Skip to main content

Research Repository

Advanced Search

Physarum in silicon: the Greek motorways study

Tsompanas, Michail Antisthenis I.; Sirakoulis, Georgios Ch; Adamatzky, Andrew I.; Tsompanas, Michail-Antisthenis; Sirakoulis, Georgios; Adamatzky, Andrew

Authors

Michail Antisthenis I. Tsompanas

Georgios Ch Sirakoulis

Andrew I. Adamatzky

Georgios Sirakoulis



Abstract

© 2014, Springer Science+Business Media Dordrecht. Physarum polycephalum has repeatedly, during the last decade, demonstrated that has unexpected computing abilities. While the plasmodium of P. polycephalum can effectively solve several geographical described problems, like evaluating human–made transport networks, a disadvantage of a biological computer, like the aforementioned is directly apparent; the great amount of time needed to provide results. Thus, the main focus of this paper is the enhancement of the time efficiency of the biological computer by using conventional computers or even digital circuitry. Cellular automata (CA) as a powerful computational tool has been selected to tackle with these difficulties and a software (Matlab) CA model is used to produce results in shorter time periods. While the duration of a laboratory experiment is occasionally from 3 to 5 days, the CA model, for a specific configuration, needs around 40s. In order to achieve a further acceleration of the computation, a hardware implementation of the corresponding CA software based model is proposed here, taking full advantage of the CA inherent parallelism, uniformity and the locality of interconnections. Consequently, the digital circuit designed can be used as a massively parallel nature inspired computer for real–time applications. The hardware implementation of the model needs six orders of magnitude less time than the software representation. In this paper, in order to develop a proof of concept and depict the applicability of the proposed hardware oriented CA approach, the topology of Greece is used as an input of the biological computer. The network formed by the in vitro experiments, along with the one designed by the CA model and implemented in hardware are compared with the real motorways and the proximity graphs of the topology.

Journal Article Type Article
Publication Date Jun 1, 2016
Journal Natural Computing
Print ISSN 1567-7818
Electronic ISSN 1572-9796
Publisher Springer (part of Springer Nature)
Peer Reviewed Peer Reviewed
Volume 15
Issue 2
Pages 279-295
DOI https://doi.org/10.1007/s11047-014-9459-0
Keywords unconventional computing, slime mould, cellular automata, hardware, motorway networks, biological computer
Public URL https://uwe-repository.worktribe.com/output/907732
Publisher URL http://dx.doi.org/10.1007/s11047-014-9459-0