We study a simple-harmonic-oscillator quantum computer solving oracle decision problems. We show that such computers can perform better by using nonorthogonal Gaussian wave functions rather than orthogonal top-hat wave functions as input to the information encoding process. Using the Deutsch-Jozsa problem as an example, we demonstrate that Gaussian modulation with optimized width parameter results in a lower error rate than for the top-hat encoding. We conclude that Gaussian modulation can allow for an improved trade-off between encoding, processing and measurement of the information.
@article{arxiv.1206.1035,
title = {Gaussian quantum computation with oracle-decision problems},
author = {Mark Adcock and Peter Hoyer and Barry C. Sanders},
journal= {arXiv preprint arXiv:1206.1035},
year = {2013}
}