Time Synchronization In Wireless Sensor Networks Computer Science Essay

Published: Last Edited:

This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.

The demand for economical wireless sensor nodes have been increased due to advancement in processor, memory and wireless communication techniques. To give the on time and fast response for different actions, wireless sensor nodes work together in cooperation with each other. The most important requirement to maintain the co-ordination among different nodes during execution of an application is perfect time synchronization between them. Message packets are used for time synchronization in wireless sensor networks but transmission of these packets require significant amount of energy which is supplied by wireless sensor nodes. As these sensor nodes have limited battery capacity, usage of large number of packets has a negative impact upon their lifetime. So, to reduce the number of message packets required for the time synchronization in wireless sensor networks, we have used Optimal Performance Reference Broadcast Synchronization (OPRBS). This protocol performs synchronization with ranging i.e. select a particular range of broadcast according to transmission power of sensor nodes. With this decrease in the message packets for transmission, the convergence time among wireless sensor nodes is reduced and the lifetime of wireless sensor nodes is also extended as much as the amount of saved battery energy.

1. Introduction

Time synchronization is a crucial aspect of any networked system. The majority of research in this field has concentrated on traditional high-speed computer networks with few power restraints, leading to the global positioning system (GPS) [7] and the network time protocol (NTP) [5]. These conventional networks are effective for communication of large amounts of data, typical of local area networks (LAN).

With the emergence of WSNs, current LAN synchronization methods will not work efficiently. GPS provides good synchronization accuracy, but requires a very large amount of power from the sensors. In a power-constrained sensor, this synchronization is infeasible. NTP is also infeasible since it is designed for traditional computer networks and will not scale well for wireless sensor networks. Some new synchronization methods have been developed specifically for sensor networks, such as the timing-sync protocol for sensor networks (TPSN) [4], Time diffusion protocol (TDP) [6] and the reference broadcast synchronization method (RBS) [1].

In the case of NTP and GPS method, when the timestamp in messages is broadcasted to exploit as a global time, the sending time and the access time of messages incurs time delay problem for time synchronization. To solve this problem, the Reference Broadcast Synchronization (RBS) [2] method has been proposed. The RBS has an advantage that it greatly reduces the time period for the message sending and access. However, as the number of sensor nodes is increased, the associated messages for time synchronization are also increased. In wireless sensor nodes, the message transmission via the RF antenna takes a major part of the energy consumption. Therefore, in the RBS method, the increase of messages for time synchronization results in a significant energy waste in wireless sensor nodes.

To reduce the energy waste problem and number of messages for the time synchronization in wireless sensor network, we propose Optimal Performance Reference Broadcast Synchronization (OPRBS). In addition the reduced messages will also shorten the synchronization time for total time synchronization.

As a result the saved energy has contributed to the extension of the life time of the wireless sensor nodes and the short converges time has guaranteed the quick response to events inadvertently happened.

These advantages become achievable thanks to the method of Ranging. In Ranging, a particular size of nodes is decided by transmission power of sensor nodes. In this paper we compare and analyzed the performance of RBS and OPRBS mechanism according to number broadcast messages for synchronization for sensor networks.

2. Synchronization Mechanisms

2.1 TDP

The TDP [6] architecture classifies into three large procedures as illustrated in figure 2.1. Sensor nodes self-determine to become master nodes with the election/reelection of master/diffused leader node procedure (ERP), which consists of the false ticker isolation algorithm (FIA) and load distribution algorithm (LDA). At the end of procedure ERP, the elected master nodes start the peer evaluation procedure (PEP) while others do nothing. The procedure PEP helps to remove false tickers from becoming neither a master node nor a diffused leader node. After procedure PEP, the elected master nodes start the time diffusion procedure (TP), where they diffuse the timing information messages at every δ second (round interval) for duration of τ seconds. Each neighbor node receiving these timing information messages self-determines to become a diffused leader node using the procedure ERP. Furthermore, all neighbor nodes adjust their local clocks using time adjustment algorithm (TAA) and clock discipline algorithm (CDA) after waiting for δ seconds.

Figure 2.1. TDP architecture

2.2 TPSN

TPSN provides synchronization for a whole network [4]. First, a node is elected as a synchronization master (details for this are not specified), and a spanning tree with the master at the root is constructed by flooding the network. In a second phase, nodes synchronize to their parent in the tree by means of round­trip synchronization.

Synchronization is performed in rounds and initiated by the root root broadcast­ing a synchronization ­request message to its children. Each child then performs a round­trip measurement to synchronize with the root. Nodes further down in the tree overhear the messages of their parents and start synchronization when their parents have synchronized. To eliminate message delay uncertainties, time stamping for the round­trip synchronization is done in the MAC layer. In case of node failures and topology changes, master election and tree construction must be repeated.

The TPSN works in two phases:

1. Level discovery phase: this is a very similar approach to the flooding phase in RBS, where a hierarchical tree is created beginning from a root node.

2. Synchronization phase: in this phase, pair-wise synchronization is performed between each transmitter and receiver.

In the level discovery phase, each sensor node is assigned a level according to the hierarchical tree. A pre-determined root node is assigned as level 0 and broadcasts a level discovery packet. Sensors that receive this packet are assigned as children to the transmitter and are set as level 1 (they will ignore subsequent level discovery packets). Each of these nodes broadcasts a level discovery packet, and the pattern continues with the level 2 nodes.

2.3 RBS

In RBS scheme, nodes periodically send beacon messages to their neighbors using the network's physical-layer broadcast. Recipients use the message's arrival time as a point of reference for comparing their clocks. The message contains no explicit timestamp, nor is it important exactly when it is sent. The accuracy of RBS is mostly determined by the amount of time it takes either node to receive and process the reference packet [2].

Figure 2.2: Comparison of traditional synchronization systems to RBS

The traditional sources of time synchronization error characterize it as having four distinct components. To better understand the source of these errors, it is useful to decompose the source of a message's latency. There are four main sources of delays that must be accounted for to have accurate time synchronization:

Send time: this is the time to create the message packet.

Access time: this is a delay when the transmission medium is busy, forcing the message to wait.

Propagation time: this is the delay required for the message to traverse the transmission medium from sender to receiver.

Receive time: similar to the send time, this is the amount of time required for the message to be processed once it is received.

RBS eliminates error introduced by the Send time and Access time from critical path. And for our purposes [9], RBS consider the Propagation time to be effectively 0, because for a LAN or ad-hoc network spanning tens of feet, propagation time is at most tens of nanoseconds, which does not contribute significantly to our micro-scale error budget. The largest sources of nondeterministic latency can be removed from the critical path by using the broadcast channel to synchronize receivers with one another.

The main advantage of RBS is that it eliminates the uncertainty of the sender by removing the sender from the critical path. By removing the sender, the only uncertainty is the propagation and receives time. The propagation time is negligible in networks where the range is relatively small. It is claimed that the reference beacon will arrive at all the receiving nodes instantaneously. By removing the sender and propagation uncertainty the only room for error is the receiver uncertainty. Figure 6.1 illustrates this concept.

As seen here, the critical path in a traditional system, which is the top diagram, includes the sender. Since RBS is a receiver to receiver synchronization the sender is removed from the critical path. The critical path on contains the propagation and the receiver uncertainty. If, however, the transmission range is relatively small, then we can eliminate the propagation time and the critical path only contains the uncertainty of the receiver.


Based on the study I propose a new mechanism called Optimal Performance Reference Broadcast Synchronization, which is an improved version of RBS. This OPRBS can be used in environment with a large number of sensor nodes. In RBS, the reference packet sender broadcast a reference packet to all sensor nodes. Each sensor node records the time at which the reference packet was received, according to its local clock and then, the sensor nodes exchange their observation. So in case of RBS the reference packet sender transmits the broadcast message to all the sensor nodes but in OPRBS we determine the transmission range of reference packet sender, according to their transmission power and broadcast is done only in this particular range. Thus this method results in performance optimization both in terms of number of messages used for synchronization as well as time taken for synchronization.

The OPRBS algorithm can be split into four major events:

1. Ranging: First of all, the range of broadcast is decided by each sensor node according to it's transmission power.

2. Flooding: A transmitter broadcasts a synchronization request packet.

3. Recording: The receivers record their local clock time when they initially pick up the synchronization signal from the transmitter.

4. Exchange: The receivers exchange their observations with each other.

For a very high accuracy sensor network the simplest design has a maximum number of two receivers for a single transmitter. But in practical situations the number of receiver is usually much greater than two. So when the number of receivers is increased more than two, the number of broadcast messages and the synchronization time both are increased. Because the receivers will exchange their observations with each other multiple times.

In OPRBS due to the ranging process the number of synchronization messages is reduced thereby reducing the synchronization time also. And hence, the above mention problem of RBS is greatly overcome. As the transmitter broadcasts the time message only to those sensor nodes, which come in its range, the number of broadcast messages and synchronization time is greatly reduced. This is verified by the result of simulation.

3. Simulation and Results

3.1 Simulation Environment

The simulations were conducted using Matlab 7.5.0 (R2007b). As for the number of sensor nodes, 800 (for the small size network) and 2000 (for the large size network) nodes are used.

Results are based on simulations of connected wireless sensor networks consisting of 800-2000 nodes. The nodes are placed uniformly within a 2-dimensional rectangular area. Additionally, it is assumed that the simulation network contains a reference node that keeps accurate time.

3.2 The simulation results of OPRBS

To measure the performance of the OPRBS method in the single hop model, firstly make Graphical User Interface. Base on this GUI, a reference packet sender node is deployed into the first position and the other sensor nodes are spread across the entire Grid map randomly or uniformly. As the performance metrics, the number of messages and the synchronization time for all sensor nodes are considered.

The figure 3.1 shows the number of messages according to the increase in the number of sensor nodes for OPRBS.

Figure 3.1: Number of Nodes versus Number of Messages for OPRBS

The above graph is drawn by result find out from the GUI of OPRBS. The table 3.1 shows the number of synchronization messages for different number of sensor nodes.

Table3.1. Number of Nodes versus Number of synchronization Messages

Number of Nodes

Number of synchronization Messages















The figure 3.2 shows the synchronization time according to the increase in the number of sensor nodes for OPRBS

Figure 3.2: Synchronization Time with

different no. of Sensor Node for OPRBS

The above graph shows the synchronization time of different number of sensor nodes which is find out with GUI of OPRBS

Table 3.2 shows the synchronization time for different number of sensor nodes

Table3.2. Number of Nodes versus Synchronization Time

Number of Nodes

Synchronization Time (sec)















Figure 3.3 shows the comparative graph of RBS and OPRBS in terms of number of synchronization messages [1].


Figure 3.3: Number of synchronization message with different number of sensor node

It is clearly shown by figure 8.3 the gradient of OPRBS is less than the RBS i.e. with the increasing of sensor node the number of synchronization message are not increasing rapidly as compared to the RBS. For 800 sensor node we require 799 synchronization messages for OPRBS while in case of RBS synchronization messages are 1250. Similarly for 2000 sensor nodes the number of messages required for OPRBS is 1999 and in case of RBS it is 2750.

Table 3.3 is the comparative table of the number of synchronization message for RBS and OPRBS.

The number of synchronization messages

The number

of nodes























1999Table 3.3 A comparative table of the number of synchronization messages

Table 3.3 shows the comparison between RBS and OPRBS based on the number of generated messages in overall synchronization in the wireless sensor network. As the number of sensor nodes increases, the time required for synchronization i.e. synchronization time and number of broadcast messages also increases. This results show that the number of nodes affects the number of generated messages in overall synchronization. When the numbers of nodes are 800, the numbers of generated messages of RBS are more than that of OPRBS.

4. Conclusions

The Optimal Performance Reference Broadcast Synchronization (OPRBS) proposed which is an improvement over the RBS algorithm. The OPRBS takes relatively less time for synchronization and also involves less number of broadcast messages. And so, this results in less power consumption in wireless sensor networks. Due to less power consumed by the sensor nodes, the lifetime of sensor nodes could we prolonged.

These advantages become achievable thanks to the method of Ranging. In Ranging, a particular size of nodes is decided by transmission power of sensor nodes.

Instead of synchronizing within the timestamp values of all sensor nodes as in the case of RBS method, the OPRBS performs the synchronization based on the reference packet and ranging. In OPRBS the messages were broadcasted within a particular range while in RBS messages broadcasted to all sensor nodes. Thus, this approach has brought down the number of synchronization message and synchronization time which were regarded as the sources of inaccurate time synchronization. In addition, since broadcast was done in a particular range, the number of messages exchanged is also reduced. Since the decrease in the number of messages, the energy consumption by sensor nodes is also reduce. Which caused the lifetime of wireless sensor nodes could be extended. Throughout the detailed experiments of OPRBS and the RBS method, it is confirmed that the OPRBS method has greatly decreased the number of messages for synchronizing all the sensor nodes in network. In addition, the reduced messages might shorten the synchronization time for synchronization.

From the result of simulation it also confirms that the number of synchronization message for OPRBS is less required than RBS. So the synchronization time also reduce of OPRBS.