Engineering Project Proposals

(Wireless sensor networks)

Group members
Hamdi Roumani (cs233017)
Douglas Stamp (cs233199)
Patrick Tayao (cs232039)
Tyson J Hamilton (cs231144)

Contact Information
Email: taal@yorku.ca
Phone: 416 318 8135

Report Summary
The following report contains four proposals, all of which rely on wireless sensor
network technologies.
Table of Content

Ocean Based Monitoring System.…. …………………………..…….…..…3

Motion Recording (node to node communication) …..……………….…….5

Soldier Monitoring System….….…………………………………………...7

Airport Baggage Delivery System….……………….………………………9

OCEAN BASED MONITORING SYSTEM
INTRODUCTION
Underwater monitoring of geological activity along plate faults could allow substantial
warning time for evacuation of target areas during the initial generation of a Tsunami
wave. Tsunamis are caused by rapid displacement of a body of water on a massive scale.
This displacement can be caused by numerous events such as earthquakes, volcanic
eruptions, underwater explosions and meteorite
impacts and the results can be unnoticeable or
devastating. There is no currently known prevention
for tsunamis however a quick, efficient warning
system can prove to be the best defense.

Monitoring an underwater area such as the ocean

using a wired network would be costly and time
consuming, however the use of wireless sensor
network could be strategically deployed. Each
device would require a particular set of sensors
based on the characteristics of a generating tsunami
wave. The formation of a tsunami wave can
generally be detected by the following:

• Vibrations indicating an earthquake.

• Large quantities of gas bubbling to the
surface.
• Unusually hot water temperatures.
• Presence of hydrogen sulfide.1

The sensors required to detect these characteristics

include a vibration sensor, temperature sensor and a Figure 1
chemical composition sensor.

The detected characteristics of the wave would be recorded and transmitted through the
ad hoc network to all nearby costal links that would alert the proper authorities. This
would allow an evacuation or prevention to begin immediately after the detection of the
tsunami potentially saving lives.

CURRENT TSUNAMI WARNING SYSTEMS

The systems used currently to detect tsunami waves rely heavily on seismic wave
detection. During geological underwater activity seismic waves are generated and
propagate through the Earth’s crust. This allows for extremely quick detection since
seismic waves generally travel with typical speeds of 4km/s (14,400km/h) where the
tsunami itself travels at approximately 500km/h to 1000km/h in open water and reduce
speeds dramatically as the water depth decreases.2 This allows for several hours of notice

1
Wikipedia, <http://en.wikipedia.org/wiki/Tsunami>
2
Wikipedia, <http://en.wikipedia.org/wiki/Tsunami_warning_system>
before arrival of the tsunami in which an alert may be generated, the entire process
requiring a total of 15 minutes.2

Since these warning systems rely on seismic waves, characteristics of the tsunami such as
energy transfer and propagation may not be detected. Seismic waves allow precise
determination of the epicenter however due to the nature of tsunami waves do not predict
the probability or progression of actual danger to those at shore.

WIRELESS SENSOR SOLUTION

The use of wireless sensor networks would allow a dynamic ad hoc network to be
installed within the large bodies of water to detect and monitor tsunami waves.
Monitoring the energy transfer of the tsunami would allow an accurate model to predict
the path of propagation giving a better idea of which coastlines to alert.

OBSTACLES / RESEARCH
The use of underwater communication is currently far less advanced than through other
mediums (e.g. Air). Water does not transmit radio waves as easily, making the range
limited. The use of acoustic waves is the current research interest and most immediately
feasible system. Acoustic waves present numerous difficulties for the designer, which
are not present in today’s wireless communication links using air as the medium of data
transfer. The signal propagation speed in underwater acoustic channel is about 1.5 x 103
m/sec; five orders of magnitude lower than the radio propagation speed (3 x 108 m/sec).
Most systems today operate below 30kHz and according to current research and
development, no commercial system can communicate faster than 40km/kbps as the
maximum attainable range/rate product.3 The bandwidth of underwater acoustic channels
operating over several kilometers are about several tens of kbps, while short-range
systems over tens of kilometers may reach speeds of hundreds of kbps. In addition to the
slow propagation delay, acoustic communication channels are also affected by factors
such as path loss, noise, multi-path, and Doppler spread. All of these factors cause high
bit-error and delay variance.

POSSIBLE SOLUTIONS
The research of underwater acoustic communication has been vigorously studied and any
improvements upon the currently manufactured systems would require further study of
transmitting data through that specific medium. The next most obvious improvement
solution would be to have a large array of wireless sensor nodes throughout the depths of
the body of water and have a buoy that would transmit the information across large
distances with current technology using air as the medium. This would allow the data to
be collected using the underwater nodes within close distances for the higher bit rate
provided the distance between nodes is limited to tens of kilometers. Another solution
involves finding the most efficient and cost friendly relation between distance of a wired
node close to the shoreline and communicating to the target area via other
communication means when the signal arrives to land.

3
D. B. Kilfoyle and A. B. Baggeroer. The State of the Art in Under-water Acoustic
Telemetry. IEEE Journal of Oceanic Engineering, OE-25(5):4–27, January 2000.
MOTION RECORDING VIA WSN (WIRELESS SENOR
NETWORKS)

(Please See “Important Note” Below)

Concept:
In certain environments motion detection / video surveillance can be difficult to
implement and manipulate due to logistical limitations. For example, say video
documented research is being done for a documentary and the subject of study is a fast
moving creature such as a hummingbird in its natural habitat. In order to track and
survey its behavior either:
i) A mini network of surveillance equipment will have to be set up in the
hummingbird’s environ and let recording continuously, or
ii) The network of surveillance would have to individually monitored /controlled
by a surveyor to determine what and when should the subject be
recorded/tracked

The first method is inefficient in terms of usage of recording media, power consumption
and time. The second is overly complex and expensive to implement.

By incorporating the use of a wireless Senor Network (WSN) within the surveillance
equipment the efficiency of the time and cost can be drastically improved, as the new
network would be virtually autonomous.

Implementation
When the subject (in this case the humming bird) enters a camera’s area of surveillance
the motion detection activates the recording. Direction and speed information is stored in
the resident mote of the camera and is sent to adjacent

camera motes. This information is used to determine which camera in the network is to
active and record next, thus creating a motion sensitive video tracking system, which
could be set up in remote location. Because the cameras are only active when recording
is beneficial, this efficiency allows for battery-powered surveillance. The power will be
conserved as there is no continuous recording, given the subject is mobile. Video
information can be collected using less personnel and at lower costs.

Important Note:
This is a complex system and the particular application above may not be practical. We
would like to stress, however, we find the idea introduced very interesting. In general, it
illustrates a system where the nodes are instructed to carry out certain actions by other
nodes in the system rather then by some self triggered event.
SOLDIER MONITORING SYSTEM USING WIRELESS SENSOR
NETWORKS
Background

Soldier management is one of the key components in military applications. Strategic

deployment is vital and contains many variables. One variable that is key to deployment
is soldier status. Determination of health and location can provide enough information to
determine the number of soldiers to deploy or where a medic might be required.

Objective: To implement a Soldier Monitoring System using wireless sensor networks.

Wireless Sensor Networks (WSN) are networks of low power communication nodes
communicating to each other in a peer to peer networking mode transferring data
acquired from different sensors back to a base station. The communication of each node
is through a radio transceiver that allows each node to communicate to each other in a
mesh networking topology. Attached to each node are inputs that monitors different
variables. The variables that will be monitored is temperature, pulse and gps information.
These input is relayed to neighboring nodes and passed on to a base station. Once
acquired, the base station processes the information received. To ensure mobility, the
base station will be designed in a mobile robot that will determine best location to ensure
connectivity within the mesh networking.

Design
The system has three facets of design: Sensor design/interface, Network implementation
and Base station design.

Sensor Design and/or Interface

This facet will entail the design of the sensors that will be used to interface with the
wireless sensor nodes. Designing the different sensors can be time consuming, but can
eliminate a lot of the cost entailed in the project. Once the sensors are designed, each
sensor must be able to interface with each node. This entails decoding the signals
acquired from the nodes to information that can be transmitted within the sensor network.
Therefore, each node must be programmed with necessary information to translate
signals transmitted from the different sensors.

Network Implementation
This is one of the most difficult facet to implement. Since communication will be peer to
peer, an optimal way of communicating within nodes must be developed to ensure
maximum bandwidth. Different algorithms can be tested to ensure that data is efficiently
acquired and transferred within the mesh of networks. This includes ensuring minimum
hops taken to communicate from node to base station. A routing algorithm must be
developed to ensure if one of the nodes is disconnected, there are alternative routes that
can be taken that is optimal.
Base Station Design
To ensure mobility, the base station will be installed in mobile robot. There are two
components that must be implemented in the design of the base station: data processing
and localization. Data processing will entail updating the necessary user interface from
the acquired data from the sensor nodes. To ensure that the mobile robot is within
communicating distance of the mote, after every data acquisition, the mobile robot must
calculate its optimal location and move to that location. This will entail determining
specific patterns of movements on the sensor nodes and applying a probabilistic function
to determine its next location. If communication is lost to all nodes, an algorithm must be
determined to move the mobile robot to the highest probable location that will allow
connectivity.
AIRPORT BAGGAGE MONITORING SYSTEM

System Overview

We propose a system which will be used to replace the existing baggage systems
implemented in airports. The downside of the current system lies with the lack of
redundancy and the difficulty in locating baggage if it is misplaced or shipped to the
wrong destination. For customers of different airlines, one of the most common
complaints is late or lost baggage. Our system should:

1) Never have a bag misplaced or lost permanently

2) If a bag is misplaced guarantee that its location is known and the time it will take
to retrieve the bag is made clear to the client

To implement such a system we intend on placing wireless nodes along the path of the
baggage system. These nodes communicate with other nodes along the path using a peer
to peer protocol. The nodes themselves each have a RFID tag reader.

System Details
To ensure the general location of the bag is known at any time we must ensure that we
can record when a particular bag passes through the RFID tag reader of a node. Once the
tag is read, the node sends a message back to the base station (Relying on a peer-to-peer
protocol) informing it to update a database. This database will contain entries for every
bag currently in the system. They will be added to the database once the RFID tag is
placed on the bag. Each entry will contain the unique ID code of the bag as well as its
current location. In other words, once a bag passes through a node it will update the
database by changing the current position to that of the node the bag just passed through.

With this system we will know a bag’s general location at all times. Many features can
be implemented on top of this such as (It is very extendible and only a few possible ideas
are illustrated below):

1) A timer which monitors the duration between the location updates of the bag, if
too much time passes we know the bag could have fallen off.

2) An entry of all the valid locations the bag may pass through. If a location other
then these is sent back some sort of alarm can be triggered.

More information about the bag such as its weight and destination can be included. Once
a bag arrives at an airport it will need to be added to the new airport’s baggage system
(The bag can be taken off the system once the bag is collected by the passenger, this can
be done by having the passenger pass through a gateway with a node which signals the
system to remove the entry), it can check whether the destination matches the current
airport. The weight feature can be used for security to guarantee no extra content was
added during the time when the bag was in the previous airports baggage system as well
as the airplane it self.

They key concept of the system is a centralized database which contains entries for all
bags currently in the system. This database is updated via the nodes along the path the
bags travel. Wireless sensor networks allow this system to be very versatile and easy to
set up.