I. Introduction
The hypervolume indicator [1] is a well-known performance indicator in evolutionary multi-objective optimization (EMO). It has been widely-used for performance evaluation of EMO algorithms since it can evaluate their convergence and diversity performance simultaneously [2]. The ability to evaluate the convergence performance is due to its Pareto compliance property [3]. Better solution sets with respect to the Pareto dominance relation always have larger hypervolume values. The ability to evaluate the diversity performance is based on the following commonly-believed implicit assumption: A larger hypervolume value means a more diverse solution set. However, this issue is under-investigated in the EMO community.