Title: Evaluating Service Oriented Architectures (SOA) in Pervasive Computing
Authors: Swaroop Kalasapur, Mohan Kumar and Behrooz Shirazi
Abstract: Increasing popularity of the pervasive computing paradigm on one hand and
the technological developments on the other, have paved the way for development and deployment of
pervasive services in the everyday habitat. Web services models will evolve to provide essential
support to users through specific services. In such a service oriented environment, users can utilize
the available resources to accomplish their tasks. To support a number of user tasks, a number of
simple services will be dynamically grouped together into complex services through the process of
service composition. The service providers, who build, operate and manage services will want to
maximize their revenues, and at the same time, the users will seek QoS guarantees to support
their applications. While there are a number of mechanisms proposed to build and operate service
oriented architectures (SOAs) in pervasive computing arena, very little work has been done in terms
of evaluating such schemes. The work presented in this paper proposes a technique for evaluating SOAs
in pervasive computing. We also present a mechanism to evaluate service composition techniques, which
are an effective way of delivering services to the end user. In earlier work, we proposed a service
composition mechanism in pervasive computing environments called Seamless Service Composition (SeSCo).
We evaluate and compare SeSCo with a Discover + Match mechanism for the deployment of pervasive
services within an automobile. We also present performance results of the two approaches through
simulation.
Title: Service Oriented Pervasive Computing for Emergency Response Systems
Authors: Swaroop Kalasapur, Kumaravel Senthivel and Mohan Kumar
Abstract: Pervasive computing paradigm is characterized by the presence of a diverse
variety of computing and communicating devices. Typically a modern day user carries a number of
personal devices with them. The vision of pervasive computing is to enable effective support for
user tasks by utilizing the devices carried by the users and those available within the infrastructure
around the users. Emergency response is a very critical to crisis situations. The process of emergency
response can benefit greatly by the application of pervasive computing principles. In this paper,
we describe a prototype designed to enhance the existing emergency response mechanism in attending
to automobile accident victims. The prototype has been built using our event oriented middleware
called Pervasive Information Communities Organization (PICO), and is built on top of a popular P2P
framework called JXTA. By utilizing the middleware, we model the useful features of devices available
within a car as services. Our proposed framework allows the creation, discovery and maintenance of
continuous services such as crash detection. Further more, the framework allows the composition of
complex services on occurrence of events by utilizing the available basic services with a goal to
provide essential support in case of an accident. We present the design and implementation details
of the prototype and snapshots of the working prototype.
Biography: Swaroop Kalasapur is a PhD candidate at Department
of Computer Science and Engineering, University of Texas at Arlington.
He is a member of PICO research group since 2002, and his advisor is Professor Mohan Kumar.