Resource and Quality of Service Management for the Next Generation 
Internet 

(White Paper)

Raymond Paul
Deputy Director
Testing and Evaluation, Dept. of Defense, 3000 Pentagon, Washington 
D.C. 
Tel: (703) 697-5806, Email: paulra@acq.osd.mil 


A. Bashandy
Graduate Student
Distributed Multimedia Systems Laboratory School of Electrical and 
Computer Engr., Purdue University, 
West Lafayette, IN, 47907
Tel: (317) 494-9917, Email: bashandy@ecn.purdue.edu 



Arif Ghafoor (Contact Author)
Associate Professor
Distributed Multimedia Systems Laboratory School of Electrical and 
Computer Engr., Purdue University, 
West Lafayette, IN, 47907
Tel: (317) 494-0638, Fax: (317) 494-4471, Email: ghafoor@ecn.purdue.edu 



1. Objective

Demand to support large-scale multimedia services over the Internet is 
expected to
grow substantially in the near future. The next evolution of the Internet is 
geared
towards supporting quality-oriented services for a broad range of 
applications. One
of the goals is to provide a balance between the cost-effectiveness of the 
technology
on one hand and, on the other hand, supporting a large array of quality-
oriented applications in a flexible manner. The objective of our research is 
to design and
develop basic infrastructure for the Next Generation Internet (NGI). 
However, prior
to its deployment, considerable security and performance enhancements are 
required
for the proposed NGI. In this white paper, we describe our ongoing research 
efforts
in these arenas. Our effort is based in the distinguished research facilities of 
the
Distributed Multimedia Systems Laboratory at Purdue University. 




2 Quality of Service, Resource Management and Routing Policies for NGI 

Included among the major technical issues being addressed by us is the 
design and
development of network resource management and route establishment 
policies for Internet-based multimedia applications. A primary 
consideration for such development
is to support the notion of application-specific quality of service. The 
quality
needs for many existing and future NGI applications can be specified in 
terms of several parameters including precision, resolution, timeliness, 
security, accuracy
etc. These diverse requirements place considerable constraints on the 
network resources and consequently demand a high degree of flexibility in 
network control and
resource management. This necessitates intelligent design of robust, 
efficient and
scalable policies for management of NGI. 

In particular, our research is focused on the following three major issues: 

- How various application level quality parameters can be mapped onto 
network resource
requirements such as bandwidth, buffering, and caching? 

- How can the network resources be controlled and allocated in an optimal 
manner?
What types of quality-of-service based routing policies can be designed for 
the NGI?
What levels of overheads and performance result from these policies? What 
role, if
any, can distributed computing play in the implementation of these policies 
in the NGI
environment?

- How should resource negotiation take place between an application and 
the network?
In resource constrained environments, how can the end-systems utilize the 
network
resources in an intelligent manner to ensure highest possible levels of 
quality? 


A major step forward in this direction is the ongoing effort by several 
industrial
and academic organizations to standardize application/Internet service 
request protocols such as RSVP. However, the current functionality of 
RSVP is limited to
just the resource procurement aspects of the overall route establishing 
procedure.
To make RSVP useful for the NGI environment, considerable 
enhancements to its functionality
are required. As a result of our research we expect to incorporate resource 
negotiation
process within the RSVP. Along with this we expect to make a seamless 
integration of
RSVP as part of the resource management mechanism of the NGI. 

3. References

1. A. Ghafoor, A. Lazar, I. Leslie, and R. Herrtwich, "Distributed 
Multimedia Systems
and Technology", IEEE Jour. on the Selected Areas in Communications, 
14(7):1209-1213,
September 1996.

2. S. Baqai, S. Seiichi, M.F. Khan, M. Woo and A. Ghafoor, "Performance 
Evaluation of
Multimedia Synchronization Protocols Using Server-Based Scheduling 
Mechanisms",IEEE
Jour. on the Selected Areas in Communications, 14(7):, September 1996. 

3. S. Baqai, M. Woo, and A. Ghafoor, "Network Resource Management for 
Enterprise-wide
Multimedia Services", IEEE Communications, 34(1):78-83, January 1996 

4. Z. Ali and A. Ghafoor, "Synchronized Delivery of Multimedia 
Information over ATM
Networks", to appear in the Communications of the ACM.