Future Directions for Microphone Arrays

Abstract

The telephone, invented more than 140 years ago, has changed very little with respect to acoustic design in the last 40 years. The most recent technological improvements were the replacement of the carbon microphone with an electret microphone and the development of the speakerphone for hands-free telephony in the early sixties. It might seem quite surprising that in spite of the rapid changes in communication systems: digital switching, digital wireless, packet-based data communications and the like, that very little has changed to the acoustic front-end in telephony systems. For instance, with the general availability of broadband communication networks, why have we not significantly increased the audio communication bandwidth? Why don’t we see stereo or multichannel in telephony and communication systems? Why don’t we see more ubiquitous adoption of voice recognition in consumer products? Why can’t we tailor the audio response of our communication systems the same way we do with our entertainment systems? Why do our cell phones sound so tinny, distorted, and noisy? Why is it so difficult to understand people who call us from noisy environments or call us in noisy environments? These are good questions when one ponders the explosion of inexpensive digital signal processing integrated circuits combined with the current bandwidth of communication systems and those that are about to be introduced. Clearly front-end acoustic signal processing will play a large role in addressing these questions.