VigilNet:

An Integrated Sensor Network System
for Energy-Efficient Surveillance

VigilNet Design Goals

Develop an operational self-organized wireless network to provide tripwire-based surveillance with a sentry-based power management scheme, in order to achieve minimum 3 ~6 months life time with current hardware capability. The system should also support timely detection, tracking and coarse granular classification of vehicle and personnel targets over all kinds of terrain.

The general objective of VigilNet system (pervious known as SOWN) is to alert the military command and control unit in advance to the occurrence of events of interest in hostile regions. The event of interest for our work is the presence of moving vehicles in the deployed region. The deployed sensor devices must have the ability to detect and track vehicles in the region of interest. Successful detection and tracking requires that the application obtain the current position of a vehicle with acceptable precision and confidence. When the information is obtained, it has to be reported to a remote base station within an acceptable latency.

VigilNet currently consists about 40,000 lines of NesC and Java code, running on XSM, Mica2 and Mica2dot platforms. The complete system is designed to scale to at least 1000 XSM motes and cover minimal 100,1000 square meters to ensure operational applicability. Download is available through Sourceforge

VigilNet Software Architecture

The above graph provides an overview of the VigilNet architecture. The VigilNet architecture is built on top of TinyOS. TinyOS is an event driven computation model, written in NesC specifically for the motes platform. TinyOS provides a set of essential components such as hardware drivers, scheduler and basic communication protocols. These components provide low-level support for VigilNet modules, which are also written in NesC.

The application components are specially designed for the surveillance purpose.

EnviroTrack, which provides entity-based tracking service,
Classification components, which provides four types of target differentiation ,
Velocity regression , which provides target speed and bearing estimation,
False alarm Filtering, which differentiates the real and false targets.

The Middleware components are designed to be application independent.

Time synchronization module is responsible for synchronizing the local clocks of the motes with the clock of the base station.
Localization module is responsible for ensuring that each mote is aware of its location.
Configuration module is responsible for dynamical reconfiguring the system when system requirement changes.
Asymmetric Detection is designed to aid routing module to select high-quality links to the base.
Radio Wakeup module is used to alert non-sentry motes when a significant event happens.
The Sentry Service component and Tripwire Management are responsible for power management.
The Group Management component is responsible for collaborative detection and tracking of events.

Time synchronization, localization, and routing comprise the lower-level components and form the basis for implementing the higher-level middleware services, such as aggregation and power management.

The sentry and tripwire service conserves energy of the sensor network by selecting a subset of motes, which we define as sentries, to monitor events. The remaining motes are allowed to remain in a low-power state until an event occurs. When an event occurs, the sentries awaken the other motes in the region and the group management component dynamically organizes the motes into groups in order to enable collaborative tracking. Together, these two components are responsible for energy-efficient event tracking.

This project is sponsored by DARPA (Defense Advance Research Projects Agency). For more information, please visit the website http://www.cs.virginia.edu/~control