A new coding scheme for image transmission over noisy channel is proposed.Similar to standard image compression, the scheme includes a linear transformfollowed by successive refinement scalar quantization. Unlike conventionalschemes, in the proposed system the quantized transform coefficients arelinearly mapped into channel symbols using systematic linear encoders. Thisfixed-to-fixed length "linear index coding" approach avoids the use of anexplicit entropy coding stage (e.g., arithmetic or Huffman coding), which istypically fragile to channel post-decoding residual errors. We use linear codesover GF(4), which are particularly suited for this application, since they arematched to the dead-zone quantizer symbol alphabet and to the QPSK modulationused on the deep-space communication channel. We optimize the proposed systemwhere the linear codes are systematic Raptor codes over GF(4). The ratelessproperty of Raptor encoders allows to achieve a "continuum" of coding rates, inorder to accurately match the channel coding rate to the transmission channelcapacity and to the quantized source entropy rate for each transform subbandand refinement level. Comparisons are provided with respect to theconcatenation of state-of-the-art image coding and channel coding schemes usedby Jet Propulsion Laboratories (JPL) for the Mars Exploration Rover (MER)Mission.
展开▼
