RIA: Vector Quantization for Noisy Channels

点击翻译按钮获取中文摘要

9309008 McLaughlin The Principal Investigator is analyzing and designing quantizers for the compression and transmission of analog signals such as speech, images, video and music across noisy channels such as satellite links or telephone lines. The research considers joint source-channel codes, which are essentially source codes designed so that their outputs can be transmitted directly across a noisy channel without channel coding. The investigator is developing an asymptotic theory, giving insight and direction into the design of locally and possibly globally optimum noisy channel quantizers. Source-channel codes then will be designed to achieve or exceed the overall system performance of the traditional approach of tandem source and channel coding. The award provides support for three years.