Study on Hardware Implementation of Artificial Immune System

Negative selection algorithm is one of the most widely used techniques in the field of artificial immune systems. This paper proposed an approach to implement negative selection algorithm based on FPGA, aiming at fault detection problems. The negative selection algorithm generally uses binary matching rules to discriminate self from non-self. Firstly, three most widely used binary matching rules were presented. Then we gave the details of our implementation of binary matching rules technique on FPGA. Experiment result shows that this approach for negative selection algorithm is very effective. This paper also put forward the hardware immune system architecture with an immunologically inspired approach to fault tolerance. It demonstrates a finite state machine can be used to detect every faulty state during a normal operating cycle in FPGA.