Objective functions for information-theoretical monitoring network design: what is “optimal”?

Abstract. This paper concerns the problem of optimal monitoring network layout using information-theoretical methods. Numerous different objectives based on information measures have been proposed in recent literature, often focusing simultaneously maximum and minimum dependence between chosen locations for data collection stations. We discuss these objective functions conclude that a single-objective optimization joint entropy suffices to maximize given number argue widespread notion minimizing redundancy, or monitored signals, as secondary is not desirable has no intrinsic justification. The negative effect redundancy total collected already accounted entropy, which net any redundancies. In fact, two networks equal one with higher amount redundant should be preferred reasons robustness against failure. attaining objective, we investigate exhaustive optimization, more computationally tractable greedy approach adds station at time, introduce “greedy drop” approach, where full set stations reduced time. show can exist guaranteed reach global optimum.