Rate-Compatible Low-Density Parity-Check (RC-LDPC) codes are proposed to provide unequal error protection (UEP) for robust and efficient transmission of progressively compressed images over noisy channels. The total bit budget is partitioned between the source and channel coding, and the optimal rates are assigned to source encoded packets with a trellis-based rate-allocation algorithm using a forward error correction (FEC) scheme. We propose an improved rate-allocation algorithm by which the speed could be increased up to about three orders of magnitude due to the reduction in the complexity of the trellis. Simulation results based on Joint Photographic Experts Group (JPEG) 2000 and set partitioning in hierarchical trees (SPIHT) source encoders show that the proposed FEC schemes over binary symmetric channel (BSC) outperforms similar schemes based on turbo codes and irregular repeat-accumulate codes by up to about 1.1 dB and 1 dB in the expected peak signal-to-noise ratio (PSNR) of reconstructed images. Moreover, the proposed FEC scheme achieves up to about 1.7 dB and 3 dB improvements over EEP and UEP with product codes respectively over Rayleigh fading channel. For applications where delay is not a major concern, we present an alternative UEP scheme using automatic repeat request (ARQ), where parity check bits are incrementally appended until source packets are correctly recovered.
展开▼
