CS 536 Project Topics (update in progress) 
  Following is a list of possible project topics for CS 536 (Fall 1997).
  The list is not meant to be exhaustive; it provides only a guideline.
  Independent project ideas are encouraged.
  Three categories of projects can be distinguished: first, the
introduction of an original idea and its exploration; second, implementing
and investigating a nontrivial variation of an existing 
algorithm, implementation, or idea; third, survey of a current research
area.
  The stringency of ``professionalism'' expected
(i.e., the level of polishedness
of the final report) increases as we go from the first category to the
third.
  Care should be taken to choose a scope of work that can be accomplished
within a period and workload specified in Assignment XII. 
  The final report should be 8-12 pages long, including figures
and references.
Multi-access control
Congestion Control
-  Congestion control for reliable transport:
  	Build a reliable transport mechanism on top of UDP incorporating
	congestion control; to be used mainly for file transfers.
	Compare performance against TCP-based implementation.
	Approaches:
  
  -  Rate-, window-based control.
  
-  Link-based control.
  
-  Others.
  
 
-  Congestion control for real-time multi-media transport:
  	Build an unreliable transport mechanism on top of UDP incorporating
        congestion control; to be used for transfer of real-time voice
	and video.  The goal is to provide quality of service (QoS)
	``guarantees'' (various levels of stringency) in terms of
	packet loss rate, end-to-end queueing delay, and their variance
	(e.g., jitter).
	Approaches:
  
  -  Reservation-based control.
  
-  Forward error-correction.
  
-  Others.
  
 
-  Survey/discussion of congestion control issues in ATM networks:
	Classification of congestion control approaches; special problems
	arising in B-ISDN systems including ATM; possible solutions and
	limitations.  New approaches?
Routing
-  Routing of multi-media conversations with QoS constraints:
	Design/implement a routing algorithm that incooperates QoS
	constraints such as end-to-end delay, bandwidth, etc.  The goal
	is to craft a fast method using the greedy (i.e., ``least-cost path'')
	paradigm.
	Approaches:
  
  -  Dijkstra, Bellman-Ford.
  
-  Others.
  
 
-  Routing using competitive ratio: Exploration of the
	feasibility of implementing practical routing algorithms using
	the notion of competitive ratios; evaluation of the
	method vis-a-vis shortest path algorithms.
Quality of service provision
-  Fair queueing and QoS: Implement an efficient scheme for
	weighted fair queueing.  Demonstrate its performance with
	respect to facilitating various levels of QoS for
	traffic streams tagged by different priority numbers.
	Issues:
  
  -  Queue management.
  
-  Multi-dimensional QoS vectors (e.g., packet loss rate,
	queueing delay, etc.).
  
 
-  Admission control and traffic shaping: If a new
	conversation with a certain QoS requirement is to be admitted,
	determine whether the QoS requirements can be met; also,
	upon admission, police individual traffic flow by
	administering traffic shaping.
	Approaches:
  
  -  Route set-up and bandwidth reservation (e.g., RSVP).
  
-  Leaky bucket.
  
-  Others.
  
 
Traffic characterization
Transparent distributed services/computing
Network tools
-  RTT estimation: Design new schemes for round-trip time
	estimation and its use in timeout selection.  Characterize
	the distribution of RTTs.
-  Inference of network topology/state: Use packet probing
	to infer the topology of a network and its state including
	link delay, effective bandwidth, and their variance.