“…There is also an increasing interest in applying machine learning techniques to optimize the performance of QKD systems [5][6][7]. The security of practical implementations has also been addressed by devising hacking strategies [8][9][10], with their corresponding countermeasures, as well as by accounting for realistic imperfections and finite-key effects in the security analysis [2,11]. The feature issue also covers some less conventional protocols [12,13], relying on ambiguous state discrimination, as well as insight into multi-mode and high-dimensional QKD systems [14,15].…”