검색 상세

System Design and Image Enhancement for SAR Satellite Constellations

  • 김경록 (아주대학교 일반대학원)

초록/요약

With the remarkable advances in space technologies, small satellite and synthetic aperture radar (SAR) are innovated for satellite communication and remote sensing at low Earth orbit. In small satellite, the payload, such as SAR and other sensors, should be designed the small size and the light weight while preserving functional aspects. Thus, the small satellites have relatively limited function than large satellites. The SAR constellations, which is a group of small satellites, collect the global data during the mission and downlink the enormous data to the ground station within a few minutes. The ground stations which are distributed over a world-wide should process the downlinked SAR data for providing various services, e.g. SAR images, global disaster monitoring, object detection. Therefore, this dissertation proposes optimization of SAR system and image applications for spaceborne SAR constellations. First, the system design with multi direct digital synthesizer using polynomial regression is proposed for high-quality and small signal generator in X-band SAR system. The proposed scheme is able to generate stable wideband waveform using less system clock than conventional method. The size of developed hardware is reduced containing digital part as well as analog part. Second, the adaptive selection algorithm considering the queue-backlog of ground station is proposed to maximize time-average SAR image quality in data-burst SAR constellations. The algorithm selects the time-average optimal filter under consideration of queue state. As the selected result, the satellite SAR data is processed to the corresponding images for providing to the users such as institute, military, academic and general user. Lastly, the optimization of image processing method is designed to utilize spaceborne SAR constellations. The proposed algorithm maximizes object detection accuracy while providing stabilized service in the ground station. To achieve this goal, the area of interest and the quality requirements are defined in terms of users. In conclusion, this dissertation enables that small SAR can have the improved image application with the intelligent ground stations in satellite constellations.

more

목차

1. Introduction 1
1.1 Background and motivation 1
1.2 Contributions 3
1.3 Organization 6
2. Related work 8
2.1 High quality signal generation and its related work 8
2.2 Efficient data processing for SAR application in satellite constellations 12
3. System Design by Predistortion Polynomial Regression 15
3.1 Signal model 18
3.2 M-DDS signal generator 21
3.2.1 Design of M-DDS 21
3.2.2 Development of signal generator 23
3.2.3 Phase calibration in signal generator 27
3.3 Implementation results 33
3.3.1 Simulation results 33
3.3.2 Hardware measurement results 35
3.4 Concluding remarks 43
4. Image Enhancement of Spaceborne SAR using Adaptive Filtering 45
4.1 Mathematical model 48
4.2 Adaptive speckle filtering 50
4.2.1 SAR rawdata processing 51
4.2.2 SAR image despeckling 59
4.3 Performance evaluation 66
4.3.1 Qualitative performance of despeckling 66
4.3.2 Performance evaluation by Lyapunov optimization 68
4.4 Concluding remarks 69
5. Stabilized Service Providing by Lyapunov-Queue Optimization 71
5.1 System model 75
5.2 Performance evaluation 79
5.2.1 Evaluation setup 79
5.2.2 Vessel detection accuracy with adaptive filter selection 80
5.2.3 Evaluation results 86
5.3 Concluding remarks 91
6. Conclusion 92
References 95

more
반출 Meta View 목록