Towards the Optimal Amplify-and-Forward Cooperative Diversity Scheme

TL;DR

This paper introduces a sequential slotted amplify-and-forward scheme that approaches the transmit diversity bound in slow fading channels, outperforming previous AF schemes especially at high multiplexing gains and larger networks.

Contribution

It proposes a novel sequential SAF scheme that achieves the DMT upper bound and improves diversity gains over existing AF schemes in high multiplexing regimes.

Findings

Sequential SAF scheme achieves the DMT upper bound as M increases.

The three-slot SAF scheme outperforms NAF schemes for two relays at certain multiplexing gains.

Numerical results show significant gains in spectral efficiency and network size regimes.

Abstract

In a slow fading channel, how to find a cooperative diversity scheme that achieves the transmit diversity bound is still an open problem. In fact, all previously proposed amplify-and-forward (AF) and decode-and-forward (DF) schemes do not improve with the number of relays in terms of the diversity multiplexing tradeoff (DMT) for multiplexing gains r higher than 0.5. In this work, we study the class of slotted amplify-and-forward (SAF) schemes. We first establish an upper bound on the DMT for any SAF scheme with an arbitrary number of relays N and number of slots M. Then, we propose a sequential SAF scheme that can exploit the potential diversity gain in the high multiplexing gain regime. More precisely, in certain conditions, the sequential SAF scheme achieves the proposed DMT upper bound which tends to the transmit diversity bound when M goes to infinity. In particular, for the two-relay case, the three-slot sequential SAF scheme achieves the proposed upper bound and outperforms the two-relay non-orthorgonal amplify-and-forward (NAF) scheme of Azarian et al. for multiplexing gains r < 2/3. Numerical results reveal a significant gain of our scheme over the previously proposed AF schemes, especially in high spectral efficiency and large network size regime.