Error exponents for backhaul-constrained parallel relay networks

In this paper, we assess the random coding error exponents (EEs) corresponding to decode-and-forward (DF), compress-and-forward (CF) and quantize-and-forward (QF) relaying strategies for a parallel relay network (PRN), consisting of a single source and two relays. Moreover, through numerical analysis we show that the EEs achieved by using QF relaying along with non-Gaussian signaling (coded modulation, M-QAM) at the source and symbol-by-symbol uniform scalar quantizers (uSQs) at the relays is better than that achieved by DF and CF relaying strategies when the system is in the low signal-to-noise ratio (SNR) regime and the backhaul capacity is sufficient. This behavior is due to the structure of coded modulation, as opposed to Gaussian signaling, which leads to better EEs for simple relaying strategies compared to its more complex counterparts.