Published online by Cambridge University Press: 19 April 2010
Applications of the theory of the optimal design of radial-velocity planet-search surveys are discussed. Two important practical problems are considered. The firstproblem is finding the time for future observations to yield the maximum improvement ofthe accuracy of exoplanetary orbital parameters and masses. In this case, the optimalscheduling rules are designed to maximize the determinant of the Fisher information matrix(the so-called D-optimality criterion). This method is asymptotically equivalent to themaximization of the expected gain of the Shannon information provided by making extraobservations. The second problem is finding the most favourable observing time fordistinguishing alternative orbital fits (the design of discriminating experiments). Inthis case, the optimal scheduling rules are designed to maximize the Kullback-Leiblerdivergence information. We also consider the potential efficiency of these methods of optimal planning of radialvelocity observations for multi-planet systems.