Reviewing Speaker Recognition Influence in Strengthening VOIP Caller Authentication

This paper focuses on authentication issues and user ID detection depending on the voice signature of the incomingVoice over IP (VoIP) calls. Firstly, a complete speaker recognition system based on LPC feature extraction and decision basedon Artificial Neural Network was described and built using MATLAB package. Then, MATLAB package was used to develop asimplified prototype version of the modified SIP protocol which is used to manage calls establishment among the networknodes. These codes were written for both SIP server and client nodes. The modified SIP protocol was implemented on anexperimental network which is built to demonstrate the operation of the enhanced authentication procedure.