Wireless communication systems suffer from a phenomenon called small scale fading which causes rapid and unpredictable fluctuations in the received signal strengths. Traditional methods to mitigate small scale fading effects impose restrictions on the data exchange process and/or the used hardware which limit their applicability.
Recently, information theory has proven that cooperative diversity is an effective means to combat the effects of small scale fading. In such an approach common neighbors of a communication pair can act as relays for packets which would otherwise be lost due to bad channel conditions.
These studies, however, have mainly focused on physical layer aspects such as link capacity, coding, and relay positioning. In such studies, source, destination, and relaying nodes are known a priori. Setting up cooperation in a network has found less attention. In this thesis, we illustrate that obtaining theoretical gains of cooperative diversity is not straightforward and we address issues regarding networking and protocol aspects of cooperation. We first elaborate on the relay selection process. We propose and evaluate relay selection methods which increase the efficiency of cooperative diversity in terms of energy, time, and success. We show how to incorporate channel and routing information into the selection process to reduce the energy consumption and time overhead of cooperation. We illustrate how the relay selection depends on the knowledge of relay candidate cardinality. Such knowledge is also of interest in numerous other fields like medium access, routing, and information dissemination to name a few. To this end, we propose and evaluate methods based on probabilistic trials which aim to estimate the number of neighbors with a desired accuracy and minimum time overhead.
As a final step, we integrate our proposed mechanisms into the design of a novel medium access protocol which facilitates cooperative diversity by handling the cooperative packet flow, selecting a relay node, and making the necessary resource reservations. We evaluate the performance of the resulting system in terms of throughput and delay for different network parameters such as node density, channel coherence time, and data packet size. Our findings indicate that by using the proposed simple yet efficient mechanisms that are not restricted by hardware, we can deploy cooperative diversity with substantial gains in the large-scale wireless networks of the future consisting of low-end devices.