Siva: Artificial Life Based on Programmed

Oohashi et al. proposed a biological hypothesis \programmed self-decomposition (PSD) model", in 1987, which assumes that a mechanism of decomposing itself and restoring substances and space to the environment is intrinsically programmed in every cell of terrestrial life. I have previously developed a simulator SIVA (Simulator for Individuals of Virtual Automata) and examined the e ects of the self-decomposition mechanism through computer simulations of an ecosystem of arti cial life (ALife) based on the PSD model. In this thesis I report SIVA-2 implemented anew on a parallel computer AP1000+. SIVA-2 was provided by this parallelization with the ability to simulate about 26 times as large space as before within a practical period of time. I observed using SIVA-2 how the evolutionary adaptation of life advanced under various conditions with di erent probabilities of mutation. The results of these experiments indicated that the mechanism of selfdecomposition could be considered to have the function to keep the evolutionary adaptability of life high, even though a probability of mutation was small. Several problems remained in SIVA-2 will be solved in a new simulator SIVA-3 which is now being designed.