Mobile　data　generated　by　a　mobile　phone　in　a　GSM　network　can　indirectly　reflect　the　change　of　a　user’s　positions.　However,　it　is　difficult　to　get　these　data　due　to　the　privacy　issue.　The　generation　of　simulated　mobile　data　provides　an　alternate　method　to　solve　the　problem.　This　paper　designs　a　system　called　a　generation　system　of　mobile　data　based　on　real　user　behavior.　This　system　can　simulate　the　communication　events　to　generate　mobile　data　and　extract　the　trajectory　points　of　users’　past　activities　with　some　appropriate　semantic　annotations.　©　Springer-Verlag　Berlin　Heidelberg　2014.