Design And Simulation Of Phase Synchrony Based Physical-Layer Network Coding

Network coding brings throughput improvement for wireless networks, by encoding data flows at the intermediate nodes. To encode naturally, physical-layer network coding (PNC) employs superposition of electromagnetic waves, allowing wireless signals from end nodes overlapping each other at the relay. Therefore, the throughput can be further improved.This thesis focuses on the practical issues on PNC and studies phase-synchronous PNC, which includes developing a phase synchronization scheme for PNC and analyzing the average error performance of M-ary quadrature amplitude modulation (M-QAM) based PNC with synchronization errors. Then, the thesis further studies interactions between the synchronization scheme and data transmission.Currently, the implementation of PNC is still a work in progress, and phase synchrony is significant but difficult. This thesis firstly proposes a synchronization scheme for PNC, and then studies the synchronization accuracy, when adopting frequency and phase estimation based on maximum likelihood estimate (MLE) and phase-locked loop (LLP), respectively.To associate the accuracy with average symbol error rate, the thesis investigates symbol error rate for PNC with deterministic phase errors. Based on the results for deterministic phase errors, by formulating the phase error model for the proposed scheme, the thesis derives the average symbol error rate for accumulative phase errors. Numerical results indicate that the analytical result is accurate enough for further studies.Considering the proposed scheme, the thesis studies the synchronization overhead. The thesis constructs the analytical relationship among overhead, accuracy, average symbol error rate, and the goodput, and then proposes an optimized scheme. Simulation results reveal that phase-synchronous PNC can outperform conventional network coding; the optimized scheme can make the throughput gain approach to theoretical upper bound.