Information Technology Reference
In-Depth Information
when the parameter
a
is smaller, i.e., it is more necessary to use these two new strategies
in HIS when the self set has a bigger potential of growing. And from the
△
C
's point of
view, the progress of strategy II is more notable than that of strategy I.
In addition, strategy I needs an additional space to store every partial matching
length of every detector. And strategy II needs two bits for every bit in a detector
because '#' is adopted. For example, “00” means '0', “01” means '1', and “11” (or
“10”) means '#'. Therefore, the space costs of both strategy I and II are larger than that
of strategy 0. Furthermore, the implementation complexities of strategy I and strategy
II in hardware immune systems are also a little higher than that of strategy 0. However,
compared with the advantages of these two new strategies, these disadvantages are not
crucial factors in the implementation of many hardware immune systems.
6 Conclusions
Self-Tolerance is a key issue in the research of Hardware Immune Systems. Two novel
detector set updating strategies are proposed in this paper. Compared with previous
detector set updating strategies in [8] and [9], results of simulation experiments prove
that, no matter the self set is available or not, the detector sets being updated by these
two new strategies are less affected by the growing of the self set, and the new
strategies have a clearly lower increasing on the false negative ratio in a dynamic
environment.
There are also some future works that should be studied for improvement, such as
embedding these strategies into a real hardware immune system for special
applications.
Acknowledgements.
This work is supported by National Natural Science Foundation
of China (No.60404004) and Nature Science Major Foundation from Anhui Education
Bureau (No. 2004kj360zd).
References
1.
Bradley, D.W., Tyrrell, A.M.: Immunotronics - Novel Finite-State-Machine Architectures
with Built-In Self-Test Using Self-Nonself Differentiation. IEEE Transactions on
Evolutionary Computation, Vol. 6(3) (2002) 227-238
2.
Tyrrell, A.M.: Computer Know Thy Self!: A Biological Way to Look at Fault Tolerance.
Proceedings of 2nd EuroMicro / IEEE Workshop Dependable Computing Systems (1999)
129-135
3.
Canham, R., Tyrrell, A.M.: A Learning, Multi-Layered, Hardware Artificial Immune
System Implemented upon an Embryonic Array. Proceedings of 5th International
Conference on Evolvable Systems, ICES (2003) 174-185
4.
Canham, R., Jackson, A.H., Tyrrell, A.M.: Robot Error Detection Using an Artificial
Immune System. Proceedings of NASA/DoD Conference on Evolvable Hardware (2003)
199-207
5.
Bradley, D.W., Tyrrell, A.M.: The Architecture for a Hardware Immune System.
Proceedings of 3rd NASA/ DoD Workshop on Evolvable Hardware. Long Beach, Cailfornia
(2001) 193-200
