Abstract

In order to improve the performance of the wireless channel, collaborative communications has recently been proposed. In such a scenario, a source wishing to transmit a signal to a destination can be aided by an otherwise idle transmitter (labeled a relay). Due to the half duplex constraint, the relay node cannot transmit and receive at the same time. It has been shown that having a variable amount of time for which the relay will receive data, can provide further gains in collaborative communications. In this thesis we develop and study methods to implement the use of a variable time-fraction in collaborative communications. We study our proposed method under different channel state information scenarios. We design channel codes that allow for a relay to collaborate with a finite set of time-fractions. Through analysis, we provide design criteria that allow variable time-fraction collaborative codes to optimize their error rate performance. Our variable time-fraction collaborative codes are shown to approach the outage probability of collaborative channels. Furthermore, through the use of an upper bound on the error rate, we show the robustness of variable time-fraction collaboration over all relay locations when compared to traditional fixed time-fraction collaboration. Next we study the effect of imperfect channel state information at the receivers. It is shown that the effect of estimation errors is twofold in collaborative communications. Firstly the relay will have diminished collaborative capabilities and secondly, the destination suffers performance degradation. Assuming full and perfect channel state information at the transmitters, we design and study the use of power allocation algorithms ( PAA ). Our proposed optimal PAA ( OPAA ) optimizes the error rate performance of variable time-fraction collaborative communications. We also provide a more robust suboptimal PAA ( SPAA ) whose results suffer slight degradation compared to the OPAA . Lastly, we study the effect of imperfect channel state information at both the transmitters and the receivers. Our analysis shows the independence of the optimal number of pilot symbols to the location of the relay.