Forwarding Strategies for Uplink Gaussian Multi-source Multi-relay Network

In this paper, we introduce the multi-source multi- relay channel model. In this model, the transmit signals of all users are received at multiple distinct relays. Each of these transmit signals can be either decoded or amplified at any one or a combination of the relays and then forward to the destination. For both decode-and-forward and amplify-and-forward strategies applied at the relays, we provide schemes to compute their achievable regions respectively. The achievable regions for both strategies are then compared by numerical simulations in both symmetric and asymmetric channel scenarios. Our numerical results show that when the power constraints of the relays are small, we can achieve larger rates by using decode-and-forward. Further, decode-and-forward is strictly optimal for the maximal sum rate since it can achieve the capacity outer bound. When the power constraints of the relays are large, amplify-and-forward outperforms decode-and-forward in general. For decode-and- forward, the optimal rates are achieved by forwarding the signals using the maximal power available at the relays. However, for amplify-and-forward, the power should be carefully adjusted to maximize the achievable rates.