Cooperative Transmission Techniques in Wireless Communication Networks

Abstract

Cooperative communication networks have received significant interests from both academia and industry in the past decade due to its ability to provide spatial diversity without the need of implementing multiple transmit and/or receive antennas at the end-user terminals. These new communication networks have inspired novel ideas and approaches to find out what and how performance improvement can be provided with cooperative communications. The objective of this thesis is to design and analyze various cooperative transmission techniques under the two common relaying signal processing methods, namely decode-and-forward (DF) and amplify-and-forward (AF).

For the DF method, the thesis focuses on providing performance improvement by mitigating detection errors at the relay(s). In particular, the relaying action is implemented adaptively to reduce the phenomenon of error propagation: whether or not a relay’s decision to retransmit depends on its decision variable and a predefined threshold. First, under the scenario that unequal error protection is employed to transmit different information classes at the source, a relaying protocol in a singlerelay network is proposed and its error performance is evaluated. It is shown that by setting the optimal signal-to-noise ratio (SNR) thresholds at the relay for different information classes, the overall error performance can be significantly improved. Second, for multiple-relay networks, a relay selection protocol, also based on SNR thresholds, is proposed and the optimal thresholds are also provided. Third, an adaptive relaying protocol and a low-complexity receiver are proposed when binary frequency-shift-keying (FSK) modulation is employed and neither the receiver nor the transmitter knows the fading coefficients. It is demonstrated that large performance improvements are possible when the optimal thresholds are implemented at the relays and destination. Finally, under the scenario that there is information feedback from the destination to the relays, a novel protocol is developed to achieve the maximum transmission throughput over a multiple-relay network while the bit-error rate satisfies a given constraint.

With the AF method, the thesis examines a fixed-gain multiple-relay network in which the channels are temporally-correlated Rayleigh flat fading. Developed is a general framework for maximum-ratio-combining detection when M-FSK modulation is used and no channel state information is available at the destination. In particular, an upper-bound expression on the system’s error performance is derived and used to verify that the system achieves the maximal diversity order. Simulation results demonstrate that the proposed scheme outperforms the existing schemes for the multiple-relay network under consideration.