Abstract

Motivated by workforce management in the IT services industry, we consider a number of algorithmic and structural issues in stochastic loss networks. The very popular Erlang fixed-point approximation can be shown to provide relatively poor performance estimates. We propose a general algorithm for estimating the stationary loss probabilities in stochastic loss networks based on properties of the exact stationary distribution, which is shown to always converge, exponentially fast, to the asymptotically exact results. We also determine structural properties of the inverse Erlang function characterizing the region of resource capacities that ensures offered traffic will be served within a given set of loss probabilities. Numerical experiments using real-world datasets further investigate various issues of both theoretical and practical interest.

Joint work with Kyomin Jung, Yingdong Lu, Devavrat Shah and Mayank Sharma.

Bio: Mark S. Squillante received the Ph.D. degree from the University of Washington, Seattle, WA, in 1990. He has been a Research Staff Member at the IBM Thomas J. Watson Research Center, Yorktown Heights, NY, since 1990, and an adjunct faculty member at Columbia University, New York, NY, from 1991 through 1996. He currently is a Research Staff Member in the Mathematical Sciences Department. From 1982 to 1985, he was a Member of the Technical Staff at Bell Telephone Laboratories, Murray Hill, NJ. His research interests concern mathematical foundations of the analysis, modeling and optimization of the design and control of stochastic systems, including stochastic processes, probability theory, stochastic optimization and their applications. Dr. Squillante is a Fellow of IEEE, and a member of AMS, ACM, IMS, INFORMS, IFIP W.G. 7.3 and SIAM. He currently serves on the Editorial Boards of Operations Research and Performance Evaluation.

MONDAY, April 28, 2008
Time: 5:00 P.M.
Location: WeH 5302

Refreshments at 4:30, Math Lounge, WeH 6220.