Abstract
In this thesis, we study the effect of channel estimation on the performance of distributed space-time coding (DSTC) in amplify-and-forward (AF) cooperative networks. The transmission is divided into two phases. In phase I, the source first transmits a block of data to the relays after which each relay will retransmit an encoded block of data to the destination in phase II. For AF networks, the signal transmitted by the source experiences a combined channel that consists of the multiplication of channel coefficients on the source-to-relay and the relay-to-destination links. The combined channel must be estimated at the destination in order to perform coherent detection. Based on the concept of TDM training, we propose two pilot insertion methods: the Source-Assisted (SA) Training Method and the Source-and-Relay Assisted (SRA) Training Method. In the SA method, pilot is generated only at the source while the relay amplifies-and-forwards the source's pilot as if it would a data symbol. In the SRA method, an additional pilot is generated by each relay in addition to relaying the source's pilot. The SRA allows the destination to estimate explicitly the relay-to-destination channel, but requires a larger block size in order to achieve a desired data rate. Given a total power constraint, a tradeoff exists in the amount of power that is allocated between pilot and data symbols. Therefore, by minimizing the pairwise error probability of a block fading approximation model, we also derive the optimal power allocation over pilot and data symbols. The performance of the proposed pilot insertion and power allocation strategies are compared through computer simulations.