Analysis of Cooperative Communication and Link Layer Performance under Imperfectly Known Fading Channel Gains

One of the main issues of wireless communication systems is to cope with random variationof their channel conditions. In many wireless communication systems, the receiver estimatesthe channel for symbol detection. It is usually assumed that channel estimation is perfect,but in practice, channel estimation error (CEE) can become signicant and can degrade theperformance of wireless communication systems. This thesis addresses the eects of channelestimation error in two contexts: CEE in cooperative communication systems and CEE inlink layer performance.Cooperative diversity, which has been recently presented as an eective way of mitigatingthe eect of deep fades in wireless channels and improving spectral and power eciencyof the wireless networks, is considered in the rst part of this dissertation. Taking intoconsideration the presence of CEE, part I of this thesis analyzes a few cooperative communicationsystem models, which display dierent levels of practicality and represent largeclasses of cooperative systems in the literature. This thesis spells out delineating aspectsof these models and rigorously compares their error probabilities. Furthermore, a novelsignal detection scheme in the presence of CEE is presented for large classes of cooperativecommunication systems.Part II of this dissertation focuses on cross layer issues between the physical and linklayers of wireless communication systems. In particular, the frame error probability (FEP)is derived for a wireless communication system over fading channels in the presence ofCEE. Part II also explores the issue of optimizing bit transmission power for minimizing theexpected energy required to reliably deliver a frame to the destination node through an ARQmechanism over a fading channel. Also, an optimization algorithm is designed to minimizethe expected energy for reliable delivery of a frame for the systems under consideration.