Several gaps and errors in [1] are identified and corrected. While accommodating these corrections, a rigours proof is given that the successive convex approximation algorithm in [1] for secrecy rate maximization (SRM) does generate an increasing and bounded sequence of true secrecy rates and hence converges. It is further shown that its convergence point is a KKT point of the original SRM problem and, if the original problem is convex, this convergence point is globally-optimal, which is not necessarily the case in general. An interlacing property of the sequences of the true and approximate secrecy rates is established.
@article{arxiv.2010.12938,
title = {Comments on "Precoding and Artificial Noise Design for Cognitive MIMOME Wiretap Channels"},
author = {Mahdi Khojastehnia and Sergey Loyka},
journal= {arXiv preprint arXiv:2010.12938},
year = {2020}
}