Loading...

Real-Time Fusion of Asynchronous Data in Distributed Sensor Networks

Talebi, Hadi | 2013

793 Viewed
  1. Type of Document: Ph.D. Dissertation
  2. Language: Farsi
  3. Document No: 44663 (19)
  4. University: Sharif University of Technology
  5. Department: Computer Engineering
  6. Advisor(s): Hemmatyar, A. M. Afshin
  7. Abstract:
  8. Real-time asynchronous data fusion for high-speed phenomena is an important and challenging task in the sensor networks. Examples of data fusion applications in sensor networks are: managing the traffic of maneuvering airplanes and ground vehicles in airside areas of an airport, traffic management in streets and roads, Driver Assistance Systems, guidance of antiaircraft and antimissile missiles. In all the data fusion applications the estimation of the required variables is necessary.
    In this research two methods are introduced for real-time asynchronous data fusion, especially for track-to-track fusion of high-speed phenomena in sensor networks. The effectiveness and usability of these methods are evaluated by mathematics reasoning and simulation. The first method is based on the estimation of time of sample. Each sensor uses its Kalman filter to obtain the best estimation of sample data. The outputs of Kalman filter along with a criterion of error are transmitted to the center of fusion. In the fusion center upon arrival of data from a sensor the actual time of the sample is estimated with respect to the time axis of the fusion center. The estimated time is used to predict the received data for the start of the next fusion period. This process is done for all the related incoming data. These data are pseudo-synchronized for the start of the next fusion period, and can be fused in time. The fusion law is element-wise Linear Minimum Variance Unbiased Estimator. The second method is based on the synchronized clocks in all of the sensors with at most one millisecond difference with respect to the fusion center clock. Here each sensor knows the exact fusion times of the fusion center, e.g. 1, 2, 3, etc. seconds. The input data of each sensor passes through a Kalman filter. The filtered data are locally predicted for the next fusion time, and then the predicted data are transmitted to the fusion center. In the fusion center the related state estimates are pseudo-synchronized and can be fused with the element-wise Linear Minimum Variance Unbiased Estimator. Also in this paper it is declared that the correlation between state estimates of different sensors is eliminated after pseudo-synchronization. Therefore the element-wise Linear Minimum Variance Unbiased Estimator is equivalent to a matrix Linear Minimum Variance Unbiased Estimator.
  9. Keywords:
  10. Asynchronous Data Fusion ; Distributed Sensor Networks ; Pseudo-Synchronization ; Kalman Filters ; Data Fusion ; Common Process Noise ; Linear Minimum Variance Unbiased Estimator (LMVUT)

 Digital Object List

 Bookmark

No TOC