Automatic speech recognition over error-prone wireless networks

This paper concludes with a discussion of applicability of different techniques based on the channel characteristics and the system requirements.