This thesis describes architecture of IP telephony networks based on signaling protocol SIP and transport protocol RTP. Also tools used for analyzing VoIP traffic are described. Main objective is to design and implement a system for the detection of calls and extraction of voice payloads from  the captured packets. The system first recognizes the signaling messages of SIP protocol. These messages are analyzed afterwards. Output statistics are generated based on gathered data. Voice data will be stored in form, which is suitable for further processing.