Skip to Main Content
As both the cost and size of wireless sensor network (WSN) nodes decreases, the scale on which these networks can be deployed will increase dramatically. However, human-centric management of such large-scale networks will be both costly and slow. Without automated or self-management, WSNs will not be able to grow to achieve their potential size and function. The concept of self-managing systems is the primary focus of autonomic computing and applying this approach to WSNs may have the potential to curtail this inherent scalability issue. One of the primary concerns for a practical WSN deployment is the maximization of its operational lifetime, which can be achieved through judicious management of the networks limited power supplies. An established practice for delivering this behavior is the opportunistic hibernation of nodes, whereby redundant nodes are temporarily deactivated and thus consume no power. In this paper we analyze a technique for achieving this, based on interpolation. We investigate its ability to, not only perform initial hibernations, but also to autonomically hibernate and activate nodes in response to data perceived from a dynamic environment. We verify this autonomic behavior experimentally using a deployed wireless sensor network and also demonstrate some of the properties of the interpolation technique.