On Information Rates over a Binary-Input Filtered Gaussian Channel

We consider the capacity of a band-limited additive white Gaussian noise channel, with bipolar inputs, a framework motivated by power-efficient faster than Nyquist signaling. We improve the available Ozarow-Wyner-Ziv (OWZ) lower bound on capacity by introducing a new communication scheme, exhibiting two advantages over the OWZ approach based on peak-power constrained pulse-amplitude modulation. We demonstrate analytically, a moderately improved information rate of the suggested technique, exhibiting also considerably less sign transitions of the bipolar signal. The gap between the known upper-bound based on spectral constrains of bipolar signals and the new achievable lower bound is reduced to 1.1 bits per Nyquist interval at the high signal-to-noise ratio regime.