Minimum Expected Distortion in Gaussian Layered Broadcast Coding with Successive Refinement - Computer Science Information TheoryReport as inadecuate




Minimum Expected Distortion in Gaussian Layered Broadcast Coding with Successive Refinement - Computer Science Information Theory - Download this document for free, or read online. Document in PDF available to download.

Abstract: A transmitter without channel state information CSI wishes to send adelay-limited Gaussian source over a slowly fading channel. The source is codedin superimposed layers, with each layer successively refining the descriptionin the previous one. The receiver decodes the layers that are supported by thechannel realization and reconstructs the source up to a distortion. In thelimit of a continuum of infinite layers, the optimal power distribution thatminimizes the expected distortion is given by the solution to a set of lineardifferential equations in terms of the density of the fading distribution. Inthe optimal power distribution, as SNR increases, the allocation over thehigher layers remains unchanged; rather the extra power is allocated towardsthe lower layers. On the other hand, as the bandwidth ratio b channel uses persource symbol tends to zero, the power distribution that minimizes expecteddistortion converges to the power distribution that maximizes expectedcapacity. While expected distortion can be improved by acquiring CSI at thetransmitter CSIT or by increasing diversity from the realization ofindependent fading paths, at high SNR the performance benefit from diversityexceeds that from CSIT, especially when b is large.



Author: Chris T. K. Ng, Deniz Gunduz, Andrea J. Goldsmith, Elza Erkip

Source: https://arxiv.org/







Related documents