The three-node wireless network: achievable rates and Cooperation strategies

Abstract

We consider a wireless network composed of three nodes and limited by the half-duplex and total power constraints. This formulation encompasses many of the special cases studied in the literature and allows for capturing the common features shared by them. Here, we focus on three special cases, namely, 1) relay channel, 2) multicast channel, and 3) three-way channel. These special cases are judicially chosen to reflect varying degrees of complexity while highlighting the common ground shared by the different variants of the three-node wireless network. For the relay channel, we propose a new cooperation scheme that exploits the wireless feedback gain. This scheme combines the benefits of the decode-and-forward (DF) and compress-and-forward (CF) strategies and avoids the noiseless feedback assumption adopted in earlier works. Our analysis of the achievable rate of this scheme reveals the diminishing feedback gain in both the low and high signal-to-noise ratio (SNR) regimes. Inspired by the proposed feedback strategy, we identify a greedy cooperation framework applicable to both the multicast and three-way channels. Our performance analysis reveals the asymptotic optimality of the proposed greedy approach and the central role of list source-channel decoding in exploiting the receiver side information in the wireless network setting.