Expressing Sensor Network Interaction Patterns Using Data-Driven Macroprogramming

Abstract

Wireless sensor networks (WSNs) are increasingly being employed as a key building block of pervasive computing infrastructures, owing to their ability to be embedded within the real world. So far, pervasive applications for WSNs have been developed in an ad-hoc manner using node-centric programming models, focusing on the behavior of single nodes. Instead, macro-programming models provide much higher levels of abstractions, allowing developers to reason on the sensor network as a whole. In this paper, we demonstrate how a wide range of interaction patterns commonly found in pervasive, embedded applications can be expressed using ATaG, a data-driven macro-programming language. To support this, we showcase real-world applications developed in ATaG, and consider both homogeneous, sense-only scenarios, and heterogeneous settings involving actuation on the environment under control