Traditionally, the suitability of radio receivers and similar devices for automotive use has been evaluated by evaluating their reception characteristics in relation to transmitted waves via repeated driving tests. This method of evaluation presents issues in terms of reproducibility and objectivity. A method of evaluating the suitability of FM receivers for vehicle fitting using a virtual propagation environment created on a PC (termed the Two-Stage method) has been developed in order to address these issues.The major challenge in the Two-Stage method is the creation of an actual propagation environment on a PC. A test-based incoming wave estimation technology able to accurately estimate the characteristics of actual propagation environments is therefore essential.The estimation of incoming FM waves necessitates large array antennas. In addition, the incoming waves become coherent multipath waves. It is challenging to separate multipath waves, and this represents an issue for accurate estimation.Seeking to address these challenges, the research discussed in this paper developed a Multiple Signal Classification (MUSIC) method using a virtual array antenna. The developed method has made it possible for the first time to accurately estimate arrival direction and delay time in an actual FM radio wave environment.Spatial smoothing preprocessing (SSP) and modified spatial smoothing preprocessing (MSSP) were applied to the virtual array antenna in order to boost performance in separating incoming waves. This made it possible to freely combine array element numbers and subarray element numbers, helping to enable decorrelation, and thus increasing the method's performance in separating multipath waves.