[fod] [Fwd: [academic] ITEE seminar : Dr Ben Pourbabai, 10.00AM, Thu 23

Dr. Ben Pourbabai has served the academic community, private sector, and the government in different capacities.

His academic career started at 1984 the Department of Systems Engineering at the University of Arizona. In 1985, he joined NYU’s Stern Graduate School of Business Administration. In 1988, he joined Department of Systems Engineering at USC. In 1989, he joined University of Maryland at College Park, as Director of Center for Production, Operations Management, and Engineering. In 1991, he taught as a distinguished visiting Professor at Tilburg University in the Netherlands. Subsequently, he joined MSCA, a systems integration company with a focus on defense and aerospace sector, as a Vice President. In 1996, he joined US Department of Transportation (USDOT), as a Chief Scientific and Technical Advisor. In his official capacity

· he collaborates with NASA, US Air Force, and aerospace industry on development of advanced aircraft, avionics, and control systems;

he collaborated with DARPA on R&D programs in domains of broadband wireless telecommunication systems, Next Generation Internet, and design of secure broadband-wireless communications infrastructure.

At USDOT, among his executive responsibilities, he coordinates R&D activities of the USDOT’s National Centers of Excellence including: National Center of Excellence for Aviation Operations Research; Center of Excellence in Computational Modeling of Aircraft Structures; and Air worthiness Assurance Center of Excellence, while promoting government-industry-university sponsored research and education.

His research interests include autonomous systems, networking and telecommunication systems, simulation and modeling of complex systems, war gaming, C3I, and avionics. His applied mathematics research interests include stochastic optimization and applied probability.

He has published a significant number of scientific papers in refereed journals and edited a Special Volume for Int. J. of Systems Automation Research on Computer Integrated Manufacturing Systems, and continues to serve as a referee for a number of journals and funding agencies.

Optimal Routing Strategies in Stochastic Communication Networks

Ben Pourbabai

US Department of Transportation

Washington, DC, USA

Abstract

Let G = (A) be a directed network with a source and a sink. Let the flow of units (i.e., packets of data) arriving at each arc (i.e., a device) be a Poisson process; Let n_ij be a random variable, representing the number of units that are found flow along (i, j) A; Let N_ij be the maximum allowable number of units flowing along (i, j) A; Let _ij be the desirable/allowable probability of finding more than N_ij units along (i, j) A; Let _ij be the flow rate along (i, j) A; Let ^*_ij be the minimum allowable flow rate along (i, j) A; Let ^**_ij be the maximum allowable flow rate along (i, j) A; Let x_ij be a random variable, representing the time it takes for an arbitrary unit traveling along (i, j) A, and let it be exponentially distributed; Let _ij =1/ be the processing rate (of device situated) along link (i, j) A; Let c_ij be the cost per unit flow rate along (i, j) A. The problem is to find a minimum cost feasible flow vector of specified value .

Minimize c_{ij ij}

Subject to:

-_{ij + ji} = , if i is source

-_{ij + ji} = 0 , if i is an intermediate point

-_{ij + ji} = - , if i is sink

Pr (n_ij > N_ij ) _ij , for all (i, j) A

^*_ij _ij minimum(_ij ,^**_ij ) , for all (i, j) A

The above network can model any communication and network topology.

In this presentation, initially optimal routing applications of the above model for enterprise-wide communication networks would be delineated. An exact algorithm to solve the above stochastic optimization problem would be proposed. Finally, a number of complementary extensions and applications would be discussed.

[fod] [Fwd: [academic] ITEE seminar : Dr Ben Pourbabai, 10.00AM, Thu 23 May 2002]