LLR Compression for BICM Systems Using Large Constellations

Digital video broadcasting (DVB-C2) and other modern communication standards increase diversity by means of a symbol-level interleaver that spans over several codewords. Deinterleaving at the receiver requires a large memory, which has a significant impact on the implementation cost. In this paper, we propose a technique that reduces the de-interleaver memory size. By quantizing log-likehood ratios with bit-specific quantizers and compressing the quantized output, we can significantly redue the memory size with a negligible increase in computational complexity. Both the quantizer and compressor are designed via a GMI-based maximization procedure. For a typical DVB-C2 scenario, numerical results show that the proposed solution enables a memory saving up to 30%.