62
Mobile and Pervasive Computing (CoMPC–2008)
A Hypothetical Scrutiny of Pervasive Computing to an
Ubiquitous Computing through Web Technique
S.M. Roychoudri1, P.N.V. Satyasree Choudri2 and P.A. Abdul Saleem3
Department of Computer Science & Engineering,
ASR College of Engineering, Tanuku, (Affiliated to JNTU), West Godavari District, Andhra Pradesh
E-mail: 1roychoudri_76@yahoo.co.in, 2satyasreekpnv@gmail.com, 3saleemhod@yahoo.co.in
ABSTRACT: Ubiquitous computing refers to building a global computing environment where seamless and invisible
access to computing resources is provided to the user. Pervasive computing deals with acquiring context knowledge
from the environment and providing dynamic, proactive and context-aware services to the user. A Ubiquitous
computing environment is created by sharing knowledge and information between pervasive computing
environments. In this paper, we propose a framework that uses the potential of the Web Technique to weave
pervasive computing environments into a Ubiquitous-computing environment. We discuss how the collaboration of
these pervasive environments can create an effective Ubiquitous computing environment referred herein as the
Integrated Global Pervasive Computing Framework (IGPF). We test the effectiveness of the Ubiquitous environment
through a small scenario from a prototype system that we have implemented over this framework to handle Home
application, Automotive, Mobile team work, and medical emergency scenario.
Keywords—Ubiquitous Computing, Pervasive Computing, Web Technique.
INTRODUCTION
W
e believe that Ubiquitous computing is the next
wave of computing after the Internet wave.
Ubiquitous computing aims to revolutionize the current
paradigm of human computer interaction. Computers have
been used in various aspects of human life, but in most
cases human beings have to adapt their behavior to each
system. Ubiquitous computing as envisioned by Weiser is a
computing environment computing systems weave
themselves in the fabric of everyday life and become
invisible. Invisibility is the most important aspect of
Ubiquitous computing. The user is exposed to a few sets of
services available to him/her and is oblivious to the
complex system implementing those services. This takes
the human-computer interaction into a whole different
dimension, where the user is surrounded by a complete
smart environment with devices/sensors communicating
with each other and aggregating their functionalities to
provide a set of consolidated services. The terms
Ubiquitous computing and Pervasive computing are used
interchangeably, but they are conceptually different.
Ubiquitous computing uses the advances in Mobile
computing and Pervasive computing to present a global
computing environment. Mobile computing is about
elevating computing services and making them available on
mobile devices using the wireless infrastructure. The focus
here is to reduce the size of the computing devices so that
they can be carried anywhere or by providing access to
computing capacity through high-speed networks. But
Mobile computing has some limitations. Pervasive
computing, on the other hand, is about acquiring context
from the environment and dynamically building computing
models dependent on context. Pervasive computing is
invisible to human users and yet provides useful computing
services. Ubiquitous computing aims to provide Pervasive
computing environments to a human user as she/he moves
from one location to another. We present an alternate
approach and use Web technique for Pervasive computing
environments. This allows context information to be stored
on the Web and then shared across Pervasive computing
environments via the Web to provide context-aware
services.
MOTIVATION
Existing methodologies for implementing a Ubiquitous
computing environment use smart devices, which have
some processing power and are specialized in a set of
specific tasks. Usually the user needs to carry these devices
with her/him as s/he moves either within or across
pervasive environments. These devices are not readily
available and are often difficult to build. Our solution
eliminates the need for smart devices by using the Web
Technique to build dynamic context models as a user
moves from one environment to another. We can achieve
dynamic building of contexts by sharing knowledge and
context information between local pervasive environments
through the Semantic Web.
PROPOSED SOLUTION
The idea of creating global smart environments doesn’t
need to expand a smart space to a global level. This is
similar to the idea of the global Internet, where numerous
small physical networks connect together to form a large
A Hypothetical Scrutiny of Pervasive Computing to an Ubiquitous Computing through Web Technique
homogenous network. Similarly, a better approach here
is to connect these smart environments together
‘Semantically’ so that they conglomerate into a virtual
global smart environment similar to the Internet, creating a
Ubiquitous environment.
Fig. 1: User rôle Model in Ubiquitous Environnent
65
A Home Application Scenario
You arrive at home, tired after a day’s work. The house
identifies you and unlocks the door. Recognizing the
fatigue on your face, the home entertainment system starts
playing your favorite relaxing music, and replaces the
artwork on the electronic wallboards with soothing images
of nature. As you enter the bathroom, you are asked if a
warm bath should be started. As you enter the kitchen, the
display on the refrigerator door suggests a light meal and a
specific recipe on the basis of the available ingredients in
the refrigerator and fitting your mood. If you accept the
suggestion, step-by-step instructions are displayed, helping
you through the steps of the recipe. Once you place the
bowl on the stove, the temperature is automatically
adjusted to control the cooking time on the basis of your
decision whether to take a bath.
In daily life, an average human being is associated with
many different environments and has different roles in
each. For example, in Figure 1, Dr. Smith is associated
with environments like hospital, bank, and pub and plays
the role of doctor, client, and customer, respectively. A
Ubiquitous system should truly reflect this real life
scenario.
Each of these environments may be an independent
smart system since they belong to different
domains/organizations they are disparate. What’s important
is that as the user moves from one environment to another,
the system transparently recognizes the user and associates
the user’s specific roles with the domain the user enters.
The system passes information required by the user
(files/contact information, etc.) between domains,
accessing/updating the user profile as needed, etc. which
allows the user to receive appropriate services. This true
sense of Ubiquitousness can be achieved only if the two
environments are semantically integrated.
An Automotive Scenario
PERVASIVE COMPUTING SCENARIOS
A Health Scenario
Scenarios
Your health monitor, possibly installed in your watch,
notices that your blood sugar is suddenly elevated. It starts
to monitor other conditions more closely to uncover
possible causes and potential problems. It contacts your
medicine cabinet to ensure that you have adequate
medicine in your home and contacts the pharmacy for
additional medic- ine if necessary. It reports the change to
your electronic medical record and sends a short message
to your doctor, informing the doctor of your current
location and evening plans in case the doctor needs to
contact you. Back at home, the kitchen notifies the
refrigerator to mark all sweets in your home off-limit for
you and suggests a healthy recipe for the evening! These
scenarios may appear far-fetched, although they rely mostly
on known technologies. The same technologies may be
used today for more near-term scenarios. In our European
The possibilities for applications of pervasive computing
are only limited by our imagination. Many imaginative
scenarios have been expounded in the literature. Such
scenarios help in recognizing the many possibilities and
also help to identify research goals. Some scenarios are
visionary ones that aim to establish long-term goals, the
parameters of the needed research, and the interdependence
of the research in pervasive computing with that of other
domains. The less visionary ones are used to define shortto mid-term research or product-development projects.
Some scenarios are created for advocating the importance
of the field and drumming up support. Let us consider the
first four different application domains: home, automotive,
mobile and health.
You are driving to a distant area for a two-week vacation.
You plan to explore the area by car during your stay. On
the way, your engine detects an irregularity in the operation
of the motor. It flashes a warning sign to you cautioning
you to drive more carefully and starts a diagnostic
procedure. It detects the offending part and sends a report
to your car manufacturer. The manufacturer locates the
garage nearest to your location and dispatches the part to
the garage from a nearby depot. The garage sends a
replacement car to meet you on the highway. In the
meantime, the police have been notified to look for possible
traffic tie-ups in your area. The car display directs you to a
specific parking area off the highway to meet the driver that
delivers your replacement car. You continue on your
vacation and the driver takes your car to the garage. Once
the car is repaired, it will be delivered to your hotel, using
the information from your itinerary posted by your travel
agent, aided by the positioning system in your car.
66
project called MOTION, we are building an infrastructure
and tool to support the following scenario:
A Mobile Teamwork Scenario
A multi-national organization has many sites around the
world and employees who travel frequently. Traditional
working groups, with well-defined management and
reporting structures are being augmented with, and in some
cases replaced with, adhoc teams formed for a welldefined, focused, short-term purpose. For example, a
particular design flaw is suspected in a product, such as a
mobile phone, that is deployed in the field. A team must be
formed from potential expert employees around the world
to examine, discuss, and propose a solution to the problem.
Virtual meetings must be held to support communication
and problem solving. Some members may be traveling and
have access to limited computer and display capabilities.
They may be in different time zones. The infrastructure
must support access to large amounts of information and
documents, synchronous and detached communication
among people, and event notification. For example, one
member may need to be notified as soon as some particular
document becomes available or another member comes online. Many of the needed technologies to integrate such a
system, such as chat and instant messaging are already
available. But they need to be integrated in new ways to
meet the demands of the new environment. The pervasive
computing environment is characterized by heterogeneity
(of devices and networks) and unpredictability and lack of
structure of communication patterns.
INFRASTRUCTURE FOR PERVASIVE
APPLICATIONS
Infrastructure
The architectures must enable high interconnectivity,
highly dynamic interactions among diverse components,
and a high level of coordination among different and
heterogeneous, possibly independently-designed activities.
An infrastructure for pervasive computing must therefore
enable components to recognize, communicate with, and
coordinate with each other. While the infrastructure should
enable free and rapid interaction among components, not
all accesses should be allowed.
Software Components for Pervasive Computing
Pervasive services will have to be composed from
individual “components” residing in the large number of
heterogeneous computing elements. The hardware
environment itself will force a natural boundary between
components. This may be the most clear-cut definition of a
component. A component will be an independently
deployable piece of software that resides on one hardware
element and provides a service element. Of course, there
Mobile and Pervasive Computing (CoMPC–2008)
may be more than one component on each hardware
element. Just as Web Services are emerging in the Web
computing infrastructure as a “component”.
Components are basically of the same “size” and
“power.” For example, JavaBeans components are for
desktop environments while Enterprise JavaBeans are for
server and enterprise-wide components. To make
application development manageable, we probably need a
single component model that is “scalable”.
One of the key problems of building applications out of
components—component based software engineering—is
what to do if the component you need is not available in the
catalogs you have. Clearly, no catalog will have every
component that an application developer needs. But, often,
there will be a related component, or one that is “almost”
the one needed. There are several possible paths to take in
this case. One is for the developer to modify the related
component to make it fit his needs. A more effective
approach is to automatically “adapt” the existing
component to the need of the application.
Device Heterogeneity
The basic premise of pervasive computing everything
connected guarantees heterogeneity at all levels:
infrastructure, hardware, software, and people. All kinds of
devices must be supported. In general, the environment
must anticipate the existence of a wide variety of devices.
If we consider devices used by the user to interact with the
system, they can range from standard ones such as laptops,
PDAs, and phones, to emerging ones such as those
embedded in clothing and eyeglasses.
One is the kind of input-output devices: textual and
graphic input-output will not be the only forms of humanmachine interaction. Audio, visual, and other sensory
modes of communication will be prevalent. Another
implication is the requirement that the environment must be
prepared to adapt to the device currently used by the user.
For example, if the user is requesting information and he is
currently driving, the retrieved data should be relayed to
him with an audio message through the car radio.
Access Control
The wide availability of services and the high mobility of
users among different environments require the provision
of security mechanisms to ensure the safe usage of services
by legitimate users and the protection of services from
unauthorized uses. Because of the wide range of services,
many diverse and flexible security models and mechanisms
will be needed.
One of the most important aspects of security is access
control, to ensure that services are only available to
authorized users and those authorized users are allowed
appropriate privileges. For example, a guest at a hotel may
be allowed to print on the hotel’s printer available in the
A Hypothetical Scrutiny of Pervasive Computing to an Ubiquitous Computing through Web Technique
lobby but not change the contents of the event display in
the same lobby. We have been working in two directions to
help application developers include sophisticated access
control in their applications.
The access control subsystem and the architectural style
mentioned here are two building blocks that may be used
for pervasive computing applications. But much more is
needed. In fact, more flexible and powerful security models
are needed to protect services that are composed of
distributed service components. It is reasonable for a user
to provide a password once to authenticate him or her. But
the user should not have to provide passwords each time a
new service is being contacted.
ARCHITECTURE OF IGPF
Following points were considered while designing
Integrated Global Pervasive Computing framework (IGPF).
IGPF should be generic. That means we should be able to
use IGPF to setup a pervasive computing environment for
any domain. For example, the IGPF can be used to setup a
pervasive computing environment for a hospital or for an
office with little modification. For example, one of the core
components, the KB Handler is used to query and inference
on the knowledge base of the system. The domain specific
components provide knowledge and services specific to the
domain. If the IGPF is used to setup a pervasive computing
environment in a hospital, then the knowledge base will
consist of domain specific ontology’s like the drug
ontology and patient ontology; IGPF should use shared
knowledge. We use the Web Technique based approach
where the shared knowledge is stored on the Web and it
can be accessed by different environment.
UI Manager
Profile Manager
KB Handler
Resource Manager
Unified
Information Base
Autonomous System Head
Fig. 2: IGPF Architecture
E
V
E
N
T
M
A
N
A
G
E
R
67
CONCLUSION
The approach presented in this paper introduces a whole
new way of implementing Ubiquitous computing by
semantically connecting different and diverse pervasive
environments. Connecting these environments through the
Web Technique enables sharing of knowledge and
information across pervasive environments, thus allowing
environments to dynamically create context models for a
user as s/he walks from one environment to another. The
impact of the designed framework is that it creates the
generic base for achieving these basic functionalities along
with offering the user a set of abstract services.
Furthermore, the environments can be ‘programmed’ to
achieve specialized applications that use the services within
a pervasive environment and also share and utilize the
services offered through access to other environments. We
believe that the impact of such a framework will speed up
the process of creating local pervasive environments and
build true Ubiquitous environments based on them. As a
whole, our work aims to use computing resources to create
Ubiquitous environments to improve the way we utilize
services that pervasive computing environments provide
and in general, benefit our daily lives.
REFERENCES
[1] Berners-Lee, T. (1994). Universal Resource Identifiers in
WWW: A Unifying Syntax for the Expression of Names and
Addresses of Objects on the Network as used in the WorldWide Web Retrieved December 3, 2004.
[2] Berners-Lee, T. (2003). Primer: Getting into RDF &
Semantic Web using N3. retrieved December 3, 2004, from
http://www.w3.org/2000/10/swap/Primer
[3] Berners-Lee, T., Hendler, J. & Lassila, O. (2001). The
Semantic Web. Scientific American.
[4] Berners-Lee, T., Masinter, L. & McCahill, M. (1994).
Uniform Resource Locators (URL). (RFC 1738) CERN,
Xerox Corporation, University of Minnesota.
[5] [Fenkam01] Fenkam, P., A Dynamic User Management
Access Control for Web Sites”. Distributed Systems Group,
Master’s Thesis, Technical University of Vienna and
Institute for Software Technology, Technical University of
Graz, 2001.
[6] [Fenkam02] Fenkam, P., Gall, H. and Jazayeri, M., “DPS—
An Architectural Style for Development of Secure
Software.” In Proceedings of the Infrastructure Security
Conference (Bristol, 1–3 Oct, 2002), Springer Verlag.
[7] [Garlan02] Garlan, D., Siewiorek, D.P., Smailagic, A. and
Steenkiste, P., “Project Aura: Toward Distraction-Free
Pervasive Computing,” Pervasive Computing, April–June
2002.
[8] [Gschwind02] Gschwind, T., Adaptation and Composition
Techniques for Component-Based Software Engineering,
PhD dissertation, Distributed Systems Group, Technical
University of Vienna, 2002.