Resource Allocation in OFDM Based Wireless Relay Networks ...

5.1 Introduction
Figure 5.2: Linear multi-hop network
Recently, the authors in [76]-[79] considered the resource allocation problem
in linear multi-hop networks. The so-called linear relay network consists of one
dimensional chain of nodes including a source node, a destination node, and several
intermediate relay nodes, as shown in Fig.5.2. It can be viewed as an important
special case of a sensor network where only a single route is active. As a natural
consequence of the multi-hop relaying and the OFDM transmission, the allocation of
per-hop resources is crucial in optimizing the end-to-end network performance. The
authors in [76] considered the optimum power allocation problem in linear multi-hop
networks where each relaying node uses AF transmission mode, where DF mode
of transmission was considered in [77]. The authors in [78] and [79] considered
the per hop transmission time and the power allocation to maximize the system
throughput and minimize the end-to-end outage, respectively. However, all the
works in [76]-to-[79] considered the optimization problem under the system-wide
power constraint. Such a power constraint may not be a good choice to see the
performance in practical systems where all the nodes are located distributively.
In multi-hop relay networks, AF relays are generally not used because of the
large noise enhancement. Hence, most literatures on resource allocation in multi-hop
relay networks consider only the DF relays [78]. However, the use of DF relays
requires energy consumption and time delay for decoding and encoding the message.
We consider a new network model that consists of mixed AF-DF relay nodes. This
new network structure halves the disadvantages of purely using AF or DF nodes.
For the time being, we focus on the simplest alternating structure, that is, each DF
relay node is preceded and followed by one AF relay node, as shown in Fig. 5.3.
In this chapter, we consider the lifetime maximization in AF-DF alternating
84