Distributed Sensor Network Solution
The emergence of miniature low-power sensor technologies that possess on-board signal processing and wireless communication capabilities has stimulated great interests in utilization of distributed sensor networks in a wide variety of applications. These sensor networks offer a new and promising paradigm for military, commercial, environmental, remote sensing, security and situation awareness applications. They can be used to capture a wide variety of signals including acoustic, seismic, IR, etc.
Benefits & Capabilities
Among the benefits of the distributed sensor networks are: simplicity and ease in deployment especially in urban areas, larger coverage area, and better spatial resolution for separating multiple closely spaced sources, less hardware complexity and hence significantly lower costs, and more flexibility in configuring different dynamic sensor arrays. However, there are many challenging technical issues that must be addressed.
Our Innovation
ISTI developed several technologies to overcome the challenges associated with the real-time tracking and localization of ground and low flying airborne vehicles as well as for transient events (e.g., gunshots) detection and localization using low-cost commercially off-the-shelf (COTS) wireless sensor nodes. More specifically, the innovations are in developments of,
- New wideband direction of arrival (DOA) estimation methods to successfully resolve and track multiple closely spaced moving sources. These methods are robust to sensor failure and sensor location errors and/or array mismatch effects.
- A robust triangulation method that uses the estimated DOAs to locate the moving vehicles.
- A new method to separate the acoustic signatures of multiple vehicles and subsequently classify the vehicles.
- Low-cost COTS-based sensor nodes with associated synchronization and communication utilities and a user-friendly GUI for pre- and post-deployment sensor network data analysis and processing.
Figure below shows the results of localizing a ground vehicle using 15 distributed acoustic sensor nodes deployed in three random clusters of 5 nodes each. Each cluster generates a DOA of the source, which are subsequently used to localize the moving vehicle in real-time
Figure below shows a MATLAB-based GUI developed to run different sensor configurations (distributed and regular) prior to deployment. The GUI allows the users to examine the effects of sensor location errors and acoustic transmission loss, which depends on atmospheric and ground conditions, sensor locations and source frequency and range to the sensors. Different DOA estimation methods may also be tested for their accuracy. In particular, the user may examine beampattern and bearing responses of different array configurations under various conditions. Furthermore, this powerful GUI also allows the users to capture and process the actual sensor data on-line after the deployment. Among the processing capabilities are: DOA estimation using a wide variety of narrow-band and wideband methods, and source localization, tracking, separation and classification methods. This GUI and the embedded tools and algorithms can be used in a wide variety of distributed sensing applications using different sensing modalities. Please contact mo@infsyst.biz for further information.
