DMT of Multi-hop Cooperative Networks - Part II: Half-Duplex Networks with Full-Duplex Performance

TL;DR

This paper analyzes the diversity-multiplexing tradeoff (DMT) in multi-hop cooperative networks with half-duplex relays, demonstrating that full-duplex performance can be approximated and that simple amplify-and-forward protocols are often optimal.

Contribution

It characterizes the DMT of various half-duplex relay network families and shows they can achieve the cut-set bound using explicit amplify-and-forward protocols.

Findings

Half-duplex networks can achieve full-duplex performance in terms of DMT.

Explicit protocols using amplify-and-forward relaying are optimal for these networks.

A linear DMT tradeoff is achievable in layered networks.

Abstract

We consider single-source single-sink (ss-ss) multi-hop relay networks, with slow-fading links and single-antenna half-duplex relay nodes. In a companion paper, we established some basic results which laid the foundation for the results presented here. In the present paper, we consider two families of networks of half-duplex networks. KPP networks may be viewed as the union of K node-disjoint parallel relaying paths. Generalizations of these networks include KPP(I) networks, which permit interference between paths and KPP(D) networks, which possess a direct link between source and sink. We characterize the DMT of these families of networks completely and show that they can achieve the cut-set bound, thus proving that full-duplex performance can be obtained even in the presence of the half-duplex constraint. We then consider layered networks, and prove that a linear DMT between maximum diversity and maximum multiplexing gain is achievable. All protocols in this paper are explicit and use only amplify-and-forward relaying. We also construct codes that achieve the optimal DMT for all the proposed schemes. Two key implications of the results in the paper are that the half-duplex constraint does not entail any rate loss for a large class of cooperative networks and that AF protocols are often optimal.