Cooperative Diversity: Capacity Bounds and Code Designs - Adaptive Signal Processing in Wireless Communications - page 338

In contrast to a traditional infrastructure wireless network (e.g., a cellular network),
where information is transmitted from one user to another via a control base station, an
ad hoc network allows peer-to-peer communication from a sending node to a destina-
tion node. That is, the information can travel directly from a sending to receiving node
in a single hop.
However, since wireless channels are often poor, single-hop routing requires either
high transmission power, and consequently causes increased interference, or complex
multiple access schemes. To achieve significant power savings and keep complexity low,
information should be conveyed to a destination through multiple intermediate nodes.
Whereas transmission over a single-hop channel has already been intensively studied
and is well understood, cooperative communication in multinode networks is still an
open research problem, which recently has received considerable attention, inspired by
the papers [1, 2].
Communication over a wireless channel is limited by interference, fading, multipath,
path loss, and shadowing. The main design challenge in ad hoc networks lies in devising
communication methodologies in a decentralized manner, based on the knowledge of
local conditions only, to overcome these limitations. An additional design issue has to
do with the high dynamics of an ad hoc network, where nodes frequently join and leave
the network.
One way of achieving high performance is to employ multiple transmitter and receiver
antennas at nodes. This multiple antenna system increases the capacity and improves
robustness to fading and interference by means of spatial diversity and data rate multi-
plexing [3, 4]. However, building multiple antennas at each node can be expensive,
impractical, and often infeasible, especially for small and simple nodes such as those
used in sensor networks.
Another recently proposed solution for achieving spatial diversity without requir-
ing multiple antennas at any node is cooperative diversity [1, 2]. It is based on grouping
several nodes (each with only one antenna) together into a cluster to form a large trans-
mit or receive antenna array. Collaborative clusters are formed in an ad hoc fashion
by negotiations among neighboring nodes without centralized control (see Figure 12.1 ).
Cooperative diversity naturally arises in ad hoc networks as it enables great power sav-
ings with cheap, simple, and mobile nodes, while supporting decentralized routing and
control algorithms. However, it is not limited to ad hoc networks, as it can be useful in
infrastructure networks as well.
The simplest nontrivial setup is when the nodes form pairs, i.e., clusters of two. In
a two-transmitter two-receiver cooperative channel, the two single-antenna transmit-
ters want to communicate messages to the two remote single-antenna receivers over
the same wireless radio channel. In transmitter cooperation , the two transmitters first
exchange their messages, and then start to act as a single two-antenna broadcast trans-
mitter. On the other hand, in receiver cooperation , the two receivers exchange their
received signals and act as a single two-antenna multiple access receiver. In general, the
two transmitters as well as the two receivers can collaborate among each other to form a
virtual multiple-input multiple-output (MIMO) channel with two transmitter and two
receiver antennas. The main goal of node cooperation is to achieve spatial diversity and