Virtual Sensing Range Emiliano Miluzzo, Nicholas D. Lane, and Andrew T. Campbell Computer Science Dept., Dartmouth College

1
Virtual Sensing Range Emiliano Miluzzo, Nicholas
D. Lane, and Andrew T. Campbell Computer Science
Dept., Dartmouth College
Performance Evaluation
Solution
Motivation
We have implemented VSR in TinyOS and evaluated
preliminary results using Tmotes Invent with a
2-axis accelerometer. We ran a small experiment
carrying 5 sensors along a 40 m long hallway
where the caller is placed in the middle of the
hallway. The callers transmission power is set
such that the callers transmission range is
about 7 m. In our set-up the light sensing range
is about 2 m. In the plot below the x-axis
represents the experiment duration while the
y-axis reports the distance from the caller with
reference point at 0 m.
We propose Virtual Sensing Range (VSR) with the
goal of extending the sensing range of a node
beyond the physical capabilities offered by its
modalities. VSR exploits sensor nodes mobility
patterns relying on the accelerometer and compass
mounted on the mobile sensors. We have
implemented VSR in TinyOS and on Tmote Invents
for our preliminary evaluation.
Mobile sensor networks have been gaining
considerable interest in the research
community. Due to physical impediments, cost and
scalability issues it might not be possible to
instrument an entire area with static
sensors. Because of the limited sensors sensing
range some area of the field might not be covered
by sensors. In this case the sensor network
might experience lack of sensing coverage.
With VSR it is possible to retrieve sensed data
for almost every distance within 15 m of the
caller on both sides of the hallway.
VSR Basic Idea
VSR More Details
To sense beyond Sr the caller tasks a callee
according to i) compass readings
ii) signals Link Quality Indicator (LQI) 1-
The caller picks as callee/s only mobile nodes
heading toward the AoI by means of compass
readings provided by the mobile nodes. 2- The
caller tasks a callee as soon as callees LQI
starts decreasing which is an indication that the
callee is approaching the callers radio coverage
edge. 3- A callee starts collecting sensed data
until the distance covered, measured with the
accelerometer, matches the distance specified by
the callers task. 4- Outside the callers radio
range, callees rely on mobile nodes heading in
the opposite direction to mule data back. A
callee X tasks a mule, which becomes a callee Y,
to collect sensed data for the distance covered
by callee X. 5- Once under the callers radio
range, callee Y delivers its own sensed data and
sensed data belonging to callee X.
The plot below reports the average light readings
(in Lux) as a function of the distance from the
caller. The caller is located at the brightest
point (higher Lux) along the hallway.
Without VSR we would be only able to collect
light readings from the caller within its
sensing capability range, i.e., -2 to 2 m. With
VSR we can sense far beyond the -2 to 2 m range.
A sensor, i.e., the caller, wants to sense beyond
its sensing range Sr. Tr is the transmission
range. The caller tasks one or more callees
heading toward the Area of Interest
(AoI). Data collected in the AoI is muled back to
the caller.
With support from the Institute for Security
Technology Studies (ISTS) and Intel Corp. More
information on the MetroSense Project, including
publications, technical reports, and source code
from http//metrosense.cs.dartmouth.edu/.