Decision analysis: Environmental learning automata for sensor placement

Detection systems can be designed in a way that responds to the environment. We consider a decision analysis sensor placement problem where the probability of intrusion is driven by environmental factors. We use two types of sensors; those which detect targets, and those which detect the environment (current speeds). We use a learning automata technique to build a mechanism. Our proposed approach is dynamic, and can adapt to environmental changes. The technique is superior in the sense that reoptimization happens continuously, and can be done with distributed control. Our tests show that the achieved configurations are better than spacing sensors equally: detection rates are far higher.