Nowadays, localization of cars is mostly done using GPS. Some major disadvantages of this method are that it is not very accurate and that it cannot be used indoors. These disadvantages do not form a major problem when an active human driver is controlling the car, but will become more critical when the autonomy of vehicles increases. The goal of this project is to use Synthetic Aperture Radar techniques to overcome the problems of GPS. To achieve this, different imaging algorithms have been implemented, and a single method has been extended to directly accommodate the road-mapping application. Different measurements were done to test the algorithms and the results were promising. It is hoped that the use of this technology will become wide-spread and that it can help with creating a safer and more efficient world of traffic, where increased self-driving capabilities are facilitated.

For radar applications, MIMO is often used. The problem is that the amount of available transmit- and receive channels is limited. If each channel is connected to a single antenna element, the effective aperture is small, and therefore the beamwidth is large and the angular resolution is limited. In this thesis, a novel scheme is proposed that circumvents this problem. More details are given in the full thesis. The result is an array that covers a reduced part of the visible space, with discrete beams that have improved beamwidth. The system gives the possibility to improve the angular resolution at the cost of field of view. A CST-model is made to verify this concept. In the specific example given in this thesis, the half-power beamwidth is improved from 8.4 degrees at broadside, to 5.7 degrees at broadside. At the same time the coverage is reduced from a theoretical ±90 degrees, to ±50 degrees.