Contact at or 8097636691
Responsive Ads Here

Tuesday, 13 February 2018

DA Sync A Doppler Assisted Time Synchronization Scheme for Mobile Underwater Sensor Networks(2014)

DA Sync A Doppler Assisted Time 

Synchronization Scheme for Mobile Underwater Sensor Networks(2014)

Time synchronization plays a critical role in distributed network systems. In this paper, we investigate the time synchronization problem in the context of underwater sensor networks (UWSNs). Although many time-synchronization protocols have been proposed for terrestrial wireless sensor networks, none of them can be directly applied to UWSNs. This is because most of these protocols do not consider long propagation delays and sensor node mobility, which are important attributes in UWSNs. In addition, UWSNs usually have high requirements in energy efficiency. To solve these new challenges, innovative time synchronization solutions are demanded. In this paper, we propose a pairwise, cross-layer, time-synchronization scheme for mobile underwater sensor networks, called DA-Sync. The scheme proposes a framework to estimate the doppler shift caused by mobility, more precisely through accounting the impact of the skew. To refine the relative velocity estimation, and consequently to enhance the synchronization accuracy, the Kalman filter is employed. Further, the clock skew and offset are calibrated by two runs of linear regression. Simulation results show that DA-Sync outperforms the existing synchronization schemes in both accuracy and energy efficiency.
Ø Existing time-synchronization schemes use half of the round trip time to calculate one way propagation delay.
Ø Due to the node mobility, propagation delays on the way forth and back are not necessarily identical, especially     when nodes move at high speed.
Ø This issue severely decreases the accuracy of most time synchronization approaches.
Ø Both accuracy and energy efficiency is less.
Ø Long propagation delay.
Ø In this paper, we propose a novel time-synchronization scheme, called DA-Sync, which is a fundamental cross-     layer-designed time-synchronization protocol specific for mobile UWSNs.
Ø DA-Sync provides a fundamental method to synchronize two sensor nodes, i.e., an ordinary node and a                   reference node.
Ø  The scheme proposes a framework to estimate the doppler shift caused by mobility, more precisely through            accounting the impact of the skew.
Ø Different algorithms have different sync message (including request and response messages) packet sizes             since they need to carry different amounts of information.
Ø High accuracy and high energy efficiency.
Ø Reduce the nondeterministic errors that are commonly encountered by time synchronization algorithms which       rely on message exchanges.
1.    Process Initialization
2.    Data Collection
3.    Velocity Estimation
4.    Propagation-Delay Estimation
5.    Linear Regression
6.    Calibration
Process Initialization:
For time synchronization between pairs of clocks, most of the algorithms rely on estimating the clock offset and skew, which present the relation between the time measured by two different clocks. DA-Sync also yields to this pairwise synchronization approach. In DA-Sync, an ordinary node’s clock aims to become synchronized with the reference clock of the reference node.
Data Collection:
In the beginning, an ordinary node initiates message exchanges by sending a “Sync-Req” message to a neighboring reference node. The ordinary node records the sending time stamp T1, obtained at the MAC layer, right before the message leaves. Upon receiving the Sync-Req message, the reference node estimates and records the ordinary node’s relative moving velocityv0 with Doppler shifts as specified. Meanwhile, it marks its local time as t2. Then, after a time interval tr (waiting for the hardware sending-receiving transition and avoiding collisions), the reference node sends back a “Sync-Res” message which containst2,v0 and its sending time t3. When receiving the Sync-Res message, the ordinary node records its receiving timeT4 and its relative moving velocity to the reference node, v1. The message exchange process between the ordinary node and the reference node.
Velocity Estimation:
Considering that multiple message exchanges occur in the synchronization process, estimation of the relative speed between the reference node and the ordinary node can be improved by incorporating the estimates obtained during the previous message exchanges. Assuming the relative motion between the reference node and the ordinary node follows the first-order kine-matic model.
Propagation-Delay Estimation:
In this module, we aims to estimate the long and dynamic propagation delays. In DA-Sync, two propagation delays need to be estimated, i.e., 1and 2. In terms of delay estimation, many other time synchronization protocols take half of the two-way round trip time as one-way propagation delay, which means that 1is equivalent to 2. In DA-Sync, by leveraging relative moving velocities obtained from the physical layer, we can differentiate 1and 2, which is required in mobile scenarios. Regarding the velocity, in DA-Sync, only one dimension is considered. That is because, to estimate the propagation delay, only the relative velocity is demanded. The individual moving pattern of sensor nodes does not affect the propagation delay between them. Thus, in DA-Sync, for the ordinary node, it essentially either moves toward to the reference node or moves far away from the reference node. The amazing thing is that the velocity estimated by utilizing the Doppler scaling factor is exactly what is needed, i.e., the relative velocity. Therefore, the velocities obtained from the physical layer can be directly applied in DA-Sync. The following steps show how propagation delays 1 and 2 are estimated in each message exchange process.
Linear Regression:
The ordinary node performs linear regression over the data points in to estimate the draft lock skew and offset, where determines the index of the message exchange process. WLSE is an approach for correcting the problem of hetero skedasti city by log-likelihood estimation of a weight that adjusts errors of prediction. It performs better than OLSE because it introduces an extra functional coefficient “weight” to module the power of each sample data.
The ordinary node carries out the calibration process to correct some parameters and recalculate the clock skew and offset, as shown in Fig. The calibration process will run iteratively until it reaches the stopping criteria, which is either when the number of runs reaches the maximum loop number 10 or the difference between estimated skew in this run or last run is less than 100 ppm.
Ø System                          :         Pentium IV 2.4 GHz.
Ø Hard Disk                      :         40 GB.
Ø Floppy Drive                 :         1.44 Mb.
Ø Monitor                         :         15 VGA Colour.
Ø Mouse                            :         Logitech.
Ø Ram                               :         512 Mb.
 Operating system           :         Windows XP/7.
 Coding Language          :
 Tool                                    :           Visual Studio 2010

No comments:

Post a Comment