Integrating machine learning techniques in quantum communication to characterize the quantum channel

Abstract

Free-space quantum communication technology has made significant advancements over the years. However, to achieve a global quantum network there are still obstacles to overcome. To date, free-space quantum communication channels have been faced with challenges related to losses in the quantum channel, security, and low data rates. Classical machine learning techniques provide a resourceful method for determining the properties of the free-space quantum channel. In this paper, we used supervised machine learning to predict the atmospheric strength of a free-space quantum channel in the form of the Strehl ratio. We show that using the ensemble technique called random forest we could make predictions of the Strehl ratio of the quantum channel with a mean absolute percentage error of 4.44%. By adding the feature of fidelity to the training set before and after the quantum channel, we were able to predict the Strehl ratio with an improved mean absolute percentage error of 3.86%. For comparison we used other machine learning techniques such as linear regression and support vector machine with various kernels.