Modeling and hardware implementation of an amoeba-like cellular automaton

Area of Science:

• Computational Biology
• Biomimetic Computing
• Complex Systems

Background:

• Living organisms, like the slime mould Physarum polycephalum, excel at solving complex computational problems.
• Physarum polycephalum is known for solving graph and combinatorial problems.
• Cellular automata (CAs) model emergent computation where global behavior arises from local interactions.

Purpose of the Study:

• To develop a cellular automata (CA) model that accurately replicates the computational behavior of Physarum polycephalum.
• To apply this CA model to solve path-finding and path-planning problems.
• To evaluate the computational performance of the CA model compared to biological experiments and assess its potential as a virtual laboratory.

Main Methods:

• Development of a cellular automata (CA) model simulating Physarum polycephalum's behavior.
• Application of the CA model to find minimum-length paths in labyrinths.
• Utilizing the CA model for path-planning in adaptive networks, exemplified by Tokyo's rail network.
• Implementation of the CA model in both software and hardware.

Main Results:

• The CA model achieved results in strong agreement with experiments using the living organism.
• The CA model successfully solved labyrinth path-finding and complex path-planning problems.
• Hardware implementation of the CA demonstrated faster and more effective computation due to its parallel nature.

Conclusions:

• The developed CA model accurately simulates Physarum polycephalum's computational capabilities.
• The CA model serves as a powerful, low-cost virtual laboratory for studying slime mould computation.
• The inherent parallelism of the CA enables efficient and high-performance computation for complex problems.