A learning, multi-layered, hardware artificial immune system implemented upon an embryonic array

R Canham; A M Tyrrell

A learning, multi-layered, hardware artificial immune system implemented upon an embryonic array

Electronic Engineering

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract

A simple robot control system is used to demonstrate bio-inspired fault tolerance techniques; a multi-layered hardware immune system is used for fault detection and an embryonic array for fault avoidance. The acquired layer of the immune system monitors system behaviour for unusual activity (normal behaviour is learnt during an unsupervised teaching period). An non-learning innate layer. is then employed to localise the fault if possible. The embryonic array allows simple, robust reconfiguration to avoid the fault; as many faults as spare cells can be accommodated. The complete system, including the learning algorithm, is implemented on a Virtex FPGA. Results showing the appropriate response to different types of faults axe given.

Original language	English
Title of host publication	EVOLVABLE SYSTEMS: FROM BIOLOGY TO HARDWARE, PROCEEDINGS
Editors	AM Tyrrell, PC Haddow, J Torresen
Place of Publication	BERLIN
Publisher	Springer
Pages	174-185
Number of pages	12
ISBN (Print)	3-540-00730-X
Publication status	Published - 2003
Event	5th International Conference on Evolvable Systems - TRONDHEIM Duration: 17 Mar 2003 → 20 Mar 2003

Conference

Conference	5th International Conference on Evolvable Systems
City	TRONDHEIM
Period	17/03/03 → 20/03/03

Poetic
Tyrrell, A. (Principal investigator), Howard, D. M. (Other) & Halliday, D. M. (Other)
EUROPEAN COMMISSION
1/09/01 → 28/02/05
Project: Research project (funded) › Research

Cite this

@inproceedings{cdb280ba8e4d422cbfdb53ee9b8f07c3,

title = "A learning, multi-layered, hardware artificial immune system implemented upon an embryonic array",

abstract = "A simple robot control system is used to demonstrate bio-inspired fault tolerance techniques; a multi-layered hardware immune system is used for fault detection and an embryonic array for fault avoidance. The acquired layer of the immune system monitors system behaviour for unusual activity (normal behaviour is learnt during an unsupervised teaching period). An non-learning innate layer. is then employed to localise the fault if possible. The embryonic array allows simple, robust reconfiguration to avoid the fault; as many faults as spare cells can be accommodated. The complete system, including the learning algorithm, is implemented on a Virtex FPGA. Results showing the appropriate response to different types of faults axe given.",

author = "R Canham and Tyrrell, {A M}",

year = "2003",

language = "English",

isbn = "3-540-00730-X",

pages = "174--185",

editor = "AM Tyrrell and PC Haddow and J Torresen",

booktitle = "EVOLVABLE SYSTEMS: FROM BIOLOGY TO HARDWARE, PROCEEDINGS",

publisher = "Springer",

address = "Germany",

note = "5th International Conference on Evolvable Systems ; Conference date: 17-03-2003 Through 20-03-2003",

}

TY - GEN

T1 - A learning, multi-layered, hardware artificial immune system implemented upon an embryonic array

AU - Canham, R

AU - Tyrrell, A M

PY - 2003

Y1 - 2003

N2 - A simple robot control system is used to demonstrate bio-inspired fault tolerance techniques; a multi-layered hardware immune system is used for fault detection and an embryonic array for fault avoidance. The acquired layer of the immune system monitors system behaviour for unusual activity (normal behaviour is learnt during an unsupervised teaching period). An non-learning innate layer. is then employed to localise the fault if possible. The embryonic array allows simple, robust reconfiguration to avoid the fault; as many faults as spare cells can be accommodated. The complete system, including the learning algorithm, is implemented on a Virtex FPGA. Results showing the appropriate response to different types of faults axe given.

AB - A simple robot control system is used to demonstrate bio-inspired fault tolerance techniques; a multi-layered hardware immune system is used for fault detection and an embryonic array for fault avoidance. The acquired layer of the immune system monitors system behaviour for unusual activity (normal behaviour is learnt during an unsupervised teaching period). An non-learning innate layer. is then employed to localise the fault if possible. The embryonic array allows simple, robust reconfiguration to avoid the fault; as many faults as spare cells can be accommodated. The complete system, including the learning algorithm, is implemented on a Virtex FPGA. Results showing the appropriate response to different types of faults axe given.

M3 - Conference contribution

SN - 3-540-00730-X

SP - 174

EP - 185

BT - EVOLVABLE SYSTEMS: FROM BIOLOGY TO HARDWARE, PROCEEDINGS

A2 - Tyrrell, AM

A2 - Haddow, PC

A2 - Torresen, J

PB - Springer

CY - BERLIN

T2 - 5th International Conference on Evolvable Systems

Y2 - 17 March 2003 through 20 March 2003

ER -

A learning, multi-layered, hardware artificial immune system implemented upon an embryonic array

Abstract

Conference

Projects

Poetic

Cite this