The merging of telephone and computer networks is introducing multiple and diverse types of resources into networks. Users will have access to increasing volumes of devices and information sources through a network interface. Larger bandwidth, more flexibility and better connectivity mean that what used to be implemented on separate dedicated networks are now worth integrating. Centralized services can now take advantage of greater intelligence in the network and become decentralized, more efficient, and better tailored to their users.
This multiple service, multiple resource environment poses a collection of problems not faced in a homogeneous network. Two trends are pushing this: integration of services and service on demand. An integrated services environment is one in which resources are no longer dedicated to a particular service but shared. Furthermore, service on demand implies a bursty, unpredictable usage of resources. Both of these trends pose challenges to the service provider in areas of management, control, and pricing of services and resources.
Suppose you are the system manager of such a network offering a multitude of services based on fixed numbers of multiple types of resources. Figure 1 depicts such a situation, with scripts representing services, cylinders representing resources, and the big stick figure representing you. Suppose a user (upper left in the figure) requests one of those services and offers some $$ for it. What do you do? Do you grant the user the desired service? Questions arise. How does one service affect another through shared resources? Which requests do you accept? Based on what? What do you charge? How many of each resource do you buy? What can you guarantee the users?
In our ongoing research, we have been concerned with such multiple service, multiple resource networks. Our goal is to provide a mathematical foundation with which to analyze these systems. To date, we have proposed a multidimensional Markov chain model of multiple service, multiple resource system and used this model to obtain the sensitivity of throughput of each service type with respect to traffic parameters, and to characterize optimal access control policies. In turn, this work inspired a continuous state space model that admits a simpler threshold type optimal access control policy and makes possible algorithms to determine system dimensioning and pricing.
Throughput in Multiple Service Multiple Resource Communication Networks (with P.P. Varaiya), IEEE Transactions on Communications, vol. 39 no. 8, August 1991, pp. 1216-1222.
Control of Multiple Service Multiple Resource Communication Networks (with P.P. Varaiya), IEEE Transactions on Communications, vol. 42 no. 11, November 1994, pp. 2979-2988.
A Continuous State Space Model of Multiple Service Multiple Resource Communication Networks, IEEE Transactions on Communications, vol. 43 no. 2, February 1995, pp. 477-484.Access Control to Two Multiserver Loss Queues in Series (with C.-Y. Ku), IEEE Transactions on Automatic Control, vol. 42 no. 7, July 1997, pp. 1017-1023.
Near Optimal Admission Control for Multiserver Loss Queues in Series (with C.-Y. Ku), European Journal of Operations Research, vol. 144 no. 1, November 2002, pp. 166-178.
Access Control of Parallel Multiserver Loss Queues (with C.-Y. Ku), Performance Evaluation, vol. 50 no. 4, December 2002, pp 219-231.
Portions of this work were supported by NSF. Any opinions, findings, conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation or IEEE. This material is presented to ensure timely dissemination of scholarly and technical work. Copyright and all rights therein are retained by authors or by other copyright holders. All persons copying this information are expected to adhere to the terms and constraints invoked by each author's copyright. One print or electronic copy may be made for personal use only. Permission must be obtained from the copyright holder for systematic or multiple reproduction, distribution to multiple locations via electronic or other means, duplication of any material in these papers for a fee or for commercial purposes, modification of the content of these papers, reprinting or republishing of this material for advertising or promotional purposes or for creating new collective works for resale or redistribution to servers or lists, and to reuse any copyrighted component of this work in other works.
Scott Jordan (6/24/2005)