18-02-2013, 09:44 AM
BorderSense: Border patrol through advanced wireless sensor networks
BorderSense.pdf (Size: 468.42 KB / Downloads: 135)
Abstract
The conventional border patrol systems suffer from intensive human involvement.
Recently, unmanned border patrol systems employ high-tech devices, such as unmanned
aerial vehicles, unattended ground sensors, and surveillance towers equipped with camera
sensors. However, any single technique encounters inextricable problems, such as high
false alarm rate and line-of-sight-constraints. There lacks a coherent system that coordinates
various technologies to improve the system accuracy. In this paper, the concept of
BorderSense, a hybrid wireless sensor network architecture for border patrol systems, is
introduced. BorderSense utilizes the most advanced sensor network technologies, including
the wireless multimedia sensor networks and the wireless underground sensor networks.
The framework to deploy and operate BorderSense is developed. Based on the
framework, research challenges and open research issues are discussed.
Introduction
Border patrol systems have recently gained interest to
address the concerns about national security. The major
challenge in protecting long stretches of borders is the
need for intensive human involvement in patrolling the
premises. Conventional border patrol system consists of
security checkpoints and border troops [5,27]. The security
checkpoints are set up on the international roads where all
vehicle traffic is stopped to detect and apprehend illegal
aliens, drugs, and other illegal activity. Each border troop
watches and controls a specific section of the border. The
troops patrol the border according to predetermined route
and time interval.
Existing border patrol techniques
Border patrol has extensively been based on human
involvement. However, the relative cost for the increasing
number of personnel as well as the diminishing accuracy
through human-only surveillance has required the
involvement of high-tech devices in border patrol. Among
these, Unmanned Aerial Vehicles (UAVs) for aerial surveillance
have recently been used to automatically detect and
track illegal border crossing [14,13,24]. Due to the large
coverage and high mobility of the UAVs, the intensive human
involvement in low-level surveillance activities can
be reduced. This allows valuable human resources to be
allocated to decision management activities based on
information from these devices. However, similar to the
conventional border patrol systems, UAVs alone cannot
cover the whole border at any time. There may exist times
when certain sections of the border is not being monitored.
Moreover, the UAVs have significantly higher costs and
accident rates than those of manned aircrafts and require
large human footprint to control their activities. In addition,
inclement weather conditions can also impinge on
the surveillance capability of UAVs.
Deployment of BorderSense
According to the heterogeneous architecture shown in
Fig. 1, the development problem of BorderSense system
is discussed in this section. In border patrol applications,
the established monitoring network should cover a significantly
large monitoring area. However, the sensing radius
of a single sensor node is normally limited. Thus, a large
number of sensor nodes are expected to fulfill the coverage
requirement. Moreover, different types of sensor nodes
(e.g., underground, ground, camera, and mobile sensors)
provide different coverage capabilities. In addition, each
sensor type is characterized by different cost, sensing radius,
and sensing accuracy. Thus, an optimal deployment
strategy is required to determine the number and locations
of sensor nodes with heterogeneous capabilities. The primary
objective of the deployment research is to find the
deployment strategy using the minimum number of each
type of sensors to cover the whole surveillance area and
to achieve a desired intrusion detection probability.
Operational framework for BorderSense
Based on the network architecture and the deployment
strategy, the operation framework is described in this section
to realize the basic functionalities of BorderSense.
Since the hybrid WSN consists of three types of sensors,
three types of sensing information are obtained from a spatially
distributed set of sensors with different attributes.
The three types of sensing information are generally complementary
to each other. The multimedia sensors provide
still image or video information of the border area but the
intruder behind any obstacles cannot be detected. The
ground sensors can sense the ground vibration as well as
the magnetic anomaly caused by the intrusion. However,
it is difficult to distinguish a human intruder from a large
animal, hence high false alarm rate may be caused. The
underground sensors can also sense the vibration of the
ground, but the attribute of the sensing measurements
are different from those acquired by the ground sensors.
The false alarm rate of underground sensors is also high.
Hence, these heterogeneous set of information should be
fused at certain points in the network to improve the decision
accuracy and minimize the miss rate and false alarm
rate.