09-07-2012, 11:27 AM
Research Challenges and Applications for Underwater Sensor Networking
[attachment=27227]
Abstract
This paper explores applications and challenges for
underwater sensor networks. We highlight potential applications to
off-shore oilfields for seismic monitoring, equipment monitoring,
and underwater robotics. We identify research directions in shortrange
acoustic communications, MAC, time synchronization, and
localization protocols for high-latency acoustic networks, longduration
network sleeping, and application-level data scheduling.
We describe our preliminary design on short-range acoustic
communication hardware, and summarize results of high-latency
time synchronization.
INTRODUCTION
Sensor networks have the promise of revolutionizing many
areas of science, industry, and government. The ability to have
small devices physically distributed near the objects being
sensed brings new opportunities to observe and act on the
world, for example with micro-habitat monitoring [6], [26],
structural monitoring [47], and industrial applications [33].
While sensor-net systems are beginning to be fielded in applications
today on the ground, underwater operations remain
quite limited by comparison. Remotely controlled submersibles
are often employed, but as large, active and managed devices,
their deployment is inherently temporary. Some wide-area data
collection efforts have been undertaken, but at quite coarse
granularity (hundreds of sensors to cover the globe) [40]. Even
when regional approaches are considered, they are often wired
and very expensive [12].
SYSTEM ARCHITECTURE
Before describing specific applications, we briefly review
the general architecture we envision for an underwater sensor
network. Figure 1 shows a diagram of our current tentative
design. We anticipate a tiered deployment, where some nodes
have greater resources.
In Figure 1, we see four different types of nodes in the
system. At the lowest layer, the large number of sensor nodes
are deployed on the sea floor (shown as small yellow circles).
They collect data through attached sensors (e.g., seismic) and
communicate with other nodes through short-range acoustic
modems. They operate on batteries, and to operate for long
periods they spend most of their life asleep. Several deployment
strategies of these nodes are possible; here we show them
anchored to the sea floor. (They could also be buried for
protection.) Tethers ensure that nodes are positioned roughly
where expected and allow optimization of placement for good
sensor and communications coverage. Node movement is still
possible due to anchor drift or disturbance from external effects.
We expect nodes to be able to determine their locations through
distributed localization algorithms.
APPLICATIONS
We see our approaches as applicable to a number of applications,
including seismic monitoring, equipment monitoring and
leak detection, and support for swarms underwater robots. We
review their different characteristics below.
a) Seismic monitoring: A promising application for underwater
sensor networks is seismic monitoring for oil extraction
from underwater fields. Frequent seismic monitoring is of
importance in oil extraction. Studies of variation in the reservoir
over time are called “4-D seismic” and are useful for judging
field performance and motivating intervention.
PROTOCOLS FOR HIGH-LATENCY NETWORKS
Acoustic communication puts new constraints on networks
of underwater sensor nodes for several reasons. First, the large
propagation delay may break or significantly degrade the performance
of many current protocols. For example, propagation
delay for two nodes at 100m distance is about 67ms. Second,
the bandwidth of an acoustic channel is much lower than that
of a radio. Efficient bandwidth utilization becomes an important
issue. Finally, unlike terrestrial networks, underwater sensor
networks cannot take advantage of rich existing infrastructure
such as GPS. We next examine several research directions at
the network level.
CONCLUSIONS
This paper has summarized our ongoing research in underwater
sensor networks, including potential applications and
research challenges.
[attachment=27227]
Abstract
This paper explores applications and challenges for
underwater sensor networks. We highlight potential applications to
off-shore oilfields for seismic monitoring, equipment monitoring,
and underwater robotics. We identify research directions in shortrange
acoustic communications, MAC, time synchronization, and
localization protocols for high-latency acoustic networks, longduration
network sleeping, and application-level data scheduling.
We describe our preliminary design on short-range acoustic
communication hardware, and summarize results of high-latency
time synchronization.
INTRODUCTION
Sensor networks have the promise of revolutionizing many
areas of science, industry, and government. The ability to have
small devices physically distributed near the objects being
sensed brings new opportunities to observe and act on the
world, for example with micro-habitat monitoring [6], [26],
structural monitoring [47], and industrial applications [33].
While sensor-net systems are beginning to be fielded in applications
today on the ground, underwater operations remain
quite limited by comparison. Remotely controlled submersibles
are often employed, but as large, active and managed devices,
their deployment is inherently temporary. Some wide-area data
collection efforts have been undertaken, but at quite coarse
granularity (hundreds of sensors to cover the globe) [40]. Even
when regional approaches are considered, they are often wired
and very expensive [12].
SYSTEM ARCHITECTURE
Before describing specific applications, we briefly review
the general architecture we envision for an underwater sensor
network. Figure 1 shows a diagram of our current tentative
design. We anticipate a tiered deployment, where some nodes
have greater resources.
In Figure 1, we see four different types of nodes in the
system. At the lowest layer, the large number of sensor nodes
are deployed on the sea floor (shown as small yellow circles).
They collect data through attached sensors (e.g., seismic) and
communicate with other nodes through short-range acoustic
modems. They operate on batteries, and to operate for long
periods they spend most of their life asleep. Several deployment
strategies of these nodes are possible; here we show them
anchored to the sea floor. (They could also be buried for
protection.) Tethers ensure that nodes are positioned roughly
where expected and allow optimization of placement for good
sensor and communications coverage. Node movement is still
possible due to anchor drift or disturbance from external effects.
We expect nodes to be able to determine their locations through
distributed localization algorithms.
APPLICATIONS
We see our approaches as applicable to a number of applications,
including seismic monitoring, equipment monitoring and
leak detection, and support for swarms underwater robots. We
review their different characteristics below.
a) Seismic monitoring: A promising application for underwater
sensor networks is seismic monitoring for oil extraction
from underwater fields. Frequent seismic monitoring is of
importance in oil extraction. Studies of variation in the reservoir
over time are called “4-D seismic” and are useful for judging
field performance and motivating intervention.
PROTOCOLS FOR HIGH-LATENCY NETWORKS
Acoustic communication puts new constraints on networks
of underwater sensor nodes for several reasons. First, the large
propagation delay may break or significantly degrade the performance
of many current protocols. For example, propagation
delay for two nodes at 100m distance is about 67ms. Second,
the bandwidth of an acoustic channel is much lower than that
of a radio. Efficient bandwidth utilization becomes an important
issue. Finally, unlike terrestrial networks, underwater sensor
networks cannot take advantage of rich existing infrastructure
such as GPS. We next examine several research directions at
the network level.
CONCLUSIONS
This paper has summarized our ongoing research in underwater
sensor networks, including potential applications and
research challenges.