Space-Based FMCW SAR Systems

Abstract

Monitoring critical structures such as oil pipe lines or dikes is essential for performing in-time and high-quality maintenance. Such structures cover a large area or are in abandoned places. Consequently, performing physical inspections by dedicated personnel is either impossible or very expensive. Alternatively such structures can be inspected by using satellite radar data. Synthetic Aperture Radar (SAR) data is already being used to create images of large areas. However, the current satellite SAR systems are big and require a lot of power, which makes their data expensive. In the TOFsat project the feasibility of a lightweight and low power FMCW SAR instrument for space is investigated. This work is part of the feasibility study. In the analysis of the power budget multiple feasible combinations for transmit power and antenna size where discussed, the preferred option is a option with 50~dBW transmit power and 6.25~m$^2$ antenna area. In our configuration of the radar ambiguities are present. The ambiguous projection of the ground directly under the satellite, called the nadir return, can distort the final image. However by carefully selecting the pulse repetition frequency the projection can be shifted outside the image of the area of interest. The utilization of FMCW on a SAR satellite introduces high demands on the isolation between the transmit chain and the receive chain. A single antenna solution on a single satellite with a circulator can not provide the required isolation. To avoid this problem a gating scheme was proposed. However a space based implementation of a gating scheme resulted in conflicting values of the gating frequency and pulse repetition frequency. Therefore gated FMCW did not resulted in a feasible solution. Having two separate antennas increases the isolation between the transmit chain and the receive chain. However from phase noise calculations it followed that still 100~dB of isolation was required. These high values of isolation are still difficult to obtain, together with the fact that two antennas on one satellite would have been required this led to a preference for a bi-static system. For a more accurate simulation of the system performance the far field antenna pattern was required. Multiple antenna options where considered, based on weight efficiency the slotted waveguide array was selected for this project. The available simulation tools within TNO did not take into account the mutual coupling of the slots of the array. To find the far field pattern first the slot voltages needed to be known. To compute the slot voltages a linear system of equations has been generated. After applying corrections to make the expressions convergent, a tool has been generated. With the obtained slot voltages the far field pattern can be generated. Resulting in a tool that is capable of simulating the far field pattern of a slotted waveguide array.