Modeling and hardware implementation of an amoeba-like cellular automaton

Abstract

Over the last few years, an increasing number of publications has shown that living organisms are very effective in finding solutions to complex mathematical problems which usually demand large computation resources. The plasmodium of the slime mould Physarum polycephalum is a successful example that has been used to solve path-finding problems on graphs and combinatorial problems. Cellular automata (CAs) computational model can capture the essential features of systems in which global behavior emerges from the collective effect of simple components, which interact locally (emergent computation). We developed a CA that models exactly the Physarum's behavior and applied it in finding the minimum-length path between two points in a labyrinth, as well as in solving a path-planning problem by guiding the development of adaptive networks, as in the case of the actual rail network of Tokyo. The CA results are in very good agreement with the computation results produced by the living organism experiments in both cases. Moreover, our CA hardware implementation results in faster and more effective computation performance, because of its inherent parallel nature. Consequently, our CA, implemented both in software and hardware, can serve as a powerful and low-cost virtual laboratory that models the slime mould Physarum's computation behavior.

A Physarum-inspired approach to supply chain network design

“…Increasing the number of states in automaton could make model more realistic however fine-granular parallelism with a large number of node states leads to increased time consumption during modelling. The time consumption could be reduced when cellular automata models of Physarum are implemented in hardware [62][63][64]. We did not consider cellular automata as truly alternative models because they usually focus on the morphology of the patterns and their qualitative dynamics rather then quantitative transitions that are characteristic for supply chains.…”

A Cellular Automata Bioinspired Algorithm Designing Data Trees in Wireless Sensor Networks

“…The different modeling approaches have, also, various implementations, from the purely spatial CA models to mathematical representations of flux canalization, oscillatory behavior, two-variable Oregonator model of Belousov-Zhabotinsky (BZ) medium, and path length as found in literature [16][17][18][20][21][22][23][24][25][26][27].…”

“…In this section a formal definition of a CA will be presented [16]. In general, a CA requires (1) a regular lattice of cells covering a portion of adimensional space;…”

Memristive Computing for NP-Hard AI Problems

“…In spite of the apparent simplicity of their structure, they have sufficient expressive dynamics to represent phenomena of arbitrary complexity. They have the ability to perform complex computations with a high degree of efficiency and robustness, as well as to model the behavior of complex systems found in nature [24,25,28,32]. Furthermore, they can easily handle complicated boundary and initial conditions, as well as anisotropies.…”

“…Owing to their potential to capture globally emerging behavior from collective interaction of simple and local components, CA have found successful application in several computational problems in physics, chemistry, biology, geology, and computer science, to name a few [20][21][22][23][24][25][26][27][28]. Additionally, when CA-based models are implemented in HW, the circuit design reduces to the design of a single cell and the overall layout results regular with exclusively local interconnections [14,29,30].…”