Skip to main content

Sensor Management

Our research explores the possibility treating the management of multiple interacting sensors as a graph problem that involves machine learning, active learning and constrained optimization. In this scenario, the interactions between sensors can be treated as a graph. The graph connections can then be grouped and their use (such as, for example, to optimize coverage) can be treated as a submodular problem. This approach is the set theoretic equivalent of convexity and assures that a single optimum solution is reached.