Professional Documents
Culture Documents
Ac 2012-3303: Remote Experimentation For Communication: From Remote Desktops To Gateways
Ac 2012-3303: Remote Experimentation For Communication: From Remote Desktops To Gateways
Ac 2012-3303: Remote Experimentation For Communication: From Remote Desktops To Gateways
Grant Huang received his M.Sc. degree from the University of Texas, San Antonio (UTSA) in 2009 and
is currently a Ph.D. student in the Department of Electrical and Computer Engineering (ECE) at UTSA.
His research interests include satellite/wireless channel modeling, Assisted-GPS/GNSS (A-GPS/GNSS),
and remote experimentation systems.
Andreas Gampe is a Ph.D. student at the University of Texas at San Antonio. He received a diploma
degree in Computer Science from the Friedrich-Schiller University, Jena, Germany, in 2006. His current
research interests include language-based security for computer systems, embedded languages in software
engineering, and online education.
Arsen Melkonyan is a Ph.D. student at the University of Texas, San Antonio. From 2005 to 2007, he was
a member of the research staff of Tampere University of Technology, Finland. He received a B.Sc. degree
in electrical engineering from the State Technical University of Armenia, Yerevan, Armenia, in 2003 and
a M.Sc. degree in electrical engineering from the University of Texas at San Antonio in 2008. His current
research interests include impact localization algorithms for structural health monitoring, WLAN Indoor
positioning technologies, and remote hardware based educational system design.
Murillo Pontual is a Senior Member of the technical staff at Oracle. He received his Ph.D. in computer
science from the University of Texas, San Antonio in 2011, and he received his M.Sc. and B.Sc. degrees in
computer science from the Federal University of Pernambuco, Brazil, respectively in 2005 and 2003. His
current research interests include formal methods, online education technologies, and computer security
with special emphasis in access control, security policies, and obligations.
David Akopian is an Associate Professor at the University of Texas at San Antonio (UTSA). He joined
UTSA in 2003 where he founded Software Communication and Navigation Systems Laboratory. He re-
ceived the M.Sc. degree in radio-electronics from the Moscow Institute of Physics and Technology in
1987 and Ph.D. degree in electrical engineering from the Tampere University of Technology (TUT), Fin-
land, EU, in 1997. From 1999 to 2003, he was a Senior Engineer and Specialist with Nokia Corporation.
Prior to joining Nokia in 1999, he was a member of teaching and research staff of TUT. His current re-
search interests include digital signal processing algorithms for communication and navigation receivers,
positioning methods and mobile applications, and remote labs
Page 25.1116.1
American
c Society for Engineering Education, 2012
Remote Experimentation for Communication: from Remote Desktops to
Gateways
Grant Huang, Andreas Gampe, Arsen Melkonyan, Murillo Pontual, and
David Akopian
Abstract
1. Introduction
Nowadays, laboratories that work with hands-on experiments are among the core components
of engineering studies. Hands-on experiments help efficiently digest theoretical concepts and
train students to rely on the facts according to physical evidence. On the other hand, high
equipment cost and the lack of the necessary maintenance and assistance have led to reduced
importance of hand-on laboratories on the curriculum. Software simulators, based on
mathematical models, can be an alternative method to replace the traditional hands-on
laboratories; however, the valuable pictorial element is removed from the practical learning
processes.
The continuous progress of information technology combined with the increasing stability of
the Internet provides a potential environment for Web-based remote applications. A web-based
Page 25.1116.2
remote laboratory application can overcome various difficulties and limitations in offering
hands-on training (i.e. minimized expenses, flexible times, and resources sharing among
students). Unfortunately, only physical experiments, which have PC-controlled features can
be implemented in remote settings, which limit the conceptual applicability of the specific
educational areas such as radio-communications. One of the main design challenges is to
establish an architecture that can provide adequate access to the remote hardware in the
remote laboratory development. Hardware that generates real time signals is the core factor in
the remote laboratory architecture, while software is the alternative hardware management
component, which provides connection between the user orders and the hardware. In addition,
a software layer provides a Graphical User Interface (GUI) where the experiment results can
be visualized.
The simplest remote laboratory may be a single user, one-by-one experiment that reports
feedback data of visualized measurement results from the hardware. Moreover, more
sophisticated systems may handle several experiments with different apparatus and manage
multiple user accesses.
The latest generation of eComLab is based on the gateway architecture, which connects
multiple instructors, experimental equipment, and users from different locations via Internet.
The GUI of the current eComLab platform includes several handy features, which help users
interact (chat rooms) and monitor the hardware with real-time video stream through the
motorized Web-camera. The eComLab uses the NI ELVIS [4] /DATEx [5]
telecommunications bundle as the hardware-based experimental equipment. eComLab is a
very flexible platform, and it is capable to connect new educational experiments (hardware
laboratories) directly to a gateway (central) server, or from remote locations. This platform
allows multiple users to access and operate with the same experiment simultaneously by
following two parallel (the queue and the group policy) methods. Finally, the eComLab is
accessible through a regular Web-browser with Java and Flash plug-ins, which eliminates the
necessity to install any additional hardware driver in the user computer. The detailed
information about the different eComLab generations is discussed in the following sections.
2.1 Architecture
The first generation of eComLab [6] is based on a straightforward system architecture that
uses conventional remote desktop application (Teamviewer [7]) to connect the users to the
experiments. The host computer known as the experiment server is attached to the hardware
equipment (NI ELVIS/DATEx). It contains all the necessary software components (NI
ELVISE and DATEx SFP) to control the hardware and monitor the results on the host
machine. Figure 1 shows the system architecture of the first generation eComLab. The first
generation of the eComLab is based on one-by-one communication system architecture.
Multiple requests are treated with the first-in-first-out queuing policy, parallel connections are
not supported.
Figure 2 is a snapshot of the first generation eComLab’s GUI. In this architecture, eComLab
uses the TeamViewer as the remote desktop software.
Page 25.1116.4
Once thhe users connnect to the host/experiiment machine, they haave complette control ov
ver the
experim
ment-relatedd software. They
T can m
manipulate hardware-ba
h sed experim
ments remottely.
Connecction is estabblished usin
ng TeamVieewer with th heir corresponding ID aand passwo ord. In
additionn, Teamviewwer has several handy bbuilt-in fun
nctions (e.g. chat and fille transfer),, which
can be uused duringg the remotee experimennt.
3. Second Generration of eC
ComLab—C
Centric Serrver
3.1 Arcchitecture
The seccond generaation of the eComLab
e w
was designeed to addresss the defectts of the firsst
Page 25.1116.5
The num
mber of virttual machin
nes that the ccentral serv
ver can host depends onn several facctors:
networkk bandwidthh, experimeental equipmments, central server haardware connstarints (e.g
g.,
memoryy size, processor speedd, etc.) The eeComLab has
h three virrtual machinnes with
ELVIS//Emona DA ATEx experiiment traineer for commmuncation ex xperiments attached. The
T
system was servingg a wirelesss communiccation class with 25 students.
Althouggh, the response collectted from thee users were very posittive. Some uusers asked
d for
some otther features, among thhem, such ass experimen nt control sw
witch functiion among the
t
users w
without the queuing
q con
nstraints, cam
mera contro ol (i.e. zoom
m in, zoom oout and chan
nge the
video sttreaming anngle), and a dedicated ggeneral disccussion boarrd for all useers.
4. Currrent Geneeration of eC
ComLab—
—Gateway Server
S
4.1 Arcchitecture
The lateest version of
o eComLab b is the com
mbination off the centrall server andd server as gateway
g
architecctures. The central serv
ver can eitheer host expeeriment macchine or connnect experiiment
Page 25.1116.7
machines located in remote lo ocations throough the nettwork. The gateway seerver architeecture,
shown iin Figure 5,, connects four
fo remote sides through the Interrnet: PC-conntrolled exp
periment,
studentss, instructorrs, and systeem administtrators.
Fig. 5. G
Gateway Server architecture
main WWeb page to the contacts, the surveey, the material area, the discussionn wall, and the
experim
ment area. The
T user can
n log off from
m the system anytime by using thhe logout bottom
located in the menuu bar.
In the suurvey sectioon, studentss can compllete a short survey, whiich helps deevelopers to
o collect
user feeedback abouut the system m in order tto improve it.
i Experimeental materiials and tuto
orial in
text or vvideo formaat are availaable in the m
material areaa. In additio
on, studentss are able to see the
list of eexperiments in the expeeriments areea. The eCo omLab defin nes the expeeriment statu us
throughh special im
mage figures,, which are shown in th he status column. It infforms user that
t the
experim
ment is free, and it is av
vailable for full use if no
n other user is in the vvirtual experriment
room. TThe system can also notify the userr about whaat experiment is being uused and wh hether
the userr can join thhe group as an observerr. When thee maximum number of users for an n
experim
ment is reachhed, the eCo omLab autoomatically blocks
b accesss to the expperiment.
After chhoosing an experiment, a user acceesses the ex xperiment arrea that inclludes the reemote
desktopp applicationn, the real tiime video, tthe chat roo
om, the statu
us and the ccontrol bars as
shown iin Figure 7. The remote desktop ap application is
i an applet that allows students to
o control
the expeeriment remmotely. Thiss applet periiodically co ommunicates with the mmain centrall server
to checkk which stuudent has thee right to coontrol the ex
xperiment, how
h many uusers are in the
experimment room, and
a automaatically updaates the user queue. No ote that the ssystem allow
ws only
one studdent to havee control ov
ver an experriment at a time.
t However, other sttudents in the
virtual eexperiment room can observe
o the experimentt, the hardw
ware from reeal-time videeo
streaminng, and disccuss with otthers using the built-in chat room.
The conntol bar is composed off four icons at the bottoom, next to the status bbar. They prrovide
several additional useful
u functtions: remotte desktop transmission
t n rate controol, quality refresh,
r
full screeen mode, and
a passing experimentt control to another useer. The last ttwo functio ons were
implem mented in ordder to fix th
he problemss that were reported
r fro
om generatioon two by thet users.
Full scrreen mode was
w designeed in order tto present th
he users a biigger workiing area. Thhe
experimment controll passing function allowws a user to
o pass the coontrol of a eexperiment to
t
Page 25.1116.10
Fig. 8. Adm
ministration in
nterface
Chart 1 summarizees survey results (meann value) con nducted amoong the 97 uundergraduaate and
34 gradduate studennts. The threee eComLabb generation
ns are noted
d as first (F)), second(S)), and
Third (TT). Chart 2 shows the correspondi
c ing standardd deviation for
f three eC ComLab
generations.
Cha
art 1. Summaarized survey rresults (mean value) for three eComLab ggeneration
The stanndard deviaations of thee “usability”” (SD=0.96 (F), SD=0.95 (S) and SD=0.85 (TT)),
“usefulnness” (SD=
=1.01 (F), SD D=0.97 (S) and SD=0.9 95 (T)), and
d “acceptancce” (SD=1.17 (F),
SD=1.009 (S) and SD=1.07
S (T))) are decreaased from one
o generatiion to anothher. The resuults
show thhat the numbber of studeents who annswer positiv
vely to the survey
s quesstions has in
ncreased
over thee eComLab developmeent generatioons.
Chart 2. Stan
ndard Deviatioon for three generation of eComLab
e
nclusion
6. Con
[1] H. Shen, Z. Xu, B. Balager, V. Kritiansen, O. Stream, M. S. Shur, T. A. Fjeldly, J. Q. Lu, and T. Ytterdal,
“Conducting Laboratory Experiments Over the Internet”, IEEE Transaction on Education, Vol. 45, No. 2,
pp. 145-151, 1999.
[2] B. Hanson, P. Cumer, J. Gallagher, K. Page, E. Read, A. Weightnam, and M. Levesley, “ReLOAD: Real
Laboratories Operated at a Distance”, IEEE Transaction on Learning Technologies, Vol. 42, No. 4, pp.
331-341, 2009.
[3] M. D. Lowe, E. Lindsay, V. Lasky, and D. Liu, “Experiences with a Hybrid Architecture for Remote
Laboratories”, Proc. 38th ASEE/IEEE Frontiers in Education Conf., 2008.
[6] A. Melkonyan, D. Akopian, and C.L. Chen, “Work in Progress—Real-Time remote Internet-Based
Communication Laboratory”, Proc. 39th ASEE/IEEE Frontiers in Education Conf., San Antonio, TX ,
October, 2009.
[8] A. Melkonyan, M. Pontual, G. Huang, A. Gampe, and D. Akopian, “eComLab: Remote Laboratory
Architecture for Radio-Communication”, Proc. 2011 ASEE Gulf-Southwest Annual Conf., Houston, TX,
March, 2011.
[9] M. Pontual, A. Melkonyan, A. Gampe, G. Huang, and D. Akopian, “eComLab: Remote Laboratory
Platform”, Proc. SPIE 8036, 806311 Conf., Orlando, FL, April, 2011.
Page 25.1116.14