Dynamic Time Sharing : A New Approach for Congestion

His current research interests include ATM networks (with focus on traac control), diierent aspects related to self-similar traac (parsimonious modeling, fast generation, implications) as well as architectures and protocols for computer networks. He is a member of the IEEE, IEEE CS and ACM. His email address is [email protected]. Masters Degree in Computer Science from Stevens Institute of Technology in 1987. After that he worked for Bellcore as a Member of Technical Staa and he concentrated on conformance testing of and automated test tool development for X.25, X.75, ISDN and SS7 networks. Later he was in charge of the Internetworking and Interoperability work where he resolved many switch in-terworking problems. In 1993 he took a leave and went to the Royal Institute of Technology, Stockholm, Sweden and in 1995 he graduated with a Ph.D. degree in telecommunications systems. He continued at Bellcore for a year and was in charge of their data analysis and traac modeling work. Today he is working at Qmetrix International where he is responsible for application of fractal traac models and methods for network systems design, planning, engineering, and management. His email address is [email protected]. CONCLUSIONS 21 for service, resource availability and fairness criteria. The conclusion therefore seems to be that in obtaining delay performance limited by optics (i.e., the propagation delay on ber) one should pay with a certain resource wastage (in the control channel). These are topics for future research. The DTS approach is applicable to other topologies as well and this is a topic for further research. Other areas of interest are related to assessing the applicability of this approach to VBR video traac, study of the eeciency in utilization of network resources, diverse trade-oos between QoS guarantees and network utilization (bandwidth, buuer sizes, etc.) as well as utilization of better models for worst-case description of traac. The performance achieved by the additional complexity introduced with such control mechanisms versus schemes relying solely on statistical guarantees is also a topic worth exploring. In light of recent studies indicating the self-similar nature of VBR video traac and data traac on packet based networks, the design of eecient traf-c management and congestion control mechanisms for broadband networks must take such characteristics into account. Methods and analytical results need to be further developed before one can use self-similar traac descriptors for policing functions and for ensuring fairness without overly constraining a source which is inherently bursty and …