Waveform Design for Wiener Phase Noise Channels and Multi-Sample Receivers

We consider the problem of optimizing the pulse shape for a channel affected by Wiener phase noise and with multi-sample receiver that takes into account the effects of frontend filtering. For a specific pulse shape, the mutual information rate is evaluated with lower-bounding techniques which give tight bounds. Our contribution is to provide understanding of the interplay of information rates, spectrum constraints, and multi-sample receivers. We employ efficient numerical techniques that ensure the robustness of our optimization results. Finally we compare with previous results and show that considerable improvements in information rates and spectral efficiency can be achieved with judicious choices of the transmitted waveform.