The previous topic examined cooperative strategies when no fading or fast fading is present. The main performance metric was rate because long codes can average out the effects of noise and fading and can make the error probability approach zero. In this topic, we consider another extreme, namely, slow fading where the channel gains are random but are held constant for the duration of a codeword. For example, if a channel is in a deep fade then for high rates one cannot avoid making errors and we say there is an outage [54, 144]. We wish to characterize the tradeoff between rate and outage probability. We again assume that the nodes have CSIR (see Sections 3.3, 4.1, and 4.2.1).
Outage probability can be reduced by means of diversity, i.e., transmitting signals carrying the same information over different paths in time, frequency, or space. For example, if ARQ feedback is available, then one can create diversity with Hybrid-ARQ (see Section 2.3). One of our aims is to create diversity through node cooperation.
Transmit cooperation has nodes using DF to exchange each other’s messages. By sharing resources, the nodes create two paths to transmit their information (see Figure 3.1) and this is known as cooperation diversity [162, 163] or cooperative diversity [112, 114]. Receive cooperation has nodes forwarding information about their observations. For example, the nodes can use CF. A system with both transmit and receive cooperation resembles a multi-antenna or multiple-input, multiple-output (MIMO) system, and it is therefore sometimes called a distributed MIMO system. It is known that multiple antennas can increase capacity without sacrificing bandwidth or energy [53, 174]. For instance, if the path gains between the individual transmit and receive antennas fade independently, high data rates are achieved by sending different data streams over the independent channels, an approach known as spatial multiplexing .
Distributed MIMO systems can realize some of the usual MIMO gains. However, the absence of a high-capacity link (e.g., a cable) between the antennas limits the gains in several ways.
• The messages are known initially only to the source nodes.
• Resources such as power, bandwidth and time (delay) must be expended to enable cooperation.
• Antenna power allocation cannot be performed as in conventional MIMO systems.
• Nodes may have half-duplex constraints.
• Synchronizing signals sent from distributed antennas is more difficult than in a conventional MIMO transmitter.
The effects of such limitations will become apparent once we analyze various cooperative protocols.