Optimizing detector distribution in V-detector negative selection using a constrained multiobjective immune algorithm

In this paper, a novel constrained multiobjective immune algorithm for optimizing detector distribution in V-detector negative selection is proposed. The theory of artificial immune system (AIS) and the spirit of population evolution are introduced to generate detectors. By combining the constraint handling technique and AIS-based multiobjective optimization, the algorithm is able to steadily maximize the anomaly coverage with little extra cost, which means the distribution with maximized coverage of the non-self space and minimized overlapping among detectors with fixed size will be well realized. Furthermore, the new approach is tested on some benchmark problems. The experimental results show that compared with some state-of-the-art methods, our algorithm can remarkably outperform them in terms of enhancing the detection rate by optimizing distribution without increasing the number of detectors.