In the past few years several applications of optimal experimental designs haveemerged to optimize the measurements of certain quantities in communication networks.The optimal design problems arising from this kind of applications sharethree interesting properties: (i) measurements are only available at a small numberof locations of the network; (ii) each monitor can simultaneously measure severalquantities, which can be modeled by "multiresponse experiments"; (iii) the observationmatrices depend on the topology of the network. In this paper, we give anoverview of these experimental design problems and recall recent results for the computationof optimal designs by second order cone programming (SOCP). New resultsfor the network-monitoring of a discrete time process are presented. In particular,we show that the optimal design problem for the monitoring of an AR1 process canbe reduced to the standard form and we give experimental results.
@article{151,
title = {Network-related problems in Optimal experimental design and second order Cone programming},
journal = {Tatra Mountains Mathematical Publications},
volume = {51},
year = {2012},
doi = {10.2478/tatra.v51i1.151},
language = {EN},
url = {http://dml.mathdoc.fr/item/151}
}
Sagnol, Guillaume. Network-related problems in Optimal experimental design and second order Cone programming. Tatra Mountains Mathematical Publications, Tome 51 (2012) . doi : 10.2478/tatra.v51i1.151. http://gdmltest.u-ga.fr/item/151/