Using logical neighborhoods to enable scoping in wireless sensor networks

• Authors: Luca Mottola;Gian Pietro Picco
• Venue: Proceedings of the 3rd international Middleware doctoral symposium
• Year: 2006

Wireless Sensor Networks (WSNs) are now enabling applications whose objective is not just to monitor the environment, but also to perform actions on it so as to implement complex control loops. Unlike early WSN projects where the application tasks were mainly relegated to the fringes of the network, e.g., to a powerful base station, in sensing and acting scenarios the application intelligence is brought in the network, and distributed among the nodes [1]. These applications are often composed of many collaborating sub-tasks, each involving only a subset of the nodes in the system. Therefore, the programmers must worry about how to identify these subsets and address them, before concentrating on the application goals. This results in additional programming effort and more complex code, affecting the reliability of the resulting application.In this work, we propose a programming abstraction called Logical Neighborhood, whose goal is to raise the level of abstraction from the physical neighborhood of a node to a logical notion of proximity. The programmers can specify the nodes part of a logical neighborhood using a declarative language we devised, based on application-defined attributes of the nodes. To address the members of a logical neighborhood, our framework provides a general communication API, supported by a dedicated routing scheme. Here, we present the logical neighborhood abstraction, illustrate our dedicated routing solution briefly reporting on some performance results, and point at current and future investigations based on the logical neighborhood abstraction.