Sumetpal Singh Vladislav Tadic Arnaud Doucet A policy gradient method for SMDPs with application to call admission control. 1268-1274 2002 ICARCV http://dx.doi.org/10.1109/ICARCV.2002.1234955 conf/icarcv/2002 db/conf/icarcv/icarcv2002.html#SinghTD02