Methods for increased resolution of targets in urban environment

This report concerns the possibility to increase the resolution of an urban target within the field of view of a radar, using some different methods. An increased resolution would give improved imaging and possibility for more accurate positioning. In the first method, the possibility to use a priori knowledge of an urban scene (geometry, material) is studied to bring about the improvement. The scene that has been studied in this initial work consists of a long, straight street, or corridor. A radar consisting of several linearly distributed antennas has been simulated in various computation scenarios. The simulations, at Xband, show that increased resolution can be obtained if a priori knowledge about the scene is used for accurate estimation of the wave propagation between the transmitter and the receiver (so-called transfer function). This is then used to generate a reference signal that is correlated with a measured or calculated return from the scene. The results are sensitive to deviations in the determination of the estimated transfer function. Furthermore, there is an indication that the more accurate the transfer function, the higher the resolution obtained. A scenario has been studied, consisting of a modelled human situated in a corridor. A somewhat better resolution of the human is obtained if the multipath propagation is taken into account in the transfer function. Considerable differences have been found, depending on the choice of polarization for this scenario, and this has to be investigated further. Another interesting method is discussed for increasing the resolution in low-resolution range profiles. The considered method is called the "Cetin method" in this report, after an author name in the original article. An initial study is presented of a third method for improved target measurement, viz. MIMO radar, which is a generalization of multistatic radar and phase-controlled radar systems. The difference is primarily due to the transmission of individually modulated signals from each transmitter, which entails, e.g., that they can be transmitted simultaneously. This enables adaptive systems where transmitted signals can be adjusted to give maximum performance of the determination of characteristics of chosen objects in the scene. Another interesting property of MIMO systems with sufficiently separated antennas is that the precision of the position measurement of an object is only limited by the wave-length of the signal instead of the bandwidth, as in mono-static systems. Furthermore, the number of targets that can be measured with a MIMO system with co-located antennas increases drastically with the number of independent transmitters used. The resolution of the system can also be increased linearly with the number of transmitters by the creation of virtually enlarged apertures. An interesting method considered in connection with the MIMO radar study is Compressive Sensing, which is a technique aimed at radically reducing the data volume being collected by a sensor system, without significantly lowering the quality of the reconstructed signal or image. The technique takes as its starting point that a scene as a rule is sparse, i.e., it contains parts with no objects and the signal, hence, only consists of measurement noise. Compressive Sensing forms a complement to MIMO radar in a very attractive way, since MIMO radar is dependent on communication of raw data between sensor nodes. By then reducing the data that needs to be sent without significant loss of information, MIMO radar systems can be built without unreasonable demands on the communication systems. Furthermore, Compressive Sensing can also give improved image quality in scenes with high dynamic range, because of reduction of side-lobe patterns.