization is considered a critical problem

21
for wireless ad hoc networks due to its de-centralized nature and the timing uncertainties
introduced by the imperfections in hardware oscillators and message de-lays in both physical
and medium access control (MAC) layers. All these uncertainties cause the local clocks of
different nodes to drift away from each other over the course of a time interval.
The primary functionality of wireless sensor networks is to sense the environment and transmit
the acquired information to base stations for further processing with secure time information
[3]. The clock synchronization problem has been studied thoroughly in the areas of Internet and
local area networks (LANs) for the last several decades. Many existing synchronization
algorithms rely on the clock information from Global Positioning System (GPS). However,
GPS-based clock acquisition schemes exhibit some weaknesses: GPS is not ubiquitously
available and requires a relatively high-power receiver, which is not possible in tiny and cheap
sensor nodes. This is the motivation for developing software-based approaches to achieve in-
network time synchronization. Among many protocols that have been devised for maintaining
synchronization, Network Time Protocol (NTP) [7] is outstanding owing to its ubiquitous
deployment, scalability, robustness related to failures, and self-configuration in large multi-hop
networks. Moreover, the combination of NTP and GPS has shown that it is able to achieve high
accuracy on the order of a few microseconds [8]. However, NTP is not suitable for a wireless
sensor environment, since WSNs pose numerous challenges of their own; to name a few,
limited energy and bandwidth, limited hardware, latency, and unstable network conditions
caused by mobility of sensors, dynamic topology, and multi-hopping. The most of the time
synchronization protocols differ broadly in terms of their computational requirements, energy
consumption, precision of synchronization results, and communication requirements [1].
In the paper, we propose a more effective, lightweight multi-hop tree structured referencing
time synchronization (TSRT) approach with the goal of achieving a long-term network-wide
synchronization with minimal Message Exchanges and exhibits a number of attractive features
such as highly scalable and lightweight.
The whole paper is organized in six Sections. In Section 2, existing synchronization schemes
are reviewed. Proposed reference based tree structured time synchronization scheme is
explained in Section 3. Section 4 contains the network evaluation phase, followed by the
comparison of proposed scheme with existing work in Section 5. Finally, Section 6 contains the
conclusion of the paper.
2. EXISTING APPROACHES TO TIME SYNCHRONIZATION
Time synchronization algorithms providing a mechanism to synchronize the local clocks of the
nodes in the network have been extensively studied in the past. The most widely adapted
protocol used in the internet domain is the NTP devised by Mills [7]. The NTP clients
synchronize their clocks to the time servers with accuracy in the order of milliseconds by
statistical analysis of the round-trip time. The time servers are synchronized by external time
sources, typically using GPS. The NTP has been widely deployed and proved to be effective,
secure and robust in the internet. In WSNs, however, non-determinism in transmission time
caused by the Media Access Channel (MAC) layer of the radio stack can introduce several
hundreds of milliseconds delay at each hop. Therefore, without further adaptation, NTP is
suitable only for WSN applications with low precision demands.
Two of the most prominent examples of existing time synchronization protocols developed for
the WSN domain are the Reference Broadcast Synchronization (RBS) algorithm [9] and the
Timing-sync Protocol for Sensor Networks (TPSN) [10].
In RBS, a reference message is broadcasted. The receivers record their local time when
receiving the reference broadcast and exchange the recorded times with each other. The main

22
advantage of RBS is that it eliminates transmitter-side non-determinism. The disadvantage of
the approach is that additional message exchange is necessary to communicate the local time-
stamps between the nodes. In the case of multi hop synchronization, the RBS protocol would
lose its accuracy. Santashil PalChaudhuri et al [15] extended the RBS protocol to handle multi
hop clock synchronization in which all nodes need not be within single-hop range of a clock
synchronization sender.
The TPSN algorithm first creates a spanning tree of the network and then performs pair wise
synchronization along the edges. Each node gets synchronized by exchanging two
synchronization messages with its reference node one level higher in the hierarchy. The TPSN
achieves two times better performance than RBS by time-stamping the radio messages in the
MAC layer of the radio stack and by relying on a two-way message exchange. The shortcoming
of TPSN is that it does not estimate the clock drift of nodes, which limits its accuracy, and does
not handle dynamic topology changes.
TinySeRSync [16] protocol works with the ad hoc deployments of sensor networks. This
protocol proposed two asynchronous phases: Phase I –secure single-hop pair wise
synchronization, and Phase II–secure andresilient global synchronization to achieve global time
synchronization in a sensor network.
Van Greunen et al [18] Lightweight Tree-based Synchronization (LTS) protocol is a slight
variation of the network-wide synchronization protocol of Ganeriwal et al. [17]. Similar to
network-wide synchronization the main goal of the LTS protocol is to achieve reasonable
accuracy while using modest computational resources. As with network-wide synchronization,
the LTS protocol seeks to build a tree structure within the network. Adjacent tree nodes
exchange synchronization information with each other. A disadvantage is that the accuracy of
synchronization decreases linearly in the depth of the synchronization tree (i.e., the longest path
from the node that initiates synchronization to a leaf node). Authors discuss various ideas for
limiting the depth of tree; the performance of protocol is analyzed with simulations.
A survey study of Abolfazl Akbari et al [4] suggested as data communication and various
network operations cause energy depletion in sensor nodes and therefore, it is common for
sensor nodes to exhaust its energy completely and stop operating. This may cause connectivity
and data loss. Therefore, it is necessary that network failures are detected in advance and
appropriate measures are taken to sustain network operation. The clock synchronization
protocols significantly differ from the conventional protocols in dealing the challenges specific
to WSNs. It is quite likely that the choice of a protocol will be driven by the characteristics and
requirements of each application.
3. TREE STRUCTURED REFERENCING TIME SYNCHRONIZATION
APPROACH
In this Section we proposed Tree Structured Referencing Time Synchronization (TSRT)
scheme, which is based on the protocol, proposed by [2], that the aim is to minimize the
complexity of the synchronization. Thus the needed synchronization accuracy is assumed to be
given as a constraint, and the target is to devise a synchronization algorithm with minimal
complexity to achieve given precision. TSRT works on two phases. First phase used to
construct an ad hoc tree structure and second phase used to synchronize the local clocks of
sensor nodes followed by network evaluation phase.
The goal of the TSRT is to achieve a network wide synchronization of the local clocks of the
participating nodes. We assume that each node has a local clock exhibiting the typical timing
errors of crystals and can communicate over an unreliable but error corrected wireless link to its
neighbors. The TSRT synchronizes the time of a sender to possibly multiple receivers utilizing a

23
single radio message time-stamped at both the sender and the receiver sides. MAC layer time-
stamping can eliminate many of the errors, as observed in [11]. However, accurate time-
synchronization at discrete points in time is a partial solution only. Compensation for the clock
drift of the nodes is inevitable to achieve high precision in-between synchronization points and
to keep the communication overhead low. Linear regression is used in TSRT to compensate for
clock drift as suggested in [9].
3.1. Main Ideas
The proposed synchronization approach is flexible and self-organized. A physical broadcast
channel is required, which is automatically satisfied by the wireless medium. A connected
network is also required in order to spread the synchronization ripple to nodes network wide.
The proposed approach assumes the coexistence of reference nodes and normal sensor nodes in
a WSN. A “reference node” periodically transmits beacon messages to its neighbors. These
beacon messages initiate the synchronization waves. Multiple reference nodes are allowed to
operate in the system simultaneously. A sensor node in this approach will dynamically select the
nearest reference node as its reference for clock synchronization.
This approach exploits the usage of multi-channel radios to improve precision, minimize the
communication overhead and low energy consumption. A common control channel is shared by
all the sensor nodes for delivery of beacon messages and control packets. This control channel
can be the same one as is used for general data traffic. Each sensor node is also assigned a
unique clock channel different from all its neighbors’ clock channels. Usage of a dedicated
clock channel reduces the variation in propagation delay caused by packet collisions and
retransmissions, thereby improving the accuracy of clock estimation.
This proposed protocol used for multi-hop synchronization of the network based on pair wise
synchronization scheme suggested by [10]. This requires nodes to synchronize to some
reference point(s) such as a sink node in the sensor network and needs a tree to be constructed
first. Then pair wise synchronization is done along the n - 1 edges of the tree. In such
algorithms, the reference node is the root of the tree and has the responsibility of initiating a
“resynchronization” when required. Using the assumption that the clock drifts are bounded, and
given the required precision, the reference node calculates the time period that a single
synchronization step will be valid. Since the depth of the tree affects the time to synchronize the
whole network, and also the precision error at the leaf nodes, the depth of the tree is
communicated back to the root node so that it can use this information in its resynchronization
time decision.
An explanation of a standard two-way message exchange between a pair of nodes [13]
employing for Synchronization is helpful to understand proposed synchronization design. The
basic building block of the synchronization process is the two-way message exchange between a
pair of nodes Here we assume that the clock drift between a pair of nodes is constant in the
small time period during a single message exchange. The propagation delay is also assumed to
be constant in both directions. Consider a two-way message exchange between nodes A and B
as shown in Fig. 1. Node A initiates the synchronization by sending a synchronization message
at time t1
as per node’s local clock. This Message includes A's identity, and the value of t1
. B
receives this message at t
2
which can be calculated as
t2
= t1
+ + d (1)
Where is the relative clock drift between the nodes, and d is the propagation delay of the
pulse.
B responds at time t3
with an acknowledgement, which includes the identity of B and the values

24
t1
, t2
, and t3
. Then, node A can calculate the clock drift and propagation delay as below, and
synchronize itself with respect to node B.
= ((t2
– t1
) – (t4
– t3
))/2 (2)
d = ((t
2
– t
1
) + (t
4
– t
3
))/2 (3)
The synchronization phase is initiated by the root node's syn_begin message. On receiving this
message, nodes of level 1 initiate a two-way message exchange with the root node. Before
initiating the message exchange, each node waits for some random time, in order to minimize
collisions on the wireless channel. Once they get back a reply from the root node, they adjust
their clocks to the root node. Level 2 nodes, overhearing some level 1 node's communication
with the root, initiate a two-way message exchange with a level 1 node, again after waiting for
some random time to ensure that level 1 nodes have completed their synchronization. This
procedure eventually gets all nodes synchronized with reference to the root node, the
synchronization process described in detail in Subsection 3.3.
Figure 1. Two way message exchange between a pair of nodes.
3.2. Ad Hoc Tree Construction Phase:
Before the sensors can be synchronized, a tree structure based network topology must be
created. The proposed Algorithm 1 is used by each sensor node to efficiently flood the network
to form a logical hierarchical structure from a designated source point. Each sensor is initially
set to accept fd_pckt (flood packets) for first time, but will ignore subsequent ones in order not
to be continuously reassigned as the flood broadcast propagates. When a node receives or
accepts the fd_pckt then first it set to its parent as source of broadcast after that level of current
receiver node will be assigned one more than the level of parent node and then it broadcast the
fd_pckt along with node identifier and level. If a node receives the ack_pckt, the variable
no_receiver increments to keep track of the node’s receivers.
Algorithm 1. Tree Structure Construction
Begin
Accept (fd_pckts)
Initialize : no_reciever = 0;
Node_Level(Root)=0;
If (current_reciever = = root)
Broadcast (fd_pckts)
Else if (current_reciever != root)
Begin
Accept (fd_pckts);
Parent(curent_reciever) = Source(broadcast_msg);
Node_Level(curent_reciever)=Node_Level(Parent)+1;
Broadcast (ack_pckt, node_id);

25
Ignore (fd_pckts);
End
Else if (current_node receives ack_pckt)
no_ receiver++;
End
3.3. Hierarchical Time Synchronization Phase
The first component of TSRT’s bidirectional time synchronization service is the push-based
[12] Hierarchy Time Synchronization (HTS) Scheme. The goal of HTS is to enable central
authorities to synchronize the vast majority of a WSN in a lightweight manner. This approach
particularly based on pair wise synchronization with allusion to single reference node is
discussed in Subsection 3.1.
3.3.1. Single Reference Node
As shown in Fig. 2, HTS consists of three simple steps that are repeated at each level in the
hierarchy. First, a Reference Node (RN) broadcasts a beacon on the control channel (Fig. 2A).
One child node specified by the reference node will jump to the specified clock channel, and
will send a reply on the clock channel (Fig. 2B). The RN will then calculate the clock offset and
broadcast it to all child nodes, synchronizing the first ripple of child nodes around the reference
node (Fig. 2C). This process can be repeated at subsequent levels in the hierarchy further from
the reference node (Fig. 2D). The HTS scheme is explained in more detail as follows:
Step 1: RN initiates the synchronization by broadcasting
the syn_begin message with time t1 using the control
channel and then jumps to the clock channel. All concerned
nodes record the received time of the message
announcement. RN randomly specifies one of its children,
e.g. SN2, in the announcement. The node SN2 jumps to the
specified clock channel.
Step 2: At time t3, SN2 replies to the RN with its
received times t2 and t3.
Step 3.1: RN now contains all time stamps from t1 to t4.
It calculates clock drift and propagation delay d, as
per equation (2) and (3), and calculate t2 = t1 + + d,
and then broadcasts it on the control channel.
Step 3.2: All involved neighbor nodes, (SN2, SN3, SN4 and
SN5) compare the time t2 with their received timestamp
t2’.
i.e. SN3 calculates the offset d’ as:
d’ = t2 - t2’
Finally, the time on SN3 is corrected as:
T = t + d +d’
Where t is the local clock reading.
Step 4: SN2, SN3, SN4 and SN5 initiate the syn_begin to
their downstream nodes.
We assume that each sensor node knows about its neighbors when it initiates the
synchronization process. In Step 1, the response node is specified in the announcement. It’s the
only node that jumps to the clock channel specified by the RN. The other nodes are not
disturbed by the synchronization conversation between RN and SN2 and can conduct normal
data communication while waiting for the second update from the RN. A timer is set in the RN
when the syn_begin message is transmitted. In case the message is lost on its way to SN2, the

26
RN goes back to the normal control channel after the timer expires and thus avoids indefinite
waiting.
As the synchronization ripple spreads from the reference node to the rest of the network, a
multi-level hierarchy is dynamically constructed. Levels are assigned to each node based on its
distance to reference node, i.e. number of hops to the central reference point. Initiated from the
reference nodes, the synchronization steps described above are repeated by the nodes on each
level from the root to the leaves.
Figure 2. (A) Reference node broadcasts (B) A neighbor replies (C) All neighbors are
synchronized (D) Repeat at lower layers
4. NETWORK EVALUATION PHASE
In this phase, the network examines the total amount of message exchanges for synchronization
during the last synchronization period, and then adjusts the duration of the next synchronization
period to minimize the overall energy consumption for synchronization. As per [14] when the
network traffic occurs rarely and synchronization delay is not a critical problem, applying the
sensor initiated (SI) mode is a better choice to save network resources instead of using the
always on (AO) mode. In addition, for some applications, the sensor clocks might be allowed to
go out of synchronization unless sensing events happen. Another critical problem is to
determine the required number of timing message exchanges (beacons) per pair wise
synchronization. To fulfill higher requirement of synchronization accuracy, a larger number of
message transfers and corresponding signal processing is needed for pair wise synchronization.
However, as the number of required timing messages per pair wise synchronization increases,
the overall number of timing messages in a synchronization period increases. Hence, there is a
tradeoff between accuracy and energy consumption.
To address these design challenges, we consider various factors to deter-mine network
parameters such as the synchronization mode, the re-sync period τ, and the number of beacons
per pair wise synchronization N. Indeed, it is aiming at efficient usage of network resources
(message exchanges) in synchronization. The network parameters are summarized as follows:
B: number of branches in a Ad hoc tree of the network.
τ: re-sync period
h: average number of hops per unit time
δ: latency factor reﬂecting the amount of allowed delay in data transmission
N: number of beacons per pair wise synchronization
The number of branches in the network B can be obtained after the Ad hoc Tree Construction
phase. The latency factor δ should be ﬁxed according to the type of a sensor network and its
range is from 0 to 1. The higher latency factor means higher concern for network delays. In
every sensing event, its destination node adds the number of hops that have occurred in that

27
particular transmission to its storage. During the synchronization phase, the reference node
collects the information of the total number of hops occurred in the last synchronization period
and determines the average number of hops per unit time (h) in the network. This information
can be included in timing messages with a small overhead. As mentioned, the goal of this phase
is to minimize the number of required timing messages. In [14] suggested that the number of
timing messages per unit time is given by M=2BN/ τ in AO mode, while in the SI mode,
M=2hN. To minimize the number of timing messages per unit time M, the synchronization
mode should be selected as follows:
2BNδ/τ <> 2hN (4)
Where the latency factor δ varies from 0 to 1 such that more delay dependant networks assumes
a larger value of δ and vise versa (0 ≤δ≤1). For example, δ is set to be 0 for sensor networks
requiring network synchronization all the time. On the other extreme, for delay-independent
networks, δ should be close to 1. As the clock synchronization period τ increases, the network
becomes more power efficient. Thus, τ should be chosen as large as possible. However, a too
large value of τ induces a critical synchronization problem since the clock difference (offset)
between nodes keeps generally increasing with time. Hence, there exists a maximum timing
synchronization period (τmax) which is determined by the oscillator regulations (hardware
specifications) and the accuracy of estimators. Notice that sensing data transmission is not
available during the synchronization phase (τsync), so the re-sync period τ = τmax + τsync. In
continuation, (4) can be rewritten as
τ < Bδ/h (5)
From (5), the synchronization mode changes from AO into SI when τ is smaller than Bδ/h and
vise versa. In the SI mode, the reference node periodically asks the number of hops that
occurred during the past time interval, and then make a decision whether or not to switch to the
AO mode. Actually, τ is also dependent on N since it strongly depends on the accuracy of
timing offset estimators. A more detailed analysis of τ is provided in next Subsection.
4.1. Optimum Number of Beacons (N) and Resynchronization Period (τ)
The number of timing messages (beacons) per pair wise synchronization (N) is a critical
parameter to determine both the synchronization accuracy and power efficiency. Suppose that
the clock timing mismatch ε between the two nodes is modeled as follows: ε = εo + εst, where t
denotes the reference time, εo and εs stand for the clock offset and skew errors, respectively. Let
εo,i and εs,i denote the clock offset and skew estimation errors when i message exchanges occur
between the two nodes. In general, it is difficult to determine any specific mathematical model
for either clock offset or skew errors. In this paper, we model both clock offset and skew errors
by normal distributions based on the experimental results reported in [9], [10]:
εo,i = N (0, σ2
εo,i
) 1 ≤ i ≤ N,
εs,i = N (0, σ2
εs,i
) 1 ≤ i ≤ N,
where εs,1 stands for the clock skew error when no skew estimation occurred. Then, the
maximum clock mismatch can be modeled another normal distribution ε = N(0,σε
2
), where σε
2
= σ2
εo,n + σ2
εs,n τmax
, (t= τmax
). Imposing the upper-limit εmax for the clock error via the
probabilistic measure:
Ps= Pr (|ε|≤ εmax) = erfc (εmax / √2σε
)
where erfc(x) ≡ (2/√ π) ∫x
∞
e
−t2
dt and Ps denotes the network wide sync error probability. Thus,
σε can be determined when εmax and the limit of sync error probability are fixed. For instance,

28
when Ps is limited to 0.1% and εmax is 10 ms, σε becomes 3.04 m.
The maximum timing sync period with N beacons can be written as
τ max
= √ ((σ2
ε
- σ2
εo,N
)/ σ2
εs,i,N
) (6)
Based on the lower bounds and asymptotic performance of the estimators, one can easily infer
closed-form expressions of the variances εo,N and εs,N in terms of the variances εo,1 and εs,2
respectively.
5. ANALYSIS OF TSRT
The TSRT protocol exploits the broadcast nature of the wireless medium to establish a single
common point of reference, i.e. the arrival time of the broadcast message is the same on all
neighbor peers. This common reference point can be used to achieve synchronization in Step 4
of the Subsection 3.3, i.e. t2 at node SN2 occurred at the same instant as t2’ at node SN3. As the
RN is synchronizing itself with SN2, the other neighboring nodes can overhear the RN’s initial
broadcast to SN2 as well as the RN’s final update informing SN2 of its offset d2. If in addition
the RN includes the time t2 in the update sent to SN2 (redundant for SN2), then that allows all
neighbors to synchronize.
The intuition is that, since t2 and t2’ occurred at the same instant, then overhearing t2 gives SN3
its offset from SN2’s local clock and overhearing d2 gives SN3 the offset from SN2’s local
clock to the RN reference clock. Thus, SN3 and all children of the RN can calculate their own
offsets to the RN reference clock with only three messages (2 control broadcasts and 1 clock
channel reply) TSRT is highly scalable and lightweight, since there is only one lightweight
overhead exchange per hop between a parent node and all of its children. In contrast,
synchronization in RBS happens between a pair of neighbors, which is called pair verification,
rather than between a central node and all of its neighbors.
As a result, RBS is susceptible to high overhead as the number of peers increases [1]. The TSRT
approach eliminates the potential broadcast storm that arises from pair wise verification, while
at the same time preserving the advantage of reference broadcasting, namely the common
reference point. Also, since the TSRT parent provides the reference broadcast that is heard by
all children, then TSRT avoids the problem in RBS when two neighbors of an initiating peer are
“hidden” from each other. The parameters used in the protocol dynamically assign the hierarchy
level to each node during the spread of the synchronization ripple and no additional routing
protocol is required. TSRT is lightweight since the number of required broadcasting messages is
constant in one broadcasting domain.
Only three broadcast messages are necessary for one broadcasting domain, no matter how dense
the sensor nodes are. In TSRT, the sender error is eliminated by comparing the received time on
each node.
TSRT’ current policy for selecting the child node to respond to the sync begin message is a
random selection. However, it is possible to incorporate historical knowledge from previous
TSRT cycles in the selection of the next child responder. Moreover, previous TSRT responses
may be combined to broadcast a composite value in Step 3. This may be useful to account for
propagation delay differences between neighbors within a local broadcast domain, which we
can assumed to be small, but which may become more relevant when the distances between
neighbors becomes very large in a highly distributed WSN.
The performance comparison of TSRT and TPSN in terms of the average number of message
exchanges M with respect to the number of beacons N is shown in fig. 3 and fig. 4 when Ps is
assumed as 0.01% and 1% respectively. This simulation is based on the linear network model
where the depth of the network B = 5, εmax =10ms, σεo
= 16.67µ, d = 10ms, t = 400ms, and σεs

29
=1.58µ have assumed.
It can be observed that TSRT requires a less number of timing messages than TPSN when there
multiple numbers of beacon transmissions are required. Moreover, the gap of the average
number of required timing messages between TSRT and TPSN signiﬁcantly increases as N
increases, and thus TSTP is by far more efficient than TPSN for large value of N. It can be also
seen that a few number of beacons is enough to minimize M for TSRT. Besides, as expected, a
larger number of beacons required to meet a more strict constraint of the network-wide error
probability Ps. In practice, a lower number of N is highly preferable since it is proportional to
the synchronization time, i.e., a lower N induces better latency performance. Although, it may
not be optimal in terms of energy consumption.
Figure 3. Average Number of Messages (M) Figure 4. Average Number of Messages (M)
for Ps=0.01% for Ps=1%
6. CONCLUSION
WSN have tremendous advantages for monitoring object movement and environmental
properties but require some degree of synchronization to achieve the best results. The proposed
TSRT synchronization approach is able to produce deterministic synchronization with only few
pair wise message exchanges.
While the proposed approach is especially useful in WSN which are typically, extremely
constrained on the available computational power, bandwidth and have some of the most exotic
needs for high precision synchronization. The proposed approach was designed to switch
between TPSN and RBS. These two algorithms allow all the sensors in a network to
synchronize themselves within a few microseconds of each other, while at the same time using
the least amount of resources possible. In this work two varieties of the algorithm are presented
and their performance is verified theoretically with the existing results and compared with
existing protocols. The comparison with RBS and TPSN shows that the proposed
synchronization approach is lightweight since the number of required broadcasting messages is
constant in one broadcasting domain.
ACKNOWLEDGEMENTS
The authors are thankful to all synchronization papers that are credited as references below.
This paper contains comprehensive study of papers on synchronization algorithms or
approaches collectively, cited from various contributions from the authors mentioned below.

30
REFERENCES
1. Rahamatkar, S., Agarwal, A. & Kumar, N., (2010) “Analysis and Comparative Study of Clock
Synchronization Schemes in Wireless Sensor Networks”, Int. J. Comp. Sc. & Engg., Vol. 2, No. 3, pp
523-528.
2. Rahamatkar, S., Agarwal, Ajay & Sharma, V., (2009) “Tree Structured Time Synchronization
Protocol in Wireless Sensor Network”, UbiCC Journal, Vol. 4, pp. 712-717.
3. Abolfazl, A., Beikmahdavi, N. & Naderi, B. S., (2010) “Cluster-based and cellular approach to fault
detection and recovery in wireless sensor network”, Int. J. Wireless & Mobile Network, Vol. 2, No. 1,
pp 97-108.
4. Ranganathan, Prakash & Nygard, K., (2010) “Time Synchronization in Wireless Sensor Networks: A
Survey”, Int. J. of UbiComp, Vol. 1, No. 2, pp 92-102.
5. Rhee, K., Lee, J. & Wu, Y.C., (2009) “Clock Synchronization in Wireless Sensor Networks”, An
Overview. Sensors 9, pp. 56-85,
6. Zhao, F. & Guibas, L., (2004) “Wireless Sensor Networks: An Information Processing Approach,
Morgan Kaufmann”, pp. 107-117.
7. Mills, D. L., (1991) “Internet Time Synchronization: The Network Time Protocol”, IEEE Trans.
Comm. 39 (10), pp. 1482–1493.
8. Bulusu, N. & Jha, S. (2005) Wireless Sensor Networks: A Systems Perspective, Artech House:
Norwood MA, USA.
9. Elson, J. E., Girod, L. & Estrin, D., (2002) ”Fine-Grained Network Time Synchronization using
Reference Broadcasts”, 5th
Symposium on Operating Systems Design and Implementation, pp. 147–
163.
10. Ganeriwal, S., Kumar, R. & Srivastava, M. B., (2001) “Timing-Sync Protocol for Sensor Networks”,
First ACM Conference on Embedded Networked Sensor System (SenSys), pp. 138–149.
11. Woo, Culler, D., (2001) ”A Transmission Control Scheme for Media Access in Sensor Networks”,
Mobicom, pp. 221–235.
12. Dai, H. & Han, R., (2004) “TSync: a lightweight bidirectional time synchronization service for
wireless sensor networks”, SIGMOBILE Mob. Comput. Commun. Rev. Vol. 8 (1), pp. 125-139.
13. Noh, K. L. & Chaudhari, Q., Serpedin, E., Suter, B., (2007) “Analysis of clock offset and skew
estimation in timing sync protocol for sensor networks”, IEEE Globecom, San Francisco.
14. Kyoung-Lae Noh & Erchin Serpedin, (2007) “Adaptive multi-hop timing synchronization for
wireless sensor Networks”, 9th Int. Symp. On Signal Processing & Its Applications.
15. PalChaudhuri, S., Saha, A. K. & Johnson, D. B., (2004) “Adaptive clock synchronization in sensor
networks” 3rd int. Symp. on inf. Processing in Sensor Networks, pp. 340-348.
16. Sun, K., Ning, P. & Wang, C., (2006) “TinySeRSync: Secure and Resilient time synchronization in
wireless sensor networks”, 13 ACM Conf. on Comp. Comm. Security, pp. 264-277.
17. Ganeriwal, S., Kumar, R., Adlakha, S. & Srivastav, M., (2003) “Network-wide Time
Synchronization in Sensor Networks”, Technical Report, Networked and Embedded Systems Lab,
Elec. Eng. Dept., UCLA.
18. Greunen, J. & Rabaey, J., (2003) “Lightweight Time Synchronization for Sensor Networks”, 2nd
ACM Int. Workshop on WSN & Applications, pp. 11–19.

31
Authors
Surendra Rahamatkar has received his Bachelor of Engineering in
Computer Science & Engineering from Barkatullah University, Bhopal,
India, Post Graduate Diploma in Advanced Computing from CDAC, Pune
and Master in Technology in Computer Science & Engineering from
VMRF Deemed University, India. He is a member of various Technical
Societies viz. CSI, International Association of Engineers (IEA), ISTE. He
is internationally recognized as a Member of Editorial Board of
International Journal of Computer Applications (IJCA), New York, USA,
International Journal of Advanced Engineering & Applications and
Reviewer of International Journal of Computer Theory and Engineering
(IJCTE), Singapore & ICMLC 2011 Singapore. He published many
research papers in various International/ National Journals and
Conferences. His main research interests include: Wireless Sensor
Network, Distributed & Mobile Computing and Middleware.
Dr. Ajay Agarwal has done B.Tech. Degree in Computer Science &
Engineering from Institute of Engineering & Technology, Lucknow
(India), M.Tech.(honors) Degree in Computer Science & Engineering from
Motilal Nehru Regional Engineering College, Allahabad and Ph.D. in
Computer Science from Indian Institute of Technology, Delhi (India).
Presently he is working as a Professor and Head in department of
Computer Applications at Krishna Institute of Engineering & Technology,
Ghaziabad, India. He is a member of various Technical Societies viz.
Institute of Electrical and Electronics Engineers (IEEE), Computer Society
of India (CSI), Indian Society for Technical Education (ISTE), Institution
of Engineers India, Institute of Chartered Computer Professional of India
and Indian Association of Physics Teachers. He published many papers in
various International/ National Journals and Conferences. His main
research interests include: Wireless Sensor Network, Mobile Computing
and Middleware.

A REFERENCE BASED, TREE STRUCTURED TIME SYNCHRONIZATION APPROACH AND ITS ANALYSIS IN WSN

More Related Content

What's hot (17)

Similar to A REFERENCE BASED, TREE STRUCTURED TIME SYNCHRONIZATION APPROACH AND ITS ANALYSIS IN WSN (20)

Recently uploaded (20)

A REFERENCE BASED, TREE STRUCTURED TIME SYNCHRONIZATION APPROACH AND ITS ANALYSIS IN WSN