On Compress and Forward with Multiple Carriers in the 3-Node Relay Channel Exploiting Information Bottleneck Graphs

This paper investigates the design of practical quantizers for the 3-node relay channel applying Compress-and-Forward (CF) over frequency selective channels. As this is a crucial task, a low complexity algorithm is designed by the help of Information Bottleneck Graphs (IBGs). In principle, the Information Bottleneck (IB) method is a powerful tool to obtain the optimal quantizer for a specific setup. In a simple Additive White Gaussian Noise (AWGN) channel model, the quantization design depends only on the signal to noise ratio (SNR) so that the IB algorithm can be applied straightforward. This paper however, considers transmission over frequency selective channels which are decomposed info flat fading channels via Orthogonal Frequency Division Multiplexing (OFDM). The dependence of quantization on the specific channel coefficient of each subcarrier makes the problem much more complex. Thus, low-complexity approximations are suggested to design cascaded quantizers by the help of IBGs. Doing so makes the problem feasible and shows promising results in a state of the art Bit Interleaved Coded Modulation (BICM) scheme.