Wireless Sensor Networks for Battlefield Surveillance

TLDR

Real‑time tracking with WSN is extremely challenging, requiring high robustness, real‑time decision making, high‑frequency sampling, multimodal sensing, complex signal processing and data fusion, distributed coordination, and wide‑area coverage. This position paper investigates using WSN for ground surveillance and aims to develop an outdoor prototype that detects and classifies multiple targets—such as vehicles and troop movements—using inexpensive acoustic and magnetic sensors, while enabling automatic self‑organization and calibration through a hybrid sensor network architecture. The prototype incorporates automatic self‑organization, calibration, target detection and tracking, real‑time data transmission to a command centre, and employs particle‑filter data fusion to mitigate environmental and sensor noise.

Abstract

In this position paper, we investigate the use of wireless sensor network (WSN) technology for ground surveillance. The goal of our project is to develop a prototype of WSN for outdoor deployment. We aim to design a system, which can detect and classify multiple targets (e.g., vehicles and troop movements), using inexpensive off-the-shelf wireless sensor devices, capable of sensing acoustic and magnetic signals generated by different target objects. In order to archive our goals, we intend to design a system, which is capable of automatic selforganization and calibration. Such a system would need to be capable of performing detection and tracking of targets as well as sending the real time enemy mobility information to a command centre. Real-time tacking with WSN is extremely challenging since it requires high system robustness, real time decision making, high frequency sampling, multi-modality of sensing, complex signal processing and data fusion, distributed coordination and wide area coverage. We propose a Hybrid Sensor Network architecture (HSN), tailored specifically to meet these challenges. We investigate data fusion technologies such as particle filters, to handle both environmental and sensing noises of inexpensive sensors.