TABLE OF CONTENTS In the previous section we have studied the cases of structured networks, composed of users transmitting towards an access point (multiple access channel) or a base station transmitting towards many users (broadcast channel). In this section, we consider the more challenging case of an unstructured network, where there are Q sources communicating with as many destinations, without any intermediate access point. The fundamental question, in such a context, concerns the choice of the multiplexing strategy. Typically, the strategy is chosen a priori. However, this rigidity induces constraints that limit the possibility for the system to optimize the use of the available resources.
In this section, we show how to find out the coding and multiplexing strategy that maximizes the rates of all links, under a constraint on the average transmit power of each terminal. We start formulating the problem as a multiobjective optimization problem. Then, we show how to achieve suboptimal solutions by casting the problem as a game among players (radio nodes) that compete with each other to optimize their own utility functions.
We make a few assumptions, which are useful to simplify the problem. The assumptions are the following:
| a1 | All channels are modeled as FIR filters of maximum order L h; |
| a2 | Each terminal transmits blocks of length N + L, with a cyclic prefix of length L, where L incorporates the maximum channel order L h plus the relative delay between the... |
