Adaptive Observing Synthetic Aperture Radar with Digital Beamforming

Kurzfassung

The popularity of Earth Observation imaging radars and the diversity of their applications have been growing tremendously over the past few decades. Airborne and spaceborne Synthetic Aperture Radars (SAR) are powerful imaging sensors capable of monitoring various Earth phenomena day and night, regardless of weather conditions. The increasing global need for environmental and crisis monitoring and security surveillance has driven the advancement of innovative SAR concepts and data processing techniques. A wide coverage at a fine resolution is typically desired, which implies a huge data volume and highly complex processing techniques. Therefore, a high downlink capacity is required, and data processing and information extraction are typically conducted on the ground. Getting the necessary information quickly is a major challenge for state-of-the-art SAR systems, particularly for user-driven applications requiring near real-time monitoring and identification. Furthermore, all observed areas are imaged with a fixed configuration for each operating mode, employing an open loop system without scene-dependent feedback. The data acquisition and evaluation process relies on human oversight, which limits the flexibility typically required for evolving environments. To tackle these challenges, this doctoral thesis introduces an innovative concept named ADaptive OBserving Synthetic Aperture Radar (ADOB-SAR). The proposed concept aims to transform SAR into an intelligent sensor capable of performing targeted and selective data acquisition. It operates in two concurrent imaging modes managed by the same radar instrument to create a closed loop with the environment. The overall scene is broadly analyzed at low resolution and minimal performance and resource consumption. The sensor then autonomously adapts the illumination on the fly to monitor specific areas of interest with higher optimum accuracy. Integrating this system on a High-Altitude Platform (HAP) will facilitate a new era of smart sensors, providing persistent monitoring over regions of interest via knowledge-aided control. With the adaptive observing approach, only the relevant spots are imaged and saved, thus reducing the downlinked data volume. To cope with this adaptive multi-mode operation, a reconfigurable multi-channel phased array antenna architecture is designed. It features Digital Beamforming (DBF) and two-dimensional agile beam steering on transmit and receive. Novel hardware techniques are designed to enable the use of these DBF techniques with minimal antenna array complexity and to generate multiple simultaneous transmit beams to track multiple targets. The instrument can adjust the SAR parameters during flight and switch between modes using pulse-to-pulse interleaving. It utilizes temporal and angular orthogonality techniques, as well as waveform and frequency diversity to avoid transmit/receive and mutual receive interference across all concurrent modes. An adaptive scheduling algorithm is derived, which selects the optimal timing for transmitting new mode pulses and receiving their return echoes, considering the overall interleaved multi-target operation. This algorithm also automatically selects the most suitable orthogonality technique for each scenario. An adaptive decision-making model is implemented to determine the optimal combination of instrument configuration and parameters, adjusting system performance and minimizing resources based on extracted scene properties and user requirements. It is based on a dynamic optimization function capable of changing its framework in response to the observed scenario using adaptive target-specific beam and time allocation models, along with variable multi-objective priority settings. A reference system for a HAP was designed considering a maritime scenario, and a simulator was developed in the framework of this thesis to validate the ADOB-SAR concept and verify its dynamic decision-making and scheduling algorithms. The results show that the proposed sensor achieves a huge data reduction and efficient use of radar resources, notably lowering power consumption. The algorithms derived to adapt the instrument parameters and schedule the pulses can be applied to any application and operate onboard, while managing the additional computational complexity, and effectively meeting the demands of near real-time user-driven SAR applications.