Abstract
Immune system (IS) is capable of evolving, learning, recognising and eliminating foreign molecules which invade organisms. Fault tolerant approaches consist in detecting erroneous states of an algorithm and executing recovery procedures. They are generally implemented by adding formal properties to be satisfied by state variables during execution. However, defining fault-tolerant procedures can be error-prone and sometimes equivalent to a formal proof. Moreover, it is always very difficult to make a sound assumption about the typology and frequency of real. We propose an analogy between IS and fault tolerant computing. After a brief presentation of immune algorithms and especially lymphocyte simulation (B-cells, suppressor T-cells, etc.) using genetic operators, we present an immune model for detecting and recovering erroneous states during execution. Program states are examined by procedures playing the role of B-cells. Those properties have previously been learnt, automatically, during testing phase using mutation analysis techniques. When a program state has to be recovered, T-cells procedures are activated in order to recover erroneous states. In this way, each software algorithm may develop automatically its own immune system capable of evolving and stimulating specific responses to software failures. We finish by a brief discussion of the main possibilities of an immune approach to fault-tolerant computing.
Preview
Unable to display preview. Download preview PDF.
Bibliography
V. Agrawal V.: “Proc, Coll. on Fault Tolerance” Montreal, 1985.
H. Bersini: “Immune Network & Adaptive Control” Paris, Décembre 1992.
H. Bertoni & M. Dorigo: “Implicit Parallelism in Genetic Algorithms” Artificial Intelligence, 61, 2, 307–314.
R. A. DeMillo & al.:“Mutation Analysis as a Tool for Software Quality Assurance”, Proc. COMPSAC 80, (October 1980)
J. Doyne Farmer, N.H. Packard, A. S. Perelson: “The Immune System, Adaptation, and Machine Learning” in Evolution, Game and Learning, Proc. of 5th Annual Inter. conference, Los Alamos, (May 1985).
J. Doyne Farmer: “A Rosetta Stone for Connectionism” in Emergent Computation, Ed. S. Forrest, MIT, North-Holland (1990).
D.E. Goldberg: “Genetic Algorithms in Search, Optimization, and Machine Learning” Addison Wesley Publ. Company, (1989).
W. D. Hillis: “Co-evolving parasites improve simulated evolution as an optimization procedure” in Emergent Computation, S. Forrest, MIT, North-Holland (1990).
J.H. Holland: “Schemata and intrinsically parallel adaptation”, Proce. of the NSF Workshop of Learning System Theory and its applications (pp 43–46), Univ. of Florida, (1973).
N. K. Jerne: “Towards a network theory of the Immune System”, Ann. Immunol. (Inst. Pasteur) 125 C (1974).
R. A. Maxion: “Toward diagnosis as an emergent behavior in a network ecosystem” in Emergent Computation, Ed. S. Forrest, MIT, North-Holland (1990).
Z. Michalewicz: “Genetic Algorithms+Data Structures=Evolution Programs”, Springer Verlag, (1992).
“Machine Learning — An Artificial Intelligence Approach”, ed. Michalski R. S. & al, Tioga Publishing Company, Palo Alto, CA, 1983.
A. Mili: “Program Fault Tolerance. A structured programming approach”, Prentice Hall International (1990).
Author information
Authors and Affiliations
Editor information
Rights and permissions
Copyright information
© 1996 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Xanthakis, S., Karapoulios, S., Pajot, R., Rozz, A. (1996). Immune system and fault-tolerant computing. In: Alliot, JM., Lutton, E., Ronald, E., Schoenauer, M., Snyers, D. (eds) Artificial Evolution. AE 1995. Lecture Notes in Computer Science, vol 1063. Springer, Berlin, Heidelberg. https://doi.org/10.1007/3-540-61108-8_38
Download citation
DOI: https://doi.org/10.1007/3-540-61108-8_38
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-61108-0
Online ISBN: 978-3-540-49948-0
eBook Packages: Springer Book Archive