A step frequency Radar for through-the-wall radar imaging applications, based on a vector network analyzer (VNA) and on a Multiple Input Single Output (MISO) architecture, is presented. The system was calibrated by removing the errors due to the connection cables, to the time delay introduced by the antennas and by the power amplifier. Using the Radar, the position of metallic targets both in the absence and in the presence of a wall was accurately identified. The data acquired by the Radar were processed using 4 algorithms present in the literature and a new FT-MUSIC algorithm developed by the Authors. All the considered algorithms were able to accurately reconstruct the position of the investigated targets both in the absence and in the presence of a wall.