Microphone array signal processing for advancements in robust speech systems