Adaptive control of sampling rate using a local time-domain sampling theorem

When the information rate of a signal varies widely with time, we would like to vary the amount of storage, processing and transmission capacity devoted to the signal. This applies, for example, to remote sensing, satellite transmission, and image or speech coding. The practical problem has been how to determine the sampling rate from the data stream on-line. In this work, we give a local time-domain sampling theorem and show how to use it for adaptive control of the sampling rate. After initial sampling and quantization at a high rate, we estimate the required minimum sampling rate from a certain length of the digital signal. By employing an appropriate frame structure for data transmission or storage, we can then set the sampling rate to a minimum for each frame