Perimeter surveillance of forward operating locations, such as Forward Arming and Refueling Points (FARPs), is crucial to ensure the survivability of personnel and materiel. FARPs are frequently located well outside the protective cover of the main forward operating bases. Therefore, they must provide their own organic perimeter defenses. Such defenses are manpower intensive. Research shows how cheap, remote, unattended sensors using commercial off-the-shelf (COTS) components can help reduce the manpower requirement for this task and yet not compromise the security of the operating location.
A low-cost Wireless Ground Sensor Network (WGSN), consisting of nodes built using low-cost COTS components, can transmit early warning intrusion detection information to a C2 Application Server located within the FARPs tactical command post, providing early intrusion detection capability for remote, small operating bases.
This WGSN deployed around the perimeter of the remote, small operating bases will relay information to the centralized C2 Application Server through a WiFi network. The meshed WiFi network can be achieved by leveraging COTS meshed routers. Each router is connected to a few WGS nodes that will be responsible for providing intrusion detection data back to the C2 Application Server.
The overall WGSN is designed to report any intrusion around the operating base's perimeter and attempt to recognize and track the intruder movement through the operating base using Passive Infrared (PIR) and facial image recognition technology. The intrusion related information and images would be transmitted from the respective WGS nodes to the C2 Application Server and SQL database for processing, storage, and image analytics before presenting the consolidated situational picture and intrusion alert information as a dashboard on the C2 Application. Based on the presented information, the relevant base protection force can be activated to neutralize the intruder threats. Figure 1 provides the overall system design concept.
The scenario presented in Figure 2 illustrates how the low-cost WGSN can be used to provide early intrusion detection and recognition within the operating base perimeter.
To provide round-the-clock unmanned perimeter surveillance and early intrusion detection, the WGS nodes deployed at different detection layers around the operating base perimeter must be able to actively detect an intrusion, determine the type of the intruder and send alerts to the C2 Application Server to alert the operator. To achieve this objective, the WGS node is equipped with PIR sensors to provide the initial intrusion motion detection, which in turn triggers the camera onboard the WGS node to turn on for a limited period to determine if the intruder is a human or an animal by means of facial recognition. Upon confirming that it is a human being detected, the camera captures the facial images of the human intruder(s) and transmits them to the C2 Application Server for image analysis and identification.
The core processing hardware of the WGS node is a Raspberry Pi 3 Model B, which consists of a Quad Core 1.2GHz processor, 1 GB RAM, on-board wireless LAN (802.11n) and Low Energy Bluetooth, 4 x USB 2.0 ports, 40-pin extended General Purpose Input Output (GPIO) pins and Micro SD port for storing the operating system, application and data (Raspberry Pi, n.d.)
The operating system installed on the 16 GB MicroSD card within the Raspberry Pi 3 Model B is Raspberry Pi's official operating system, Raspbian Jessie (Version November 2016). The Raspberry Pi 3 core hardware, together with Raspbian Jessie operating system, supports a wide range of USB devices, such as USB webcams, and Raspberry Pi specific peripherals such as PIR sensors and Pi Camera, that can be connected to its USB ports, GPIO pins and Pi Camera interface, respectively.
This work was done by Dingyao Hoon and Yueng Hao Kenneth Foo for the Naval Postgraduate School. NPS-0005
This Brief includes a Technical Support Package (TSP).
Low-Cost Ground Sensor Network for Intrusion Detection
(reference NPS-0005) is currently available for download from the TSP library.
Don't have an account? Sign up here.