A method for robust adaptation of the configuration of distributed sensor fields

摘要

We illustrate an approach for adapting the configuration of surveillance-based distributed sensor fields. By using a robust method of centralized adaptation of the field configuration, the performance of distributed sensor fields can be greatly extended. This extension can be in terms of improved performance due to new local knowledge of the environment, improvement to compensate for individual sensor node failures, and improvement due to changes in the operational goal. In all of these situations, it is crucial to adapt only in a manner that is consistent with the performance intent of the original deployment strategy. In this paper, we use the concept of Pareto optimality (that is, the achievement of desired tradeoffs between competing objectives) as a mechanism for describing the intent of the original deployment. We then illustrate a computational approach to adaptation of the field configuration that can be performed to maintain this intent in the presence of a variety of sources of field degradation. Numerical examples are shown that illustrate this effect using a sparse sensor network that is designed for providing surveillance coverage of nonhomogeneous environments. We conclude with an analysis of the stability properties of this adaptation procedure, along with guidelines for its use in practical systems.