The intracellular circadian clock mechanisms are known to affect various substantial cellular machinery such as cell cycle progression, inflammatory response, apoptosis, and DNA repair. Cancer growth in various tissues is still under circadian control, which may be at least partly underlain by the survived connections between the intracellular machinery and the clock. Considering such findings, chronotherapy has been applied to cancer treatments, in which anti-cancer drugs are administered in scheduled circadian times so as to resolve the trade-off between damages against the normal and cancer cells. However, any effective administration strategy has not yet been established especially in a quantitative sense. In this study, we develop an automaton model of cell division cycle interacting with circadian clock and suffering from a probability of cell death. A cancer cell is modeled by shortening/ lengthening the cell division interval and a transition to motility state under starving condition. Population proliferating dynamics in 3D space are simulated under the diffusion of nutrient factor and the anti-cancer drugs from a vessel. The simulation results show that the drug administration schedule could differentiate the damages against proliferation of normal and cancer cells. This implies the existence of optimal timing for the drug administration, which could provide an efficient strategy of chronotherapeutic treatment of cancer.