Bio-inspired prokaryotic array for fault-tolerant electronic systems

Created by W.Langdon from gp-bibliography.bib Revision:1.4192

  author =       "Mohammad Samie",
  title =        "Bio-inspired prokaryotic array for fault-tolerant
                 electronic systems",
  school =       "Faculty of Environment and Technology, University of
                 the West of England",
  year =         "2012",
  address =      "UK",
  keywords =     "genetic algorithms, genetic programming,
                 fault-tolerant, embryonic system, prokaryotic system,
                 unitronics, radiation tolerant, FPGA, SABRE,
                 self-healing system, bio-inspired system,
                 fault-tolerant electronic array",
  URL =          "",
  abstract =     "Integrated circuits and electronic systems fabricated
                 in current technologies suffer from a wide range of
                 faults that may occur during the fabrication process or
                 during the life time of each circuit/device. Soft
                 (non-persistent) faults are caused mainly by ionizing
                 radiation. It is known that the rate of soft faults
                 increases and the rate of hard faults decreases as
                 technology improves. It is therefore vital to build
                 systems able to tolerate faults, especially in order to
                 improve the reliability of critical applications, such
                 as systems working in extreme environments, satellites,
                 aircrafts, medical instruments etc. Self-healing and
                 fault-tolerant systems deal with faults through
                 mitigation, self-test and self-repair mechanisms. A
                 number of designs have already been proposed to improve
                 reliability. The bio-inspired approach to self-healing
                 systems is part of a very promising class of methods
                 that try to mimic the successful reliability solutions
                 found in living organisms to design self-healing
                 electronic systems. This project aims to further our
                 understanding of how the unicellular nature of
                 biological systems and their protective immune systems
                 could be used to enhance reliability of digital
                 electronic systems. A novel model, which takes
                 inspiration from unicellular living entities
                 (prokaryotes) to reduce the large hardware and software
                 overhead found in current bio-inspired multi-cellular
                 systems, is proposed and implemented in an architecture
                 named Unitronics (Unicellular + Electronics).
                 Unitronics proposes the application of mechanisms that
                 take place during bacterial biofilms formation to
                 improve the reliability of electronic systems. Concepts
                 characteristic to bacterial formation, such as geometry
                 of islands, void and cell, are used to define the
                 prototype of Unitronics-based applications in the
                 hardware layer. Each cell includes configuration
                 information that defines the cell's phenotype once it
                 has been programmed. Finally, other prokaryotic
                 features such as bacterial species, transposone,
                 horizontal and vertical gene transfer functions
                 (HGT-VGT) define the genotype of Unitronics systems,
                 and are used to evolve, test and repair cells within a
                 community. A single cell is not able to repair itself;
                 instead the cell includes just its own configuration
                 bits. A faulty cell is repaired based on its
                 similarities and differences to/from other cells within
                 a community of cells. This method has several
                 advantages, such as: balance between time and hardware
                 redundancies; possibility of repairing several faulty
                 cells at the same time; and compression of redundant
                 information needed for fault recovery. Several
                 applications (including an e-puck object avoidance
                 robot controller, signed and unsigned multipliers, and
                 PID controller) are used to demonstrate the underlying
                 theory and the practical viability of our bio-inspired
                 model, and to show examples of performances that can be
                 achieved using the Unitronics architecture we have
  notes =        "Is this GP?

                 March 2017 Full text not available from this

Genetic Programming entries for Mohammad Samie