Programming paradigms for networked sensing: a distributed systems’ perspective

  • Authors:
  • Amol Bakshi;Viktor K. Prasanna

  • Affiliations:
  • Department of Electrical Engineering, University of Southern California, Los Angeles, CA;Department of Electrical Engineering, University of Southern California, Los Angeles, CA

  • Venue:
  • IWDC'05 Proceedings of the 7th international conference on Distributed Computing
  • Year:
  • 2005

Quantified Score

Hi-index 0.00

Visualization

Abstract

Research in embedded networked sensing has primarily focused on the design of hardware architectures for sensor nodes and infrastructure protocols for long lived operation of resource constrained sensor network deployments. There is now an increasing interest in the programming aspects of sensor networks, especially in the broader context of pervasive computing. This paper provides a brief overview of ongoing research in programming of sensor networks and classifies it into layers of abstraction that provide the application developer with progressively higher level primitives to express distributed, phenomenon-centric collaborative computation. As a specific instance of a macroprogramming methodology, we discuss the data driven Abstract Task Graph (ATaG) model and the structure of its underlying runtime system. ATaG separates the application functionality from non-functional aspects, thereby enabling end-to-end architecture-independent programming and automatic software synthesis for a class of networked sensor systems. A prototype visual programming, software synthesis, functional simulation and visualization environment for ATaG has been implemented.