This thesis investigates the possibilities of Software-defined radios as a surveillance system by monitoring the electromagnetic spectrum. The surveillance system monitors the Bluetooth bandwidth with a multi-channel receiver that passively listens to Bluetooth packets. Furthermore, this thesis investigates the possibility of implementing an automatic k-means clustering algorithm to count unique devices in the vicinity. The Background explains the fundamental technologies used in Bluetooth and explains how devices communicate with each other. The Background also explains the proposed receiver architecture and its technologies. Section Related work and similar products investigate different approaches to detecting mobile devices and the effectiveness of the k-means algorithm. The Method explains how the receiver is modeled, how the Python script identifies Bluetooth packets in the bit stream, and how the physical imperfections are collected for the automatic k-means algorithm. Lastly, the Method explains how the labs were conducted. The Result section shows the performance of the receiver and the k-means algorithm. The Discussion section analyzes the results and discusses some design flaws and how to fix them potentially. Lastly, the Conclusion section compares the goals with the results and presents future work for further development.