RESEARCH INTERESTS
We have developed a hardware-software platform
independent, abstract
information sharing and integration model (AISIM) for designing a
middleware architecture. Using our AISIM model we designed a middleware
architecture with CORBA
(Common Object Request Broker Architecture) objects. A small prototype
of the middleware was
implemented through EJBs (Enterprise Java Beans).
Present Internet and (World Wide) Web are providing service for diverse
multimedia communication. It is predicted that in the future multimedia
communication will generate most of the traffic for Internet and Web.
Unlike CBR (constant bit rate) voice communication in POTS (plain old telephone
systems), resource requirements for VBR (variable bit rate)
multimedia communication in Internet and Web is not
identical for all users.
Design of broadband communication networks for Internet and Web services
requires good understanding of needs of multimedia systems. Focus of
current
research is analysis and modeling of VBR multimedia for such use as designing
and testing
of broadband communication systems that will support multimedia
traffic.
We have developed methods for analysis and modeling of MPEG-like encoded
full-length
VBR video. The information obtained from the analysis are used for
creating FSM (Frame Size Model). Our models are very useful for study of
loss of data during communication.
We have effectively used our sigle video FSMs to
develop a model
--- called multinomial model --- for prediction of bandwidth
requirement
when several videos are communicated simultaneoulsy over a link of a
broadband network. Our extensive tests with standard
full-lenght MPEG video traces have shown excellent agreent
between model predicted and observed bandwidth requirements.
Wired connections of 'terminals' --- computers, faxes, or phones ---
tie users
at a fixed location. However, users move, for instance, from home to
office. Mobile and wireless computing systems bring the convenience being
connected any time any where.
Thus, the demand for wireless communication is growing very fast.
However, the bandwidth for wireless
communication is limited and cannot be increased as it can be done easily
for wired communication. Apart from the limited available bandwidth,
noise and user mobility are two other major problems.
The most important issue to be addressed for the successful deployment of
mobile wireless systems is Quality of Service (QoS) to the users. Once
connected, a user will expect the connection to be as reliable as the
wired connection. This is very hard while users are moving from one cell to another
cell. Thus, a coordinated control system must be developed to support
QoS. The focus of our current research is to use such AI tools as neural
networks and fuzzy associative memory with our recently developed new call
preblocking technique to assure QoS for mobile users.
Middleware and Multimedia Systems
VBR Multimedia Modeling and Communication
Virtual Private Networks over the Internet for Multimedia
Communication
QoS in the Web
The widespread use of the Internet has created two problems: document
retrieval latency and network traffic. Caching of documents "close"
to users has helped to alleviate both problems. Different caching
policies have been proposed/implemented to make best use of limited
available cache at each caching server. A mesh of caching servers,
aided by different data diffusion algorithms and the natural hierarchical
structure of the Internet topology, has increased "virtual" size of cache.
Yet the size of available cache is small compared to the total size of all
documents served, and remains a major resource constraint. In this work,
we looked at how to improve document download time, by distributing a
fixed
amount of total storage in a network or mesh of caches. The intuition
behind
our cache distribution approach is to give more storage to the caching
nodes
in the network which experience more traffic, in the hopes that this will
reduce the average latency of document retrieval in the network. A
heuristic
was developed to estimate traffic at each cache of a network. From this
traffic estimation, each cache then receives a corresponding percentage of
the total storage capacity of the network. Through extensive simulation
it
is found that the proposed cache distribution algorithm can reduce latency
up
to 80% over prior work that includes both Harvest-type and demand-driven
data diffusion algorithms. Furthermore, the best improvement was achieved
in a cache range that corresponds to practical, real world cache ranges.
Mobile and Wireless Computing