Internet of Things – Thesis Proposals

croutonsgruesomeRéseaux et Communications

16 févr. 2014 (il y a 3 années et 6 mois)

77 vue(s)

Internet of Things


Thesis Proposals

Proposal 1:
Video Streaming through
Wireless Sensor Networks

(June
2011)

Abstract

Wireless

Visual Sensor Networks (WVSN)

are

networks of wireless

devices capable of
sensing multimedia content such as audio, still

image
s and video, as well as scalar sensor
data from the environment.

Such multimedia networks will enable several new
applications,

especially in the surveillance and monitoring area
. Such

networks are in
many

ways unique and more challenging with respect to
traditional sensor

networks: (i)
sensor devices are constrained in terms of battery,

memory and
computational capability
while multimedia processing

requires to store, process and transmit large amount of data;
(ii)

applications of multimedia sensor networ
ks require real
-
time data

from camera
networks: on the one hand these requirements struggle

with the network side limitations
(variable channel capacity, limited

bandwidth, etc...) and on the other hand with the
constraints imposed

by the limited energy re
sources and processing speed of available

embedded processors.

In this thesis the candidate will investigate the possibility to build an efficient MPEG
-
like
video streaming system, starting from a hybrid DCT
-
DPCM existing system.

The goal of this thesis i
s the study and design of comprehensive solutions to support
video delivery in WMSNs. In particular, the work will address:



The design of effective video encoding techniques based on MPEG
-
like standards



The design of communication protocols to support vide
o delivery over multi
-
hop
WMSNs

The final outcome of the thesis will have to be an experimental testbed featuring the
developed encoding techniques and communication protocols


Tools

The methodologies used during the thesis will include:

-
Simulation:
Matla
b
(encoding techniques), TinyOS
-
nesC (testbed simulation)

-
Testbed Experimental study: TintOS
-
nesC (testbed development), Java (development of
a minimal application for video delivery in WMSNs).


Background

Foundations on Wireless Networking

Foundations of

programming languages (C/C++/Java)

Matlab Fundamentals (not mandatory)

Curiosity, Initiative


Notes

and References

S
.

Paniga, L
.

Borsani, A
.

Redondi, M
.

Tagliasacchi, M
.

Cesana,
Experimental Evaluation
of a Video Streaming System for

Wireless Multimedia S
ensor Networks
,
MedHocNet
2011


Supervisors
/Contacts

Matteo Cesana, cesana@elet.polimi.it

Alessandro Redondi, Redondi@elet.polimi.it

Proposal 2:
Image Transmission through Wireless Sensor Networks

(June
2011)


Abstract

Wireless

Visual Sensor Networks (WVSN
)

are

networks of wireless

devices capable of
sensing multimedia content such as audio, still

images and video, as well as scalar sensor
data from the environment.

Such multimedia networks will enable several new
applications,

especially in the surveillanc
e and monitoring area
. Such

networks are in
many

ways unique and more challenging with respect to traditional sensor

networks: (i)
sensor devices are constrained in terms of battery,

memory and
computational capability
while multimedia processing

requires

to store, process and transmit large amount of data;
(ii)

applications of multimedia sensor networks require real
-
time data

from camera
networks: on the one hand these requirements struggle

with the network side limitations
(variable channel capacity, lim
ited

bandwidth, etc...) and on the other hand with the
constraints imposed

by the limited energy resources and processing speed of available

embedded processors.

In this thesis, d
ifferent image coding and transmission schemes
for

WSN
s

will be
analyzed, imp
lemented and evaluated.

In particular, the work will include:



The design of effective image compression techniques to be implemented on sensor
nodes (JPG, progressive JPG, etc.)



The design of effective communication protocols to deliver the compressed ima
ges
remotely through multi
-
hop networks


Tools

The methodologies used during the thesis will include:

-
Simulation:
Matlab
(encoding techniques), TinyOS
-
nesC (testbed simulation)

-
Testbed Experimental study: TintOS
-
nesC (testbed development), Java (developm
ent of
a minimal application for image delivery in WMSNs).


Background

Foundations on Wireless Networking

Foundations of programming languages (C/C++/Java)

Matlab Fundamentals (not mandatory)

Curiosity, Initiative


Notes

and References

J.

Romberg,
Imaging
via Compressive Sampling


Supervisors
/Contacts

Matteo Cesana, cesana@elet.polimi.it

Alessandro Redondi, Redondi@elet.polimi.it







Proposal 3:
Audio Assisted Living

(June 2011)

Abstract

One of the greatest challenges in Ambient
Assisted

Living

is to des
ign smart homes that
could be able to anticipate the needs of its inhabitant while maintaining their comfort and
their safety with an adaptation of the house environment and a facilitation of the
connections to the outside world.
Sensor nodes can be geared

of multi
-
media sensors
including cameras, microphones and speakers.
In the field of monitoring applications for
WSNs, such multimedia “hardware” can be leveraged to implement safety
-
oriented
application and integrated surveillance systems. As an example,
the wireless sensor
networks can be used to convey vocal commands or to pervasively recognize words
and/or sounds which may trigger some type of reaction of the integrated surveillance
system (human intervention, on demand activation of video cameras, etc.
). All these
applications build upon the capability of the sensor and the sensor networks to recognize
acoustic patterns.


In this thesis a complete audio based assisted living system will be implemented and
deployed, through the design of suitable on/off
-
mote data compression schemes and
audio analysis algorithms.

The work will include:



The study of energy efficient algorithms for recognizing audio patterns through
commercial sensor hardware.



The study of the most proper means to deliver the audio pattern
s through multi
-
hop
wireless sensor networks.

Tools

The methodologies used during the thesis will include:

-
Simulation:
Matlab
(to test audio pattern recognition techniques), TinyOS
-
nesC (to
implement/simulate the audio pattern recognition techniques on
sensor nodes)

-
Testbed Experimental study: Tin
y
OS
-
nesC (testbed development), Java (development of
a minimal application for audio recognition in WMSNs).


Background

Foundations on Wireless Networking

Foundations of programming languages (C/C++/Java)

Matla
b Fundamentals (not mandatory)

Curiosity, Initiative


Notes

and References

Vacher, M. ; Portet, F. ; Fleury, A. ; Noury, N. ;
Challenges in the Processing of Audio
Channels for Ambient Assisted Living

LIG Lab., UJF, Grenoble, France


Berisha, V. ; Homin Kw
on ; Spanias, A. ;
Real
-
time acoustic monitoring using wireless
sensor motes

Dept. of Electr. Eng., Arizona State Univ., Tempe, AZ


Supervisors
/Contacts

Matteo Cesana, cesana@elet.polimi.it

Alessandro Redondi, Redondi@elet.
polimi.it

Proposal 4:
Street Qu
ality Detection for bikers

(June 2011)

Abstract

Have you ever cursed your local administration while riding your bike through bumpy
streets full of potholes? If so, you may be interested to work on this thesis project whose
aim is to develop a pothole dete
ction systems for bikers
based on
wireless sensor
networks.
The biker (or the bike) can be geared with wireless sensors with accelerometers
and gyroscopes which can be leveraged to detect discontinuities and holes on the street
surface. The thesis will inc
lude the following contributions:



the study and design of effective algorithms to process the accelerometers signals to
detect and assess the “quality” of the street surface.



the study of transmission mechanism to deliver the sampled (and processed) signa
l to
a cell phone



the implementation of a lightweight application to
display the results of the analysis
(the application might leverage positioning information)



Tools

The methodologies used during the thesis will include:

-
Simulation:
Matlab
(to test h
oles detection algorithms), TinyOS
-
nesC (to
implement/simulate the detection techniques on sensor nodes)

-
Testbed Experimental study: TintOS
-
nesC (testbed development), Java/Android
(development of a minimal application)


Background

Foundations on Wireles
s Networking

Signal Processing Fundamentals

Fundamentals of programming languages (C/C++/Java)

Matlab Fundamentals (not mandatory)

Curiosity, Initiative


Notes

and References

J
.

Eriksson, L
.

Girod, B
.

Hull, R
.

Newton, S
.

Madden, and H
.

Balakrishnan,

The po
thole
patrol: using a mobile sensor network for road surface monitoring
,

ACM
MobiSys '08.


Advisor/Contacts

Matteo Cesana, cesana@elet.polimi.it

Alessandro Redondi, Redondi@elet.polimi.it










Proposal 5:
3D Localization

Solutions

(June 2011)

Abstract

Localization

has been an important topic in data
-
centric wireless sensor networks (WSNs)
due to the special association between location information and the relevance of sensory
data. Although most proposed approaches have been modeled for 2D space, a gra
dual but
marked shift in the focus of
3D

localization

has taken place. In this thesis the candidate
will implement and evaluate a 3D localization algorithm on a Wireless Sensor Network,
including the development of a Graphic User Interface
.


Tools

TinyOS,
Matlab, Java


Background

Foundations on Wireless Networking

Foundations of programming languages (C/C++/Java)

Matlab Fundamentals (not mandatory)

Curiosity, Initiative


Notes

and References

Chia
-
Yen Shih
,

P.J.
Marró
,

COLA:
Complexity
-
Reduced Trilateration

Approach for 3D
Localization in Wireless Sensor Networks

Dept. of Comput. & Cognitive Sci., Univ.
Duisburg
-
Essen & Fraunhofer IAIS, Duisburg, Germany


A. Redondi,; M. Tagliasacchi, M. Cesana, L. Borsani, P. Tarrío, F. Salice,
LAURA


LocAlization and Ub
iquitous monitoRing of pAtients for health care support
,
IEEE
APLEC 2010


Advisor/Contacts

Matteo Cesana, cesana@elet.polimi.it

Alessandro Redondi, Redondi@elet.polimi.it
















Proposal

6
: Virtual Trainer for Running

(June 2011)

Abstract

Recent
technological advances in integrated circuits, wireless

communications, and
physiological sensing have boosted

the development of miniaturized, lightweight, ultra
-
low power,

intelligent monitoring devices. A number of these devices

can be integrated to for
m a Wireless Body Area Network

(WBAN), a wireless sensor
network whose devices, called

motes, are positioned on the body of a person to monitor

biometric parameters and vital signs. To this extent, the latest

advances in Mirco
-
Electro
Mechanical Systems (M
EMS) technology

have opened up the opportunity to realize
miniaturized

accelerometers, a sensing element that measures acceleration.

Such system can help the user during the

exercise providing him a feedback to avoid
errors which can

cause hurts.

The goal

of this thesis is the realization of an automatic system for the supervision of
jogging/running exercises. The systems will be composed of wearable sensor nodes
positioned on body to capture and process acceleration samples gathered through
accelerometers
, and to transmit remotely the collected information. In a nutshell, the
thesis objective can be summarized in the following points:



To study effective algorithms to acquire data from accelerometers



To study effective algorithms to process the aforemention
ed data to give a “quality”
indicator to the jogging/running gesture



To study effective solutions to collect and deliver the information remotely


Tools

The methodologies used during the thesis will include:

-
Simulation:
Matlab
(to test
acceleration analys
is algorithms
), TinyOS
-
nesC (to
implement/simulate the detection techniques on sensor nodes)

-
Testbed Experimental study: TintOS
-
nesC (testbed development), Java/Android
(development of a minimal application)


Background

Foundations on Wireless Networking

Signal Processing Fundamentals

Fundamentals of programming languages (C/C++/Java)

Matlab Fundamentals (not mandatory)

Curiosity, Initiative


Notes

and References

Stefano Melzi, Luca Borsani, Matteo Cesana,
The Virtual Trainer: Supervising
Movements Throug
h a Wearable Wireless Sensor Network
, IEEE Secon 2009



Advisor/Contacts

Matteo Cesana, cesana@elet.polimi.it

Alessandro Redondi, Redondi@elet.polimi.it



Proposal

7
:
ECG through WSNs

(June 2011)

Abstract

Sensor nodes are currently available to capture Elect
ro Cardio Graphic
signals from
nomadic patients which can then be remotely delivered through wireless transmissions to
other devices (e.g., cell phones and PDAs).

This thesis will target the realization of a comprehensive system for real
-
time ECG
through
sensor nodes. The work will include:



the analysis of proper compression techniques for ECG samples



the realization of proper transmissions techniques to deliver the collected sensors
remotely



the realization of an application to provide ECG feedback to the

end user (possibly
for mobile terminals)



Tools

The methodologies used during the thesis will include:

-
Simulation:
Matlab
(to test sample acquisition and compression techniques), TinyOS
-
nesC (to implement/simulate the detection techniques on sensor node
s)

-
Testbed Experimental study: TintOS
-
nesC (testbed development), Java/Android
(development of a minimal application)


Background

Foundations on Wireless Networking

Signal Processing Fundamentals

Fundamentals of programming languages (C/C++/Java)

Matlab
Fundamentals (not mandatory)

Curiosity, Initiative


Notes

Chulsung Park; Chou, P.H.; Ying Bai; Matthews, R.; Hibbs, A.; , "An ultra
-
wearable,
w
ireless, low power ECG monitoring system,"
IEEE
BioCAS 2006.


Advisor/Contacts

Matteo Cesana, cesana@elet.polimi
.it

Alessandro Redondi, Redondi@elet.polimi.it











Proposal 8: Visual Analysis in Wireless Sensor Networks

(October 2011)

Abstract

Traditionally, smart cameras perform image acquisition, and compression locally,
sending away a compact representation

of the fetched image which is then analysed
remotely. Such paradigm may not be suited to wireless sensor network where delivering
remotely a compressed image or even a sequence of images may be impaired (or simply
impossible) due to the limited resources
(bandwidth, energy, etc.). Enabling visual
analysis in energy
-
constrained WSNs requires departing from traditional solutions and
pursuing a paradigm shift that affects the way visual data is sensed, processed and
transmitted.

The key tenet is that most vis
ual analysis tasks can be carried out based on a succinct
representation of the image, which entails both global and local features, while it
disregards the underlying pixel
-
level representation.
In a nutshell, sensor camera may
extract and send only a suc
cinct representation of the captured image rather than
delivering the full
-
fledged image itself.

To this extent
,

it is imperative to optimize the computation
, coding

and communication of
the visual
features. That is, image features are collected by sensin
g nodes,
are
processed,
and
are then
delivered to
the
final destination(s) in order to enable higher level visual
analysis tasks by means of either centralized or distributed detectors and classifiers
.

All
these steps are subject to tight application
-
depen
dent requirements (bandwidth/delay
guarantees), and may be affected by network conditions
, communication and hardware
constraints
.

Such paradigm shift in the coding/processing functionalities calls for
enhanced and innovative networking algorithms and prot
ocols.

The goal of this thesis is the study and design of comprehensive solutions to support
feature extraction in WMSNs. In particular, the work will address:



The design of effective feature extraction, coding algorithms for sensor cameras



The design of
coordination/orchestration protocols to support distributed processing
of features among neighboring sensor nodes



The design of communication protocols to support feature delivery over multi
-
hop
WMSNs


Tools

The methodologies used during the thesis will in
clude:

-
Simulation:
Matlab
(encoding techniques), TinyOS
-
nesC (testbed simulation)

-
Testbed Experimental study: TintOS
-
nesC (testbed development


Background

and Notes

Foundations on Wireless Networking

Foundations of programming languages (C/C++/Java)

Matl
ab Fundamentals (not mandatory)

Curiosity, Initiative

The proposal can give raise to different theses


Supervisors
/Contacts

Matteo Cesana,
cesana@elet.polimi.it
,
Alessandro Redondi, Redondi@elet.polimi.it

Propos
al 9: Implement
ation of a

CoAP
-
Enabled Wireless Sensor Network

(October 2011)


Abstract



A recent initiative of the IETF has been launched to standardize an application layer
protocol for constrained devices/networks. The reference protocol is called Con
strained
Application Protocol
(CoAP)
and is devised to mimic the HTTP REST
-
full
communication paradigm. CoAP provides methods and commands to interrogate a sensor
device (equivalent to HTTP GET commands) and to modify the status of the sensor itself.

The
goal of this thesis is the implementation of CoAP in a
wireless sensor network

based
on TinyOS to showcase the performances of the CoAP in terms of latency and
robustness. CoAP libraries are already available (see reference []), so the required work
load
and novelty makes this proposal more suited to a short thesis.


The final outcome of the thesis will have to be an experimental testbed
running CoAP
over TinyOS.


Tools

The methodologies used during the thesis will include:

-
Testbed Experimental study: Ti
ntOS
-
nesC (testbed development), CoAP libraries ©



Background

Foundations on Wireless Networking

Foundations of programming languages (C/C++/Java)

Curiosity, Initiative


Notes

and References

[1]
http://d
atatracker.ietf.org/wg/core/

W e b S i t e o f t h e I E T F w o r k i n g g r o u p
s t a n d a r d i z i n g C o A P


[ 2 ]
h t t p://h i n r g.c s.j h u.e d u/j o o m l a/i m a g e s/s t o r i e s/c o a p
-
i p s n.p d f

R e s e a r c h p a p e r d e s c r i b i n g
t h e i m p l e m n e t a t i o n o f t h e C o A P



S u p e r v i s o r s
/C o n t a c t s

M a t t e o C e s a n a, c e s a n a @ e l e t.p o l i m i.i t

A l e s s a n d r o R e d o n d i, R e d o n d i @ e l e t.p o l i m i.i t