Combining Source and Channel Coding for Transmission of PVQ Compressed Images over Gaussian and Rayleigh Fading Channels

Kurzfassung

An important advantage of the fixed rate source coding scheme DCT and Product Pyramid VQ with (L-K)-thresholding is, besides its very good performance, the ability to calculate very precisely the expected mean square error caused by data compression mses. This together with the a-priori ability to calculate the bit-sensitivity of the compressed image data enables us to perform a highly efficientunequal error protection for image transmission over noisy channels. We show how to optimize the ratio of source to channel rate for transmission of compressed still images over a Gaussian channel (and a Rayleigh fading channel) using unequal error protection, and we present gains of up to 6 dB (7 dB) in PSNR compared to the image compression standard JPEG.