Verification and Validation of Sensor Networks

Abstract

Sensor networks play an increasingly important role in critical systems infrastructure and
should be correct, reliable and robust.  In order to achieve these performance goals, it is
necessary to verify the correctness of system software and to validate the more broadly
defined world and system models.  This includes: 

  * Physical Phenomena (PDE models, statistical models, etc.),

  * Signals (Equations of state, physical properties, etc.),

  * Sensors (Physics models, noise models, etc.),

  * Hardware (Failure models, power consumption models, etc.),

  * RF (Antenna models, bandwidth, delay, propagation, etc.),

  * Embedded Code (Correctness, complexity, context),

  * Distributed Algorithms (Correctness, concurrency models, etc.),

  * Overall Sensor Network and Environment Models (Percolation
     theory, wave theory, information theory, simulation, etc.).

We outline some of the V & V issues involved in the various aspects of sensor networks
as well as possible approaches to their development and application both in simulation 
and in operational deployed systems.